Managing hydrometric data: from ... - The Hydrology Projecthydrology-project.gov.in/PDF/MET_Handbook_180514.pdf · This Hydrology Project Phase II (HPII) Handbook provides guidance

HP II Indian Hydrology Project

Technical Assistance (Implementation Support) and

Management Consultancy

Hydro-Meteorology Handbook:

Precipitation and Climate May 2014

Hydrological Information System May 2014

HP II Last Updated: 19/05/2014 05:01 Filename: MET Handbook.docx

Hydro-Meteorology Handbook: Precipitation and Climate Issue and Revision Record Revision Date Originator Checker Approver Description 0 21/05/14 Helen Houghton-Carr Version for approval 1 2 3



Page i

Contents Contents i Glossary iii 1. Introduction

1.1 HIS Manual 1.2 Other HPI documentation 1.3 Rainfall data in groundwater studies

1 2 3 4

2. The Data Management Lifecycle in HPII 5 2.1 Use of hydrological information in policy and decision-

making 2.2 Hydrological monitoring network design and development 2.3 Data sensing and recording 2.4 Data validation and archival storage 2.5 Data synthesis and analysis 2.6 Data dissemination and publication 2.7 Real-time data

5 6 6 6 7 8 8

3. Hydro-Meteorological Monitoring Stations and Data 10 3.1 Types of hydro-meteorological monitoring station

3.2 Hydro-meteorological monitoring networks 3.3 Site inspections, audits and maintenance 3.4 Data sensing and recording 3.5 Data processing

10 10 14 14 15

4. Rainfall Data Processing and Analysis 18 4.1 Data entry

4.2 Primary validation 4.3 Secondary validation 4.4 Correction and completion 4.5 Compilation 4.6 Analysis

18 21 24 29 33 37

5. Snow Data Processing and Analysis 40 5.1 Snow data in the Hydrology Project

5.2 Data entry 5.3 Primary validation 5.4 Secondary validation 5.5 Analysis

40 41 43 43 44

6. Climate Data Processing and Analysis 46 6.1 Data entry

6.2 Primary validation 6.3 Secondary validation 6.4 Correction and completion 6.5 Analysis

46 49 51 54 56

7. Data Dissemination and Publication 59 7.1 Hydro-meteorological products

7.2 Annual reports 7.3 Periodic reports 7.4 Special reports

59 59 62 63



Page ii

7.5 Dissemination to hydrological data users 63 References 64 Annex I States and agencies participating in the Hydrology Project 65 Annex II Summary of distribution of hard copy of HPI HIS Manual

Surface Water 66

Annex III Summary of distribution of hard copy of HPI HIS Manual

Groundwater 67

List of figures 1.1 Hydrometric information lifecycle 1 4.1 Definition of test and neighbouring stations 26 4.2 Definition sketch for double mass analysis 28 4.3 Example of basin area divided into Theissen polygons 35 4.4 Example of drawing isohyets using linear interpolation 35 List of tables 1.1 HPI hydro-meteorology training modules 4 2.1 Hydro-meteorological data processing timetable for data for

month n

8 3.1 Where to go in the HIS Manual SW for hydro-meteorological

data management guidance: rainfall and snow

11 3.1 cont/ Where to go in the HIS Manual SW for hydro-meteorological

data management guidance: climate and evaporation

12 4.1 Measurement errors for rainfall data 22 5.1 Measurement errors for snow data 44 6.1 Measurement errors for climate data 50



Page iii

Glossary ADCP Acoustic Doppler Current Profiler ARG Autographic Rain Gauge AWS Automatic Weather Station BBMB Bhakra-Beas Management Board CGWB Central Ground Water Board CPCB Central Pollution Control Board CWC Central Water Commission CWPRS Central Water and Power Research Station Div Division DPC Data Processing Centre DSC Data Storage Centre DWLR Digital Water Level Recorder e-GEMS Web-based Groundwater Estimation and Management System

(HPII) eHYMOS Web-based Hydrological Modelling System (HPII) eSWDES Web-based Surface Water Date Entry System in e-SWIS (HPII) e-SWIS Web-based Surface Water Information System (HPII) FCS Full Climate Station GEMS Groundwater Estimation and Management System (HPI) GW Groundwater GWDES Ground Water Data Entry System (HPI) GWIS Groundwater Information System (GPI) HDUG Hydrological Data User Group HIS Hydrological Information System HP Hydrology project (HPI Phase I, HPII Phase II) HYMOS Hydrological Modelling System (HPI) IMD India Meteorological Department Lab Laboratory MoWR Ministry of Water Resources NIH National Institute of Hydrology SRG Standard Rain Gauge Stat Station Sub-Div Sub-Division SW Surface Water SWDES Surface Water Data Entry System (HPI) TBR Tipping Bucket Raingauge ToR Terms of Reference WISDOM Water Information System Data Online Management (HPI) WQ Water Quality



Page iv



Page 1

1. Introduction This Hydrology Project Phase II (HPII) Handbook provides guidance for the management of hydro-meteorological data on rainfall, snow and other climate variables. The data are managed within a Hydrological Information System (HIS) that provides information on the spatial and temporal characteristics of the quantity and quality of surface water, including hydro-meteorology, and groundwater. The information is tuned to the requirements of the policy makers, designers and researchers to provide evidence to inform decisions on long-term planning, design and management of water resources and water use systems, and for related research activities. The Indian States and Central Agencies participating in the Hydrology Project are listed in Annex I. However, this Handbook is also relevant to non-HP States. It is important to recognise that there are two separate issues involved in managing hydro-meteorological information. The first issue covers the general principles of understanding monitoring networks, of collecting, validating and archiving data, and of analysing, disseminating and publishing data. The second covers how to actually do these activities using the database systems and software available. Whilst these two issues are undeniably linked, it is the first – the general principles of data management - that is the primary concern. This is because improved data management practices will serve to raise the profile of Central/State hydrometric agencies in government and in the user community, highlight the importance of hydro-meteorological data for the design of water-related schemes and for water resource planning and management, and motivate staff, both those collecting the data and those in data centres. This Handbook aims to help HIS users locate and understand documents relevant to hydro-meteorology in the library available through the Manuals page on the Hydrology Project website. The Handbook is a companion to the HIS manuals. The Handbook makes reference to the six stages in the hydrometric information lifecycle (Figure 1.1), in which the different processes of data sensing, manipulation and use are stages in the development and flow of information. The cycle and associated HIS protocols are explored more fully in Section 2. Subsequent sections cover different stages of the cycle for different hydro-meteorological variables.

Figure 1.1 Hydrometric information lifecycle (after: Marsh, 2002)



Page 2

1.1 HIS Manual The primary reference sources are the HIS Manual Surface Water (SW) and HIS Manual Groundwater (GW), two of many hundreds of documents generated during Hydrology Project Phase I (HPI) to assist staff working in observation networks, laboratories, data processing centres and data communication systems to collect, store, process and disseminate hydrometric data and related information. During HPI, special attention was paid to the standardisation of procedures for the observation of variables and the validation of information, so that it was of acceptable quality and compatible between different agencies and States, and to facilities for the proper storage, archival and dissemination of data for the system, so that it was sustainable in the long-term. Therefore, the majority of the documents produced under HPI, particularly those relating to fundamental principles, remain valid through and beyond HPII. Some parts of the guides, manuals and training material relating to HPI software systems (SWDES, HYMOS, WISDOM, GWDES, GEMS, GWIS) have been partially or wholly superseded as replacement Phase II systems (e-GEMS, e-SWIS) become active. The HIS Manual SW and HIS Manual GW describe the procedures to be used to arrive at a sound operation of the HIS in regard to rainfall, snow and climate data. The HIS Manual SW and HIS Manual GW each consist of 10 volumes. Each volume contains one or more of the following manuals, depending on the topic: • Design Manual (DM) - procedures for the design activities to be carried out for the

implementation and further development of the HIS. • Field Manual (FM) or Operation Manual (OM) – detailed instructions describing the activities to

be carried out in the field (station operation, maintenance and calibration), at the laboratory (analysis), and at the Data Processing Centres (data entry, validation, processing, dissemination, etc). Each Field/Operation Manual is divided into a number of parts, where each part describes a distinct activity at a particular field station, laboratory or data processing centre.

• Reference Manual (RM) - additional or background information on topics dealt with or

deliberately omitted in the Design, Field and Operation Manuals. Those HIS Manual SW/GW volumes relevant to rainfall and climate are: SW/GW Volume 1: Hydrological Information System: a general introduction to the HIS, its structure, HIS job descriptions, Hydrological Data User Group (HDUG) organisation and user data needs assessment. The content of the SW and GW volumes is identical.

• Design Manual • Field Manual

Part II: Terms of Reference for HDUG Part III: Data needs assessment

SW/GW Volume 2: Sampling Principles: units, principles of sampling in time and space and sampling error theory. The content of the SW and GW volumes is identical.

• Design Manual SW/GW Volume 3: Hydro-meteorology: network design, implementation, operation and maintenance. The content of the SW and GW volumes is identical.

• Design Manual • Field Manual

Part I: Network design and site selection Part II: Standard raingauge station (SRG) operation and maintenance



Page 3

Part III: Autographic raingauge station (ARG or TBR (and SRG)) operation and maintenance

Part IV: Full climate station (FCS) operation and maintenance Part V: Field inspections, audits maintenance and calibration

• Reference Manual SW Volume 8: Data processing and analysis: specification of procedures for Data Processing Centres (DPCs).

• Operation Manual Part I: Data entry and primary validation Part II: Secondary validation Part III: Final processing and analysis Part IV: Data management

GW Volume 8: Data processing and analysis

• Operation Manual Part V: Groundwater Year Book

SW Volume 10: Surface Water protocols: outline of protocols for data collection, entry, validation and processing, communication, inter-agency validation, data storage and dissemination, HIS training and management.

• Operation Manual Data entry forms

In this Handbook, individual parts of the HIS Manual SW/GW are referred to according to the nomenclature “SW/GWvolume-manual(part)” e.g. SW Volume 3: “Hydro-meteorology” Field Manual Part II: “Standard raingauge station (SRG) operation and maintenance” is referred to as SW3-FM(II), and GW Volume 8: “Data processing and analysis” Operation Manual Part V: “Groundwater Year Book” is referred to as GW8-OM(V). A hard copy of the relevant manuals should be available for the locations listed in Annex II. For example, a hard copy of SW3-FM(II) should be available at all meteorological stations where rainfall measurement with an SRG takes place. Similarly, SW8-OM(I) should be available at all Data Processing Centres where data entry and primary validation take place. As noted, there is some inevitable overlap and repetition between the HIS Manual SW and the HIS Manual GW (e.g. Volume 3). In the following sections of this Handbook, reference is generally made only to the HIS Manual SW, as the majority of hydro-meteorological reference material is incorporated in here, unless there is important additional information in the HIS Manual GW. 1.2 Other HPI documentation Other HPI documents of relevance to hydro-meteorology include: • The e-SWIS software manual, and the SWDES and HYMOS software manuals - although

SWDES and HYMOS are being superseded by e-SWIS in HPII, to promote continuity, e-SWIS contains eSWDES and eHYMOS modules.

• “Illustrations: hydrological observations” – an illustrative booklet demonstrating how to make

measurements of rainfall, water level and flow at stations, and also how to carry out an inspection at those stations.



Page 4

Table 1.1 HPI hydro-meteorology training modules Topic Module Title Meteorology 07 How to make data entry for rainfall data

08 How to carry out primary validation for rainfall data 09 How to carry out secondary validation of rainfall 10 How to correct and complete rainfall data 11 How to compile rainfall data 12 How to analyse rainfall data 13 How to report on rainfall data 14 MISSING – How to process evaporation data 15 How to make data entry for climatic data 16 How to carry out primary validation for climatic data 17 How to carry out secondary validation of climatic data 18 MISSING – How to correct and complete climatic data 19 How to analyse climatic data 20 How to report on climatic data

Hydrometry 43* Statistical Analysis with Reference to Rainfall & Discharge Data 44* How to carry out correlation and spectral analysis 45* How to review Monitoring Networks

* Hydrometry modules also relevant to Hydro-Meteorology. 43 and 44 present statistical analysis techniques as applied to, say, analyse rainfall data. • “Surface Water O&M norms” – a maintenance guide for hydro-meteorology, stage-discharge

and water quality instrumentation and equipment. • “Surface Water Yearbook” – a template for a Surface Water Yearbook published at State level. • Hydro-meteorology training modules – these relate to the entry, primary and secondary

validation, processing, analysis and reporting of rainfall and climate data using SWDES and HYMOS (see Table 1.1). Their contents have been largely incorporated into this Handbook as the underlying principles for data validation and analysis remain valid.

1.3 Rainfall data in groundwater studies Access to rainfall data is important in interpretation of groundwater level data, and for balancing recharge, discharge and storage of groundwater systems. GW3-DM, FM and RM describe the design, implementation, operation and maintenance of hydrometeorological networks, and rainfall data may be stored in the groundwater data processing and analysis software e-GEMS. However, subsequent data processing and analysis of rainfall data are covered only in SW8-OM and the surface water software e-SWIS has a wider range of validation and manipulation tools for rainfall data than e-GEMS. Therefore, it is recommended to carry out the majority of rainfall data processing and analysis in e-SWIS, and then export final datasets from e-SWIS, for import to e-GEMS.



Page 5

2. The Data Management Lifecycle in HPII Agencies and staff with responsibilities for hydrometric data have a pivotal role in the development of hydro-meteorological information, through interacting with data providers, analysts and policy makers, both to maximise the utility of the datasets and to act as key feedback loops between data users and those responsible for data collection. It is important that these agencies and staff understand the key stages in the hydrometric information lifecycle (Figure 1.1), from monitoring network design and data measurement, to information dissemination and reporting. These later stages of information use also provide continuous feedback influencing the overall design and structure of the hydrometric system. While hydrometric systems may vary from country to country with respect to organisation set-ups, observation methods, data management and data dissemination policies, there are also many parallels in all stages of the cycle. 2.1 Use of hydro-meteorological information in policy and decision-making The objectives of water resource development and management in India, based on the National Water Policy and Central/State strategic plans, are: to protect human life and economic functions against flooding; to maintain ecologically-sound water systems; and to support water use functions (e.g. drinking water supply, energy production, fisheries, industrial water supply, irrigation, navigation, recreation, etc). These objectives are linked to the types of data that are needed from the HIS. SW1-DM Chapter 3.3 presents a table showing HIS data requirements for different use functions on page 19. In turn, these use functions lead to policy and decision-making uses of HIS data, such as: water policy, river basin planning, water allocation, conservation, demand management, water pricing, legislation and enforcement. Hence, freshwater management and policy decisions across almost every sector of social, economic and environmental development are driven by the analysis of hydrometric information. Its wide-ranging utility, coupled with escalating analytical capabilities and information dissemination methods, have seen a rapid growth in the demand for hydrometric data and information over the first decades of the 21st century. Central/State hydrometric agencies and international data sharing initiatives are central to providing access to coherent, high quality hydrometric information to a wide and growing community of data users. Hydrological data users may include water managers or policymakers in Central/State government offices and departments, staff and students in academic and research institutes, NGOs and private sector organisations, and hydrology professionals. An essential feature of the HIS is that its output is demand-driven, that is, its output responds to the hydrological data needs of users. SW1-FM(III) presents a questionnaire for use when carrying out a data needs assessment to gather information on the profile of data users, their current and proposed use of surface water, groundwater, hydro-meteorology and water quality data, their current data availability and requirements, and their future data requirements. Data users can, through Central/State hydrometric agencies, play a key role in improving hydrometric data, providing feedback highlighting important issues in relation to records, helping establish network requirements and adding to a centralised knowledge base regarding national data. By embracing this feedback from the end-user community, the overall information delivery of a system can be improved. A key activity within HPII was a move towards greater use of the HIS data assembled under HPI. Two examples of the use of HIS data include the Purpose-Driven Studies (PDS) and the Decision Support Systems (DSS) components of HPII. See the Hydrology Project website for more information about DSS and PDS, and access to PDS reports. The 38 PDS, which were designed, prepared and implemented by each of the Central/State



Page 6

hydrometric agencies, are small applied research projects to investigate and address a wide range of real-world problems and cover surface water, groundwater, hydro-meteorology and water quality topics. Some examples of projects include optimisation of the river gauging station and raingauge networks in Maharashtra (PDS number SW-MH-1), and a snowmelt runoff study in the Beas basin (PDS number SW-NIH-1). The PDS utilise hydrometric data and products developed under HPI, supplemented with new data collected during HPII. Two separate DSS programmes were set up under HPII. One, for all participating implementing agencies, called DSS Planning (DSS-P), has established water resource allocation models for each State to assist them to manage their surface and groundwater resources more effectively. The other, called DSS Real-Time (DSS-RT) was specifically for the Bhakra-Beas Management Board (BBMB), although a similar DSS-RT study has also now been initiated on the Bhima River in Maharashtra. The DSS programmes have been able to utilise hydrological data assembled under the Hydrology Project to guide operational decisions for water resource management. 2.2 Hydro-meteorological monitoring network design and development Section 3.2 of this Handbook outlines the design and development of hydro-meteorological monitoring networks. Networks are planned, established, upgraded and evolved to meet a range of needs of data users and objectives, most commonly water resources assessment and hydrological hazard mitigation (e.g. flood forecasting). It is important to ensure that the hydro-meteorological, surface water, groundwater and water quality monitoring networks of different agencies are integrated as far as possible to avoid unnecessary duplication. In particular, a raingauge network should have sufficient spatial coverage that all flow monitoring stations are adequately covered. Integration of networks implies that networks are complimentary and that regular exchange of data takes place to produce high quality validated datasets. Responsibility for maintenance of Central/State hydrometric networks is frequently devolved to a regional (Divisional) or sub-regional (Sub-Divisional) level. 2.3 Data sensing and recording Sections 3.1 to 3.4 of this Handbook review hydro-meteorological monitoring networks and stations, maintenance requirements and measurement techniques. Responsibility for operation of Central/State hydro-meteorological monitoring stations is frequently devolved to a regional (Divisional) or sub-regional (Sub-Divisional) level. However, it is important that regular liaison is maintained between sub-regions and the Central/State agencies through a combination of field site visits, written guidance, collaborative projects and reporting, in order to ensure consistency in data collection and initial data processing methods across different sub-regions, maintain strong working relationships, provide feedback and influence day-to-day working practice. Hence, the Central/State agencies are constantly required to maintain a balance of knowledge between a broad-scale overview and regional/sub-region hydro-meteorological awareness. Operational procedures should be developed in line with appropriate national and international (e.g. Indian, ISO, WMO) standards (e.g. WMO Report 168 “Guide to Hydrological Practices”). For the Hydrology Project, field data from observational stations are required to be received at Sub-Divisional office level by the 5th working day of the following month (SW10-OM Protocols and Procedures). 2.4 Data validation and archival storage The quality control and long-term archiving of hydro-meteorological data represent a central function of Central/State hydrometric agencies. This should take a user-focused approach to



Page 7

improving the information content of datasets, placing strong emphasis on maximising the final utility of data e.g. through efforts to improve completeness and fitness-for-purpose of Centrally/State archived data. Section 3.5 of this Handbook summarises the stages in the processing of hydro-meteorological data. Sections 4 to 6 of this Handbook cover the process from data entry through primary and secondary validation to correction and completion of data, and also compilation and analysis of data (Section 2.5), for rainfall, snow and climate data, respectively. During all levels of validation, staff should be able to consult station metadata records detailing the history of the site and its hydro-meteorological performance, along with topographical and isohyetal maps and previous quality control logs. Numerical and visual tools available at different phases of the data validation process, such as versatile hyetograph plotting and manipulation software to enable comparisons between different near-neighbour rainfall measurement sites, assessment of basin rainfall input hyetographs and assessment of time series statistics greatly facilitate validation. High-level appraisal by Central/State staff, examining the data in a broader spatial context, can provide significant benefits to final information products. It also enables evaluation of the performance of sub-regional data providers, individual stations or groups of stations, which can focus attention on underperforming sub-regions and encourage improvements in data quality. A standardised data assessment and improvement procedure safeguards against reduced quality, unvalidated and/or unapproved data reaching the final data archive from where they can be disseminated. However, Marsh (2002) warns of the danger of data quality appraisal systems that operate too mechanistically, concentrating on the separate indices of data quality rather than the overall information delivery function. For the Hydrology Project, the timetable for data processing is set out in SW10-OM Protocols and Procedures, and summarised in Table 2.1 of this Handbook. Data entry and primary validation of field data from observational stations is required to be completed at Sub-Divisional office level by the 10th working day of the following month (e.g. for June data by 10th working day in July), ready for secondary validation by State offices. Initial secondary validation, in State DPCs for State data, and IMD local offices for IMD data, should be completed by the end of that month (e.g. for June data by 31st July). Some secondary validation will not be possible until the end of the hydrological year when the entire year’s data can be reviewed in a long-term context, and compared with IMD data, so data should be regarded as provisional approved data until then (e.g. for June data by the end of the hydrological year plus 3 months), after which data should be formally approved and made available for dissemination to external users. At certain times of year (e.g. during the monsoon season), this data processing plan may need to be compressed, so that validated hydro-meteorological data are available sooner. 2.5 Data synthesis and analysis Central/State hydrometric agencies play a key role in the delivery of large-scale assessments of rainfall data and other climate data. Through their long-term situation monitoring, they are often well placed to conduct or inform scientific analysis at a State, National or International level, and act as a source of advice on data use and guidance on interpretation of precipitation patterns. This is especially true in the active monitoring of the State or National situation or the assessment of conditions at times of extreme events (e.g. monsoonal rains, droughts) where agencies may be asked to provide input to scientific reports and research projects, as well as informing policy decisions, media briefings, and increasing public understanding of the state of the water environment. Sections 4 to 6 of this Handbook cover compilation and analysis of data, as well as the process from data entry through primary and secondary validation to correction and completion of data (Section 2.4), for rainfall, snow and climate data, respectively.



Page 8

Table 2.1 Hydro-meteorological data processing timetable for data for month n Activity Responsibility Deadline Rainfall, snow and climate data Data receipt Sub-Divisional office 5th working day of month n+1 Data entry Sub-Divisional/Divisional office 10th working day of month n+1 Primary validation Sub-Divisional/Divisional office 10th working day of month n+1 Secondary validation State DPC

State DPC Initial - end of month n+1 Final – end of hydrological year + 3 months

Correction and completion State DPC State DPC

Initial - end of month n+1 Final – end of hydrological year + 3 months

Compilation State DPC As required Analysis State DPC As required Reporting State DPC At least annually Data requests State DPC 95% - within 5 working days

5% - within 20 working days Interagency validation IMD At least 20% of State stations, on

rolling programme, by end of hydrological year + 6 months

2.6 Data dissemination and publication One of the primary functions of Central/State hydrometric agencies is to provide comprehensive access to information at a scale and resolution appropriate for a wide range of end-users. However, improved access to data should be balanced with a promotion of responsible data use by also maintaining end-user access to important contextual information. Thus, the dissemination of user guidance information, such as composite summaries that draw users’ attention to key information and record caveats (e.g. monitoring limitations, high levels of uncertainty regarding specific rainfall event accuracy, major changes in hydro-meteorological setup), is a key stewardship role for Central/State hydrometric agencies, as described in Section 7 of this Handbook. For large parts of the 20th century the primary data dissemination route for hydrometric data was via annual hardcopy publications of data tables i.e. yearbooks. However, the last decade or so has seen a shift towards more dynamic web-based data dissemination to meet the requirement for shorter lag-time between observation and data publication and ease of data re-use. Like many countries, India now uses an online web-portal as a key dissemination route for hydrometric data and associated metadata which provides users with dynamic access to a wide range of information to allow selection of stations. At least 95% of data requests from users should be processed within 5 working days. More complex data requests should be processed within 20 working days. 2.7 Real-time data During HPII many implementing agencies developed low cost real-time data acquisition systems, feeding into bespoke databases and available on agency websites. Such systems often utilise short time interval recording of data e.g. 5 minutes, 15 minutes, etc. In some instances, agencies are taking advantage of the telemetry aspect of real-time systems as a cost-effective way of acquiring data from remote locations. However, for some operational purposes (e.g. real-time flood forecasting, reservoir operation, etc), real-time data may need to be used immediately.



Page 9

Real-time data should go through some automated, relatively simple data validation process before being input to real-time models e.g. checking that each incoming data value is within pre-set limits for the station, and that the change from preceding values is not too large. Where data fall outside of these limits, they should generally still be stored, but flagged as suspect, and a warning message displayed to the model operators. Where suspect data have been identified, a number of options are available to any real-time forecasting or decision support model being run, and the choice will depend upon the modelling requirements. Whilst suspect data could be accepted and the model run as normal, it is more common to treat suspect data as missing or to substitute them with some form of back-up, interpolated or extrapolated data. This is necessary for hydrometric agencies to undertake some of their day-to-day functions and, in such circumstances, all the data should be thoroughly validated as soon as possible, according to the same processing timetable and protocols as other climate data. Real-time data should also be regularly transferred to the e-SWIS database system, through appropriate interfaces, in order to ensure that all hydro-meteorological data are stored in a single location and provide additional back-up for the real-time data, but also to provide access to the data validation tools available through the eSWDES and eHYMOS modules of e-SWIS.



Page 10

3. Hydro-Meteorological Monitoring Stations and Data 3.1 Types of hydro-meteorological monitoring station SW3-FM(I) Chapter 2.1 lists different types of hydro-meteorological stations and instruments which measure various rainfall and climate variables. For each of these, Table 3.1 (two parts) lists the relevant section in the HIS Manual SW for detailed information on design and installation, maintenance, measurement, data entry, primary and secondary validation, correction and completion of data, compilation and analysis of data, and reporting. Stations include: • SRG – a rainfall station equipped with a standard or non-recording raingauge. Additional

information on design and installation (Table 3.1) includes the appropriate capacity of raingauge containers for different Indian States.

• ARG – a rainfall station with an autographic or recording raingauge, which will also have an

SRG for check purposes. • TBR – a tipping bucket raingauge is a type of ARG, often connected to a data logger, which will

also have an SRG for check purposes. Additional information on design and installation (Table 3.1) includes the relative advantages of ARGs and TBRs, compared to each other and to SRGs.

• Snow stations – a type of station not included in HPI, where observation are made of: Snowfall since the last observation; Total depth of snow on the ground (i.e. the depth of the snowpack); Snow-water equivalent (SWE i.e. the depth of liquid precipitation contained in that snowfall

and/or the snowpack). • FCS – a full climate station where, in addition to rainfall, a comprehensive range of other

climate variables are observed for direct measurement of evaporation and/or for indirect estimation of evaporation: Pan-evaporation (direct measurement) using a pan-evaporimeter; Temperature of pan water; Sunshine duration using a sunshine recorder; Air temperature using thermometers and optional thermograph; Humidity using thermometers and optional hygrograph; Wind speed and direction using an anemometer and wind vane; Atmospheric pressure using a barometer and/or barograph.

• AWS – an automatic weather station is an FCS where the climate variables are observed by

automatic/recording means. These were not included in HPI documentation. AWS were not included in HPI.

A set of specifications for hydrometric equipment was compiled under HPI and updated under HPII. The specifications, which are downloadable from the Hydrology Project website, provide a guideline for procurement (with examples of some procurement templates and documents also on the Hydrology Project website). 3.2 Hydro-meteorological monitoring networks Monitoring networks should be considered to be dynamic entities. It is important that the current utility of well-established monitoring networks is periodically assessed to ensure that they continue



Page 11

Table 3.1 Where to go in the HIS Manual SW for hydro-meteorological data management guidance: rainfall and snow Instrument/ Variable

Design & Installation

Maintenance Measurement Data entry Primary Validation

Secondary Validation

Correction & Completion

Compilation Analysis Reporting

SRG SW3-DM 6.2.1, 8.2.2

SW3-FM(II) 1.3 SW3-FM(V) 2.2, 3.2

SW3-FM(II) 1.2

SW8-OM(I) 4.4, 4.5

SW8-OM(I) 5

SW8-OM(II) 2

SW8-OM(II) 3

SW8-OM(II) 4

SW8-OM(III) 4

SW8-OM(III) 9

ARG SW3-DM 6.2.2, 8.2.3

SW3-FM(III) 2.3 SW3-FM(V) 2.2, 3.3

SW3-FM(III) 2.2.2

SW8-OM(I) 4.6

SW8-OM(I) 5

SW8-OM(II) 2

SW8-OM(II) 3

SW8-OM(II) 4

SW8-OM(III) 4

SW8-OM(III) 9

TBR SW3-DM 6.2.3, 8.2.4

SW3-FM(III) 3.3 SW3-FM(V) 2.2, 3.4

SW3-FM(III) 2.3.2

SW8-OM(I) 4.7

SW8-OM(I) 5

SW8-OM(II) 2

SW8-OM(II) 3

SW8-OM(II) 4

SW8-OM(III) 4

SW8-OM(III) 9

Snow Not included in HPI



Page 12

Table 3.1 Where to go in the HIS Manual SW for hydro-meteorological data management guidance: climate and evaporation Instrument/ Variable

Design & Installation

Maintenance Measurement Data entry Primary Validation

Secondary Validation

Correction & Completion

Compilation Analysis Reporting

FCS SW3-DM 6.3, 8.3

SW8-OM(I) 6.4, 6.5

SW8-OM(I) 7

SW8-OM(II) 5

SW8-OM(II) 6

SW8-OM(II) 5

SW8-OM(III) 11

Pan-evaporation

SW3-DM 6.3.5, 8.3.7

SW3-FM(IV) 5.2 SW3-FM(V) 2.2, 3.9

SW3-FM(IV) 5.2 SW8-OM(I) 7.8.2

SW8-OM(I) 6.4, 6.5

SW8-OM(I) 7.8

SW8-OM(II) 5

SW8-OM(II) 6.7

SW8-OM(II) 5.2

SW8-OM(III) 11

Penman evapotran-spiration

SW3-DM 2.3 SW8-OM(II) 5.3

SW8-OM(III) 11

Sunshine duration

SW3-DM 6.3.1, 8.3.8

SW3-FM(IV) 6.3 SW3-FM(V) 2.2, 3.10

SW3-FM(IV) 6.2 SW8-OM(I) 7.7.2

SW8-OM(I) 6.7

SW8-OM(I) 7.7

SW8-OM(II) 5

SW8-OM(II) 6.6

Temperature SW3-DM 6.3.2, 8.3.4-5

SW3-FM(IV) 3.1.3, 3.2.3 SW3-FM(V) 2.2, 3.6, 3.7

SW3-FM(IV) 3.1.2, 3.2.2 SW8-OM(I) 7.3.2

SW8-OM(I) 6.4, 6.5, 6.6

SW8-OM(I) 7.3

SW8-OM(II) 5

SW8-OM(II) 6.2

Humidity SW3-DM 6.3.3, 8.3.6

SW3-FM(IV) 4.2.3 SW3-FM(V) 2.2, 3.8

SW3-FM(IV) 4.1, 4.2.2 SW8-OM(I) 7.3.2, 7.4.2

SW8-OM(I) 6.4, 6.5, 6.6

SW8-OM(I) 7.3, 74

SW8-OM(II) 5

SW8-OM(II) 6.3

Wind speed & direction

SW3-DM 6.3.4, 8.3.3

SW3-FM(IV) 1.1.3, 1.2.3 SW3-FM(V) 2.2, 3.5

SW3-FM(IV) 1.1.2, 1.2.2 SW8-OM(I) 7.5.2

SW8-OM(I) 6.4, 6.5

SW8-OM(I) 7.5

SW8-OM(II) 5

SW8-OM(II) 6.4

Atmospheric pressure

SW3-DM 6.3.6, 8.3.9

SW3-FM(IV) 7.2

SW3-FM(IV) 7.2 SW8-OM(I) 7.6.2

SW8-OM(I) 6.6

SW8-OM(I) 7.6

SW8-OM(II) 5

SW8-OM(II) 6.5

AWS Not included in HPI



Page 13

to meet changing requirements and to optimise the information they deliver. Network reviews should be done in collaboration with other agencies. SW3-FM(I) Chapter 1 and SW3-DM Chapter 3 describe network design and optimisation for monitoring rainfall and other climate variables. This is a multi-step process comprising: 1. Identification of hydrological data users and their data needs to understand what data are

required and at what frequency. 2. Definition of the purposes and objectives of the network in order to fulfill the hydrological data

need, and evaluation of the consequences of not meeting those targets, to inform a prioritisation of objectives in case of budget constraints.

3. Evaluation of the existing and required network densities, using an effectiveness measure

which takes into account the spatial and temporal correlation of the variables. This step and steps 4 and 5 may involve the development of regionalisation and network optimisation techniques (e.g. Institute of Hydrology, 1999; Hannaford et al., 2013).

4. Evaluation of the existing network versus the required one in relation to network density,

purposes and objectives, distribution with respect to surface water and groundwater monitoring networks, adequacy of existing equipment and operational procedures, and possible improvements to existing network.

5. Site and equipment selection i.e. the identification of gaps in the existing network if it is

inadequate to meet the purposes and objectives. This may require the collection of maps and background information to inform the revised network design.

6. Estimation of overall costs of installing, operating and maintaining the existing and new sites.

Achieving an optimum network design may involve an iterative process, repeating some or all of steps 3 to 6, until a satisfactory outcome is reached.

7. Preparation of phased implementation plan for optimum network that is prioritised, realistic and

achievable in the time scales allowed. 8. Selection of sites. SW3-FM(I) devotes Chapter 2 to this topic, identifying the factors that

should be taken into consideration to ensure long-term reliable data. These include: technical (positioning to minimise estimation errors and optimise integration with surface water and groundwater networks); environmental (topography around site, exposure conditions at site, future development near site, vegetation at and near site, proximity of water bodies, no water-logging at site); logistical (accessibility, communication, staffing); security (location of site, design of site, staffing); legal (land acquisition, rights of passage); and financial (cost of land, cost of civil works, equipment procurement, calibration and maintenance, operating costs, staffing) aspects. Site selection should be carried out in collaboration with IMD and should involve a site visit, which may reveal that the desired location is unsuitable, and an alternative site may need to be considered.

The selection of appropriate locations for snow stations (also known as snow courses) may be challenging because of terrain and wind effects. In flat, open areas, it is desirable to have snow stations in typical landscapes, such as in open fields and forests, with different snow accumulation conditions. In mountainous areas, additional criteria may apply:

• At sites sufficiently accessible to ensure continuity of surveys • At elevations and exposures where there is little or no melting prior to the peak

accumulation • At a site having protection from strong wind movement



Page 14

• In forested areas where the sites can be located in open spaces sufficiently large so that snow can fall to the ground without being intercepted by the trees

9. Establishment of a framework for periodic network reviews (e.g. after 3 years or sooner if new

data needs develop) i.e. starting this process again from step 1. As an example of the theory and practical aspects of network design, SW3-RM presents a pilot study for designing a raingauge network for two sub-basins of the Mahanadi river basin in Orissa. A good example of a monitoring network review under HPII is the Purpose Driven Study (PDS) on optimisation of the river gauging station and raingauge networks in Maharashtra (PDS number SW-MH-1). For more detailed information see: SW2-DM Chapter 7 which provides some generic guidance on types of network and the steps in network design; SW2-DM Chapters 3.2.1 to 3.2.6 which describe classification of stations and offer some examples of types of network; and Hydro-Meteorology Training Module 45 “How to review monitoring networks”. 3.3 Site inspections, audits and maintenance Regular maintenance of equipment, together with periodic inspections and audits, ensures collection of good quality data and provides information that may assist in future data validation queries. Table 3.1 lists the relevant section in the HIS Manual SW for maintenance of the different types of hydro-meteorological stations and instruments. Whilst this topic is largely covered in different parts of SW3-FM(II)-(V), information is collated together in the document “Surface Water O&M norms” which is a maintenance guide for hydro-meteorology, stage-discharge and water quality instrumentation and equipment. Maintenance and calibration requirements depend to a large extent of the type of station, instruments and equipment so are often site-specific. A supply of appropriate spare parts should be kept on site and/or taken on station visits in case they are needed. SW3-FM(V) Annex II lists maintenance norms for hydro-meteorological stations, including maintenance of civil works, maintenance of equipment, costs of consumable items and payments to staff (where the costs should be regarded as out of date). Formal inspections cum audits are carried out, and station log sheets completed, at a frequency dependent on the importance of the station, the type of station and the time of year and will typically vary between monthly and annually as set out in SW3-FM(V) Chapter 1, with station log sheets for inspections of rainfall and climate stations in FM(V) Annex I. Activities may include: checking the performance of and motivating the field staff; and identifying existing or potential problems with the site, instruments, equipment and observation procedures at an early stage so they can be rectified. However, a brief inspection of the site and instrumentation should be made every day that somebody is on site. 3.4 Data sensing and recording Table 3.1 lists the relevant section in the HIS Manual SW for operational instructions on the measurement of rainfall and other climate variables at hydro-meteorological stations. Note that there is some overlap between SW3-FM and SW3-DM, and between the network design and site selection topic (covered in Section 3.2 of this Handbook) and data measurement. See also the document “Illustrations: hydrological observations” which demonstrates how to make measurements of rainfall, water level and flow at stations, and also how to carry out an inspection at those stations.



Page 15

At an FCS, the observations of the various climate variables are made once or twice a day in a prescribed sequence commencing from 10 minutes preceding the scheduled hours i.e. 08:20 for 08:30 readings, and 17:20 for 17:30 readings. The sequence is: wind instruments, raingauges, thermometers, evaporation, radiation, and culminating with atmospheric pressure at exactly 08:30 and 17:30. Charts on autographic instruments are changed daily during the 08:30 reading, except the sunshine recorder card which is changed during the 17:30 reading or after sunset, whichever is later. Hourly values are abstracted from autographic charts and tabulated daily. Observers at an FCS should also make observations of the depth of snow on the ground, should it occur. For stations with only an SRG, the measurement is made at 08:30 IST only, though more frequent observations are required during heavy rain to avoid overflow due to the limited capacity of the raingauge container. SRGs at FCSs are read twice daily along with the other climate variables. ARG chart recorders are digitised at 1-hour or 15-minute time intervals depending on what is most appropriate for the location and the intensity of the rainfall (SW3-DM Chapter 5.2). TBRs with data loggers can operate in time mode where the number of tips in a pre-set time interval (e.g. 1 hour, 15 minutes, etc) are recorded, or in event mode where the times of every tip are recorded, thereby producing a more flexible record for subsequent analysis. However, note that event mode data cannot currently be stored in e-SWIS. At snow stations, observations are made of the snowfall since the last observation, the total depth of snow on the ground (i.e. the depth of the snowpack) and the snow-water equivalent (SWE i.e. the depth of liquid precipitation contained in that snowfall and/or the snowpack). Measurements may be made daily or sub-daily. The accuracy of measurements of snowfall, snow depth and SWE depends on the graduations of the scales being used, and on instrumental and subjective errors. At some snow stations, data are augmented by regular measurements of sunshine, temperature, humidity, wind speed and direction, and atmospheric pressure. The extent of snow cover is usually made from one or a combination of field observations, aerial survey data and satellite imagery. The observer should always note any occurrences which may influence the climate variables as observed by the instruments. These may include: damage to the equipment for a specified reason. The observer should also note any maintenance activities carried out at the monitoring site (e.g. change batteries, clean sensor, etc). The observer should double-check that that any manual reading is taken correctly, and transcribed correctly (e.g. decimal point in right place). If the reading is later transferred to another document (e.g. hand copied or typed in, or abstracted from a chart), the observer should always check that this has been done correctly. An experienced and suitably qualified observer should compare measurements with equivalent ones from earlier that day or from the day before, if available, as an additional form of checking. However, the observer should not, under any circumstances, retrospectively alter earlier readings or adjust current readings, but should simply add an appropriate comment. Data collected in the field are delivered to a Data Processing Centre (DPC) on a variety of media, including handwritten forms and notebooks, charts and digital data. 3.5 Data processing SW8-OM(IV) Chapter 2 sets out the steps in processing of hydro-meteorological data, which starts with preliminary checking in the field, as described in Section 3.4 of this Handbook, through receipt of raw field data at a DPC, through successively higher levels of validation in State and Central DPCs, before data are fully validated and approved in the National database. Validation ensures that the data stored are as complete and of the highest quality as possible by: identifying errors and sources of errors to mitigate them occurring again, correcting errors where possible, and



Page 16

assessing the reliability of data. It is important for staff to be aware of the different errors that may occur as described in SW8-OM(IV) Chapter 2.5.1. Hydro-meteorological data validation is split into two stages: primary and secondary. Validation is very much a two-way process, where each step feeds back to the previous step any comments or queries relating to the data provided. The diverse hydrological environments found in India mean that staff conducting data validation should be familiar with the expected climate in order to identify potentially anomalous behaviour. SW3-RM presents some tabular and graphical summary data for selected coastal and hill FCSs in the HPI Indian States for the period 1931-60, to give an insight into typical annual variation of the climate variables. The data processing steps comprise: 1. Receipt of data according to prescribed target dates. Rapid and reliable transfer of data is

essential, using the optimal method based on factors such as volume, frequency, speed of transfer/transmission and cost. Maintenance of a strict time schedule is important because it gives timely feedback to monitoring sites, it encourages regular exchanges between field staff, Sub-Divisional offices, State and Central agencies, it creates continuity of processing activities at different offices, and it ensures timely availability of final (approved) data for use in policy and decision-making.

2. Entry of data to computer, using the eSWDES module of e-SWIS, is primarily done at a Sub-

Divisional office level where staff are in close contact to field staff who have made the observations and/or collected the chart or digital data. Historical data, previously only available in hardcopy form, may also be entered this way. Each Central/State agency should have a programme of historical data entry.

3. Primary data validation which should be carried out in State DPCs for State data and IMD local

offices for IMD data, as soon as possible after the observations are made, data extracted from charts, or data downloaded from loggers, using the eSWDES module of e-SWIS. This ensures that any obvious problems (e.g. indicating an instrument malfunction, observer error, etc) are spotted at the earliest opportunity and resolved. Other problems may not become apparent until more data have been collected, and data can be viewed in a longer temporal context during secondary validation.

4. Secondary data validation which should be carried out in State DPCs for State data and IMD

local offices for IMD data, to take advantage of the information available from a large area by focusing on comparisons with the same variable at other good quality, nearby monitoring sites (analogue stations) which are expected to exhibit similar hydrological behaviours (e.g. comparison of cumulative rainfall from two raingauges), uses the eHYMOS module of e-SWIS. States should have access to IMD data during secondary validation, and may receive support from IMD in this activity.

5. Data correction and completion are elements of data validation which are used to infill missing

value, sequences of missing values or correct clearly erroneous values, in order to make the time series as complete as possible. Some suspect (doubtful) data values may still justifiably remain after this stage if correction is not possible so that the original data value remains the best estimate of the true value for that day and time.

6. Data storage. The e-SWIS HIS database, of both approved data and unapproved data

undergoing primary and secondary validation, is backed up automatically. Therefore, there is no need to make regular back-ups, unless any data are stored outside the HIS database, for instance in Excel files or other formats awaiting data entry, or in stand-alone real-time databases – such files should be securely backed up, ideally onto an external back-up device and/or backed up network server, so that there is no risk of data loss. All PCs should have up-to-date anti-virus software.



Page 17

Raw field data, in the form of handwritten forms and notebooks, and charts should also be stored in a secure manner after database entry to ensure that original field data remain available should any problems be identified during validation and analysis. Such hardcopy data should ultimately be securely archived, in the State DPC for State data or IMD local office for IMD data, possibly by scanning documents and storing them digitally.

7. Interagency data validation by IMD – IMD should aim to validate at least 20% of current and

historic data from State hydro-meteorological monitoring stations every year, on a rolling programme, so that IMD has independently validated the data from every State gauge at least once every 5 years. Interagency validation is a 2-way process and IMD should discuss any identified issues and agree final datasets with State DPCs through a 2-way consultative process, to build capacity for data validation within the States.

For rainfall, snow and climate data, Sections 4 to 6 of this Handbook, respectively, cover the process from data entry through primary and secondary validation to correction and completion of data, and also compilation (i.e. the transformation of data observed at one time interval to another time interval e.g. aggregation from daily rainfall to monthly rainfall) and analysis of data.



Page 18

4. Rainfall Data Processing and Analysis 4.1 Data entry 4.1.1 Overview Entry of data to computer is primarily done at a Sub-Divisional office level where staff are in close contact to field staff who have made the observations and/or collected the chart or digital data. Data entry is carried out using e-SWIS, the data entry module of which replicates the SWDES software from HPI, and is referred to as eSWDES. Prior to entry to computer, two manual activities are essential: registration of receipt of the data, and manual inspection of the rainfall charts, forms and notebooks from the field, for complete information and obvious errors. Data entry (see Table 3.1) and primary validation of field data from observational stations is required to be completed at Sub-Divisional office level by the 10th working day of the following month (e.g. for June data by 10th working day in July), ready for secondary validation by State offices. 4.1.2 Manual inspection of field records Prior to data entry to computer an initial inspection of field records is required. This is done in conjunction with notes received from the observation station on equipment problems and faults, missing records or exceptional rainfall. Rainfall sheets and charts are inspected for the following: • Is the station name and code and month and year recorded? • Does the number of record days correspond with the number of days in the month? • Are there some missing values or periods for which rainfall has been accumulated during

absence of the observer? • Have monthly totals of rainfall and rain days been entered? • Have the autographic rainfall hourly totals been extracted? • Is the record written clearly and with no ambiguity in digits or decimal points? Any queries arising from such inspection should be communicated to the observer to confirm ambiguous data before data entry. Any unresolved problems should be noted and the information sent forward with the digital data to Divisional/State offices to assist in initial secondary validation. Any equipment failure or observer problem should be communicated to the supervising field officer for rectification. 4.1.3 Entry of daily rainfall data Using the eSWDES module in e-SWIS, the user selects the correct station and daily series. The screen for entry (or editing) of daily rainfall is displayed, along with the upper warning level used to flag suspect values (which can be altered for different seasons), and the maximum and minimum values for that station. For rainfall the minimum value is 0.0 mm, and a rainy day is defined as that day on which the rainfall is more than 0.0 mm. The user selects the correct year and month, and enters the daily rainfall value recorded at 08:30 for each date, adding comments where appropriate. Negative and non-numerical entries are automatically rejected. For each month, the user also enters the number of rain days, total rainfall and maximum rainfall. The software also calculates the number of rain days, the cumulative rainfall and the maximum rainfall as the user enters the data. Two types of data entry checks are performed for this case of daily rainfall data: • The number of rain days, total monthly rainfall and maximum rainfall entered by the user are

compared with the values calculated by the software. In the case of a mismatch the user is



Page 19

prompted by colour highlighting and can refer back to the field documents to see if there was some error in entering the data. If cumulated values are also available in the field documents, it becomes quicker to isolate the error.

• The entered daily data are compared against the upper warning level and the maximum limit. This identifies potentially suspect values to the user who can refer back to the field documents to see if there was some error in entering the data. If values which exceed the upper warning level and the maximum limit are actually reported in the field documents, the user should add an appropriate comment.

Any mismatch remaining after thorough checking of the field documents must be due to incorrect field computations by the observer and should be communicated to the supervising field officer. The user should also view entered data graphically to identify potentially suspect data not apparent in tabular form, which may reflect an error in data entry. There are three ways in which the entered data can be plotted: daily data for the month, daily data for the year, and monthly totals for the year. Missing data When data are missing, the corresponding cell is left as -999 (not zero) and a comment entered against that day. Accumulated data Where the observer has missed readings over a period of days and an accumulated total is subsequently measured, the cells corresponding to the missed days are left as -999 (not zero) and a comment entered against the date of the accumulation to specify the period over which the accumulation has occurred (e.g. Accumulated from 23 to 27 Sep). There are occasions when the observer is legitimately absent from her/his station, for example on account of sickness. The observer should be encouraged to leave such spaces “Missing” or “Accumulated” rather than guess the missing values. The completion procedures (Section 4.4), based on adjoining information, are better able to estimate such missing values. 4.1.4 Entry of rainfall data at twice daily interval Using the eSWDES module in e-SWIS, the user selects the correct station and twice-daily series. The screen for entry (or editing) of twice-daily rainfall is displayed, along with the upper warning level used to flag suspect values (which can be altered for different seasons), and the maximum and minimum values for that station. For rainfall the minimum value is 0.0 mm, and a rainy day is defined as that day on which the rainfall is more than 0.0 mm. The user selects the correct year and month, and enters the twice-daily rainfall values recorded at 17:30 the previous day and 08:30 for each date, adding comments where appropriate. Negative and non-numerical entries are automatically rejected. For each month, the user also enters the number of rain days, total rainfall and maximum rainfall. The software also calculates the number of rain days, the total daily rainfall, the cumulative rainfall and the maximum rainfall as the user enters the data. Two types of data entry checks are performed for this case of twice-daily rainfall data: • The number of rain days, total daily rainfall, total monthly rainfall and maximum rainfall entered

by the user are compared with the values calculated by the software. In the case of a mismatch the user is prompted by colour highlighting and can refer back to the field documents to see if there was some error in entering the data. If cumulated values are also available in the field documents, it becomes quicker to isolate the error.

• The entered daily data are compared against the upper warning level and the maximum limit. This identifies potentially suspect values to the user who can refer back to the field documents to see if there was some error in entering the data. If values which exceed the upper warning level and the maximum limit are actually reported in the field documents, the user should add an appropriate comment.



Page 20

Any mismatch remaining after thorough checking of the field documents must be due to incorrect field computations by the observer and should be communicated to the supervising field officer. The user should also view entered data graphically to identify potentially suspect data not apparent in tabular form, which may reflect an error in data entry. There are four ways in which the entered data can be plotted: twice daily data for the month, daily data for the month, daily data for the year, and monthly totals for the year. Missing and accumulated data are handled in the same way as for entry of daily rainfall data (Section 4.1.3). 4.1.5 Entry of hourly rainfall data Hourly rainfall data are obtained either from the chart records of ARGs or from the digital data of TBRs. Digital data can also be imported directly, but can undergo entry checks and be viewed graphically using this option. Using the eSWDES module in e-SWIS, the user selects the correct station and hourly series. The screen for entry (or editing) of hourly rainfall is displayed, along with the upper warning level used to flag suspect values (which can be altered for different seasons), and the maximum and minimum values for that station. For rainfall the minimum value is 0.0 mm. The user selects the correct year and month, and enters the hourly rainfall values, with each row corresponding to a different day and each column to a different time, adding comments where appropriate. The rainfall value is entered against the time following the hour in which the rainfall occurred e.g. rainfall falling and recorded from 11:30 to 12:30 is recorded against 12:30. Negative and non-numerical entries are automatically rejected. For each day, the user enters the daily total. For each month, the user also enters the columnar total for each hourly period, the number of rain days, total rainfall and maximum rainfall. The software also calculates the daily and hourly totals, the number of rain days, the cumulative rainfall and the maximum rainfall as the user enters the data. Two types of data entry checks are performed for this case of hourly rainfall data: • The number of rain days, columnar total for each hourly period, total daily rainfall, total monthly

rainfall and maximum rainfall entered by the user are compared with the values calculated by the software. In the case of a mismatch the user is prompted by colour highlighting and can refer back to the field documents to see if there was some error in entering the data. If cumulated values are also available in the field documents, it becomes quicker to isolate the error.

• The entered hourly data are compared against the upper warning level and the maximum limit. This identifies potentially suspect values to the user who can refer back to the field documents to see if there was some error in entering the data. If values which exceed the upper warning level and the maximum limit are actually reported in the field documents, the user should add an appropriate comment.

Any mismatch remaining after thorough checking of the field documents must be due to incorrect field computations by the observer and should be communicated to the supervising field officer. The user should also view entered hourly data for the day and month graphically to identify potentially suspect data not apparent in tabular form, which may reflect an error in data entry. Missing and accumulated data are handled in the same way as for entry of daily rainfall data (Section 4.1.3).



Page 21

4.1.6 Import/entry of digital data Digital data from TBRs can take two forms: time mode where the number of tips in a pre-set time interval (e.g. 1 hour, 15 minutes, etc) are recorded, or in event mode where the times of every tip are recorded, thereby producing a more flexible record for subsequent analysis. TBR data can be imported directly should an appropriate import interface be available (bespoke to each type of data logger), and can undergo entry checks and be viewed graphically as described in Section 4.1.5. Time mode data are imported to an appropriate equidistant time series, whilst event mode data are imported to a non-equidistant time series. 4.2 Primary validation 4.2.1 Overview Primary validation is primarily done at a Sub-Divisional office level where staff are in close contact to field staff who have made the observations and/or collected the chart or digital data. Primary validation is carried out using e-SWIS, the data entry module of which replicates the SWDES software from HPI, and is referred to as eSWDES. Primary validation (see Table 3.1) of field data from observational stations is required to be completed at Sub-Divisional office level by the 10th working day of the following month (e.g. for June data by 10th working day in July), ready for secondary validation by State offices. This time schedule ensures that any obvious problems (e.g. indicating an instrument malfunction, observer error, etc) are spotted at the earliest opportunity and resolved. Other problems may not become apparent until more data have been collected, and data can be viewed in a longer-term context during secondary validation. Primary validation of rainfall data focuses on validation within a single data series by making comparisons between individual observations and pre-set physical limits, and between two measurements of rainfall at a single station (e.g. daily rainfall from an SRG and accumulated daily rainfall from an ARG/TBR). The high spatial and temporal variability of rainfall data compared to other climate variables makes validation of rainfall more difficult. This is particularly the case on the Indian sub-continent, experiencing a monsoon-type climate involving convective precipitation. Examples of many of the techniques described in this section are given in Hydro-Meteorology Training Module 08 “How to carry out primary validation for rainfall data” and Training Module 10 “How to correct and complete rainfall data”. 4.2.2 Typical errors Staff should be aware of typical errors in rainfall measurement, listed in Table 4.1, and these should be considered when interpreting data and possible discrepancies (SW8-OM(I) Chapter 5.2). SRG errors from most of these sources are very difficult to detect from the single record of the standard raingauge, unless there has been a gross error in reading or transcribing the value (Section 4.2.3). Errors are more readily detected if there is a concurrent record from an ARG or TBR. As these too are subject to errors (of a different type), comparisons with the SRG are very important (Section 4.2.4). The final check by comparison with raingauges at neighbouring stations should show up further anomalies, especially for those stations which do not have an ARG or TBR at the site. This is carried out during secondary validation where more gauges are available for comparison (Section 4.3). 4.2.3 Comparison with upper warning level and maximum and minimum limits Both hourly and daily rainfall data should be validated against physical limits, which are required to be quite wide to avoid the possibility of rejecting true extreme values. For rainfall data, the minimum limit is 0.0 mm. The maximum limit will vary spatially over India with climatic region and



Page 22

Table 4.1 Measurement errors for rainfall data SRG measurement errors • Observer reads measuring glass incorrectly • Observer enters amount incorrectly in the field sheet • Observer reads gauge at the wrong time (i.e. the correct amount may thus be allocated to the wrong

day) • Observer enters amount to the wrong day • Observer uses wrong measuring glass (i.e., 200 cm2 glass for 100 cm2 gauge, giving half the true

rainfall, or 100 cm2 glass for 200 cm2 gauge giving twice the true rainfall • Observed total exceeds the capacity of the gauge • Instrument fault - gauge rim damaged so that collection area is affected • Instrument fault - blockage in raingauge funnel so that water does not reach collection bottle and may

overflow or be affected by evaporation • Instrument fault - damaged or broken collector bottle and leakage from gauge ARG measurement errors • Potential measurement faults are primarily instrumental rather than caused by the observer • Funnel is blocked or partly blocked so that water enters the float chamber at a different rate from the rate

of rainfall • Float is imperfectly adjusted so that it syphons at a rainfall volume different from 10 mm • In very heavy rainfall the float rises and syphons so frequently that individual pen traces cannot be

distinguished • Clock stops; rainfall not recorded or clock is either slow or fast and thus timings are incorrect • Float sticks in float chamber; rainfall not recorded or recorded incorrectly • Observer extracts information incorrectly from the pen trace

TBR measurement errors • Funnel is blocked or partly blocked so that water enters the tipping buckets at a different rate from the

rate of rainfall • Buckets are damaged or out of balance so that they do not record their specified tip volume • Reed switch fails to register tips • Reed switch double registers rainfall tips as bucket bounces after tip. (better equipment includes a

debounce filter to eliminate double registration) • Failure of electronics due to lightning strike etc. (though lightning protection usually provided) • Incorrect set up of measurement parameters by the observer or field supervisor orography, and maximum limits for 1-hour and 1-day rainfall should be based of analysis of historic data for the station or IMD maps of 1-hour maximum rainfall and 1-day maximum rainfall. However, validation of rainfall data against a maximum value does not discriminate those comparatively frequently occurring erroneous data which are less than the prescribed maximum limit. In view of this, it is advantageous to consider an upper warning level, which can be employed to filter high data values which are not expected to occur frequently. For daily rainfall data, this limit can be set statistically e.g. to 99th percentile of actual rainfall values excluding zero values. A similar statistic can be employed for obtaining a suitable upper warning level for hourly rainfall data. Setting such warning levels and limits for each station ensures rapid filtering of potentially erroneous values, and should result is not too many and not too few, data values being flagged as suspect. During data entry, hourly and daily data are compared against the upper warning level and the maximum limit. If values which exceed the upper warning level and the maximum limit are actually reported in the field documents, the user should have considered the values suspect and added an appropriate comment, for further attention during primary validation. Data entries which are more than the prescribed upper warning level or the maximum limit may imply that the earlier maximum has indeed been crossed. In such cases, it is expected that there will be a few nearby stations recording similar higher rainfall to support such inferences, and this should be reviewed during secondary validation.



Page 23

4.2.4 Comparison of daily and sub-daily rainfall data For stations with an ARG or a TBR, an SRG is always also available. Thus, daily rainfall data are available from two independent sources. The accumulated daily rainfall from the sub-daily raingauge should be compared with the daily rainfall at the daily raingauge, to check for consistency. Differences which are less than 5% can be attributed to exposure, instrument accuracy and precision in tabulating the analogue records and are ignored, but differences greater than 5% may indicate potential errors and should be examined further. The observation made using the SRG is traditionally regarded as comparatively more reliable. This is based on the assumption that there is higher degree of possibility of malfunctioning of autographic or digital recorders owing to their mechanical and electronic systems. However, significant systematic or random errors are also possible in the SRG as shown in Table 4.1. If the error is in the autographic or digital records, then it should be possible to relate it either to instrumental or observational errors. Moreover, such errors tend to repeat under similar circumstances. Comparison of daily and accumulated sub-daily rainfall data may be carried out in tabular or graphical form, with an additional table column for those days where the discrepancy is more than 5%, or a second graph axis showing percent discrepancy: • Where the recording gauge gives a consistently higher or lower total than the daily gauge, then

the recording gauge could be out of calibration and either tipping buckets (TBR) or floats (ARG) need recalibration → Accept SRG and adjust ARG or TBR

• Where agreement is generally good but difference increases in high intensity rainfall suggests that for the ARG either (i) the syphon is working imperfectly in high rainfall or (ii) the chart trace is too close to distinguish each 10 mm trace (underestimate by multiples of 10 mm), and for the TBR (i) the gauge is affected by bounce sometimes giving double tips → Accept SRG and adjust ARG or TBRG

• Where a day of positive discrepancy is followed by a negative discrepancy and rainfall at the recording gauge was occurring at the observation hour, then it is probable that the observer read the SRG at a different time from the ARG. The sum of SRG readings for successive days should equal the 2-day total for the ARG or TBR → Accept ARG or TBR and adjust SRG

• Where the agreement is generally good but isolated days have significant differences, then the entered hourly data should be checked against the field documents. Entries resulting from incorrect entry are corrected. Check that water added to a TBR for calibration is not included in rainfall total. Otherwise there is probable error in the SRG observation → Accept ARG or TBR and adjust SRG

• Where the values reported for the daily rainfall by the SRG and ARG/TBR correspond exactly for considerable periods, it is conceivable that the observer has forcefully manipulated one or both datasets. Variation must exist due to variance in the catch and instrument and observation variations. Both hourly and daily data should be checked against the field documents to attempt to ascertain which has been manipulated.

In the case where the SRG record is accepted records from autographic gauges at neighbouring stations can be used in conjunction with the SRG at the station to correct the ARG/TBR record at the station. This involves hourly distribution of the daily total from the SRG at the station by reference to the hourly distribution at one or more neighbouring stations. Donor (or base) stations are selected by making comparison of cumulative plots of events in which autographic records are available at both stations and selecting the best available for estimation. If the daily rainfall on the day in question at the station under consideration is Dtest and the hourly rainfall for the same daily period at the selected neighbouring donor station are Hbase,i (i = 1, 24), the hourly rainfall at the station under consideration, Htest,i are obtained as:



Page 24

The procedure may be repeated for more than one neighbouring donor station and the average or resulting hourly totals calculated. In the case where the ARG/TBR record is accepted it can be used to correct the SRG record at the station. The SRG data values are made equal to the accumulated daily or twice-daily totals from the ARG/TBR. Where a shift in one of the data series has been detected, this can be removed after looking at the field documents and corresponding tabulated data and adjusting the data appropriately. The adjusted data should be flagged as corrected and an appropriate comment added. Where a doubtful or incorrect rainfall value is identified, and there is any uncertainty as to the correct action, this should be marked with an appropriate flag to indicate that it is suspect. The data flagged as suspect are reviewed at the time of secondary validation. 4.3 Secondary validation 4.3.1 Overview Secondary validation of rainfall data is primarily carried out at State DPCs, to take advantage of the information available from a larger area. Secondary validation is carried out using e-SWIS, the validation module of which replicates the HYMOS software from HPI, and is referred to as eHYMOS. Data may also be exported to Excel for secondary validation. For the Hydrology Project, secondary validation (see Table 3.1) done at State level should be completed by the end of the following month (e.g. for June data by 31st July). Some secondary validation (including comparison with IMD data) will not be possible until the end of the hydrological year when the entire year’s data can be reviewed in a long-term context, so data should be regarded as provisional approved data until then (e.g. for June data by the end of the hydrological year plus 3 months), after which data should be formally approved and made available for dissemination to external users. Data entering secondary validation have already received primary validation on the basis of knowledge of the station and instrumentation and field documents. Data may have been flagged as missing, accumulated, shifted or suspect for some other reason e.g. a mismatch in the number of rain days. Secondary validation focuses on comparisons with neighbouring stations to identify suspect values. However, data processing staff should continue to be aware of field practice and instrumentation and the associated errors which can arise in data. Some of the secondary validation checks are oriented towards the specific types of errors just mentioned, whilst others are general in nature and lead to identification of spatial inconsistencies in the data. Rainfall poses particular problems for spatial comparisons because of the limited or uneven correlation between stations. When rainfall is convectional in type, it may rain heavily at one location, whilst another only a few miles away may remain dry. Over a month or a monsoon season, such spatial unevenness tends to be smoothed out and aggregated totals are much more closely correlated. Spatial correlation in rainfall depends on type of precipitation, physiographic characteristics of the region, duration (decreases as duration decreases) and distance (decreases as distance increases). Examples of many of the techniques described in this section are given in Hydro-Meteorology Training Module 09 “How to carry out secondary validation of rainfall data” and Training Module 10 “How to correct and complete rainfall data”.



Page 25

4.3.2 Comparison with data limits for totals at longer durations Daily rainfall data are aggregated to monthly, seasonal and yearly time intervals for checking if the resulting data series is consistent with the prescribed data limits for such time intervals. This is a useful technique for identifying small systematic or persistent observation errors in rainfall that are not apparent at a daily time interval but which tend to accumulate with time and, therefore, become more visible. As well as upper warning levels or maximum limits, for monsoon months and yearly values, use of lower warning levels can be made to see if certain values are unexpectedly low and thus warrant a closer look. This is a valuable secondary check for SRGs at which there is no sub-daily raingauge for comparison. Suspect aggregated values are flagged and commented appropriately for further validation on the basis of data from nearby stations. 4.3.3 Comparison of rainfall from multiple stations A combination of graphical and tabular displays of hourly, daily, monthly, seasonal and annual rainfall data from multiple stations in a region provides an efficient way of identifying anomalies. Where only two stations are involved in the comparison, the identification of an anomaly does not necessarily indicate which station is at fault. Specific checks to make include: • Do the daily blocks of rain days generally coincide in start day and finish day? • Are there exceptions that are misplaced, starting one day early or late? • Is there a consistent pattern of misfit for a station through the month? • Are there days with no rainfall at a station when (heavy) rainfall has occurred at all

neighbouring stations? For multiple rainfall time series plots, select a set of stations within a small area with an expectation of spatial correlation including in the set, if possible, one or more stations which historically have been more reliable. Plot the rainfall series as histograms, preferably in different colours for each station. Side by side plotting of histograms permits comparison on the magnitudes of rainfall at different stations, whilst one above the other plotting makes time shifts easier to detect. This is a valuable secondary check for SRGs at which there is no sub-daily raingauge for comparison. Rainfall occurs in dry and wet spells and SRG-only observers may fail to record the zeros during the dry spells and, hence, lose track of the date when the next rain arrives. When ancillary climate data are available, this may be used to compare with rainfall data e.g. a day with unbroken sunshine in which rain has been reported suggests that rainfall has been reported for the wrong day. However, most comparisons are not so clear cut and the user should be aware that there are a number of possibilities: • Rainfall data only on the wrong day - anomalies between rainfall and climate and between

rainfall and neighbouring rainfall • Rainfall and climate data both reported on the wrong day - hence no anomaly between them

but discrepancy with neighbouring stations • Rainfall and climate both reported on the correct day - the anomaly was in the occurrence of

rainfall e.g. no rainfall at one site but at neighbouring sites. In this case, climate variables are likely to have been shared between neighbouring stations even if rainfall did not occur

Unexplained anomalies should initially be followed up by checking the field documents to check for unnoticed mistakes during data entry or primary validation, in which case the data can be corrected accordingly. If necessary, the anomaly should be communicated to the supervising field officer and observer to confirm data and/or rectify problems. Data still regarded as suspect after follow-up checking are flagged and commented appropriately for further validation on the basis of data from nearby stations.



Page 26

Figure 4.1 Definition of test and neighbouring stations

4.3.4 Spatial homogeneity testing of rainfall Spatial homogeneity testing of rainfall data is a technique where the observed rainfall at a station station is compared with an estimate of the rainfall at that station, based on the weighted average of rainfall from nearby stations. This approach depends on the degree of spatial consistency of the rainfall, which is primarily based on the actual spatial correlation. This check for spatial consistency can be carried out for various durations of rainfall accumulations. Wherever the difference between the observed and the estimated values exceeds the expected limiting value, such observed values are considered suspect values and flagged for further investigation to ascertain the possible causes of the differences. It is only possible to include a brief description of the technique below, but more detailed information is provided in SW8-OM(II) Chapter 2.6 and Chapter 3.6. Firstly, the nearby stations to be used should be chosen. The stations selected as neighbours should be physically representative of the area in which the “test station” (the station under scrutiny) is situated. The following criteria are used to select the neighbouring stations (Figure 4.1): • The distance between the test and the neighbouring station should be less than a specified

maximum correlation distance, say Rmax km • A maximum of 8 neighbouring stations can be considered for interpolation • To reduce the spatial bias in selection, it is appropriate to consider a maximum of only two

stations within each quadrant. Secondly, having selected the neighbouring stations, the estimation of the spatially interpolated rainfall is made at the test station, by computing the weighted average of the rainfall observed at neighbouring stations. The estimate of the interpolated value at the test station based on the observations at M neighbouring stations is given as:

Where: Pest,j = estimated rainfall at the test station at time j

Pi,j = observed rainfall at the neighbour station i at time j



Page 27

Di = distance between the test and the neighbouring station i Mbase = number of neighbouring stations taken into account b = power of distance D used for weighting rainfall values at individual station e.g. 2

Thirdly, this estimated rainfall is compared with the observed rainfall at the test station and the difference is considered insignificant if: • The difference between the observed and estimated rainfall at the test station is less than or

equal to the admissible absolute difference Xabs • The difference between the observed and estimated rainfall at the test station is less than or

equal to a multiplier Xrel of the standard deviation SPest,j of the neighbouring values given by:

Where differences between the observed and estimated are unacceptably high, the observed value should be flagged “+” or “-”, depending on whether the observed rainfall is greater or less than the estimated one. Typical rainfall measurement errors show up with specific patterns of “+” and “-”. The limits Xabs and Xrel are chosen by the user and have to be based on the spatial variability of rainfall, and may be altered seasonally. They are normally determined on the basis of experience with the historical data with the objective of flagging a few values (say 2-3%) as suspect values. It is customary to select a reasonably high value of Xabs and a value of Xrel which avoid having to deal with a large number of difference values in the lower range where differences are more likely to occur and have less effect on the overall rainfall total. 4.3.5 Double mass analysis Double mass analysis is a technique to detect a systematic shift, like abrupt or gradual changes, in a rainfall time series, persisting for a considerable period of time. Such inconsistencies can occur for various reasons: • The raingauge might have been installed at different sites in the past • The exposure conditions of the gauge may have undergone a significant change due to the

growth of trees or construction of buildings in its proximity • There might have been a change in the instrument, say from 125 mm to 200 mm raingauge • The raingauge may have been faulty for a considerable period etc A note may be available in the station files of the known changes of site and instruments and can be used to corroborate the detection of inconsistencies. The application of double mass analysis to rainfall data is not be possible until a significant amount of historical data is available. The accumulated rainfall at the test station (the station under scrutiny) is compared with another accumulated rainfall series that is expected to be homogeneous. The homogeneous time series can be from an individual reliable station or be an average of reliable time series from neighbouring stations, referred to as the base station(s). Accumulation of rainfall can be made from daily data to monthly or yearly duration. It is only possible to include a brief description of the technique below, but more detailed information is provided in SW8-OM(II) Chapter 2.12 and Chapter 3.5. Firstly, the double mass plot between the accumulated rainfall values in absolute or percent form at test and base stations is drawn and observed for any visible change in its slope. The tabular output giving the ratio between the accumulated rainfall values at test and base stations in absolute and percent form is also obtained. The analysis can be carried for only a part of the years



Page 28

or months, if there are some missing rainfall values within the time series (which may themselves indicate some change at the station). Secondly, any visible change in the slope of the double mass plot should be noted. If the data of the test station is homogeneous and consistent with the data of the base station(s), the double mass curve will show a straight line. An abrupt change in the test rainfall series will create a break in the double mass curve, whereas a trend will create a curve. A change in slope is not usually considered significant unless it persists for at least 5 years, and it does not imply that either period is incorrect, simply that it is inconsistent. Furthermore, double mass analysis is based on the assumption that only a part of the rainfall time series under consideration is subject to systematic shift. Where the whole rainfall time series has such a shift, the double mass analysis will fail to detect any inconsistency. Any significant inconsistency that is detected should be investigated further to explore possible causes. If the inconsistency is caused by changed exposure conditions or shift in the station location or systematic instrumental error, the rainfall series should be considered suspect. Thirdly, the suspect part of the rainfall series can be made homogeneous by suitably transforming it before/after the period of shift as required. The earlier part of the record may be adjusted so that that the entire record is consistent with the present and continuing record, or the latter part of the record may be adjusted when the source of the inconsistency is known and has been removed or where the record has been discontinued. Transformation is carried out by multiplying it by a correction factor which is the ratio of the slope of the adjusted mass curve to the slope of the unadjusted mass curve. In the double mass plot shown in Figure 4.2, there is a distinct break at time T1. If the start and end times of the period under consideration are T0 and T2, respectively, then the slopes of the curve before α1 and after α2 the break point can be expressed as:

Figure 4.2 Definition sketch for double mass analysis



Page 29

In the case that the earlier part between T0 and T1 needs to be corrected for, the correction factor and the corrected observations at the test station can be expressed as:

After making such a correction, the double mass curve should be plotted again to check that there is no significant change in the slope of the curve. 4.4 Correction and completion 4.4.1 Overview Completion – the processing of filling in missing values and correcting erroneous values – is done as a continuous process with primary and secondary validation. Although the HIS Manual SW separates correction and completion in SW8-OM(II) Chapter 3 from secondary validation in SW8-OM(II) Chapter 2, and from primary validation in SW8-OM(I) Chapter 5, there is substantial overlap between the techniques used. In this Handbook, some correction and completion techniques have been included in the appropriate parts of Sections 4.2 and 4.3, and others are described below. Examples of many of the techniques described, which should be carried out by experienced staff with appropriate training, are given in Hydro-Meteorology Training Module 10 “How to correct and complete rainfall data”. The majority of secondary validation, and therefore the majority of correction and completion, is carried out by State DPCs to take advantage of the information available from a larger area. For the Hydrology Project, correction and completion (see Table 3.1) should be completed by the end of the following month (e.g. for June data by 31st July). Some secondary validation, correction and completion will not be possible until the end of the hydrological year when the entire year’s data can be reviewed in a long-term context, so data should be regarded as provisional approved data until then (e.g. for June data by the end of the hydrological year plus 3 months), after which data should be formally approved and made available for dissemination to external users. Procedures for correction and completion depend on the type of error and the availability of suitable source records with which to estimate. It should be recognised that values estimated from other gauges are inherently less reliable than values properly measured. There will be circumstances where no suitable neighbouring observations or stations are available, such that missing values should be left as -999 and incorrect values should be set to -999, and suspect original values should be given the benefit of the doubt and retained in the record with an appropriate flag. 4.4.2 Correcting missing and erroneous data Missing values (-999 or incorrect zeros) may be the result of the observer failing to make an observation, failing to enter the observation in the record sheet, or entering the observation incorrectly. For rainy periods, missed values will be anomalous in the multiple station tabulation and plot and will be indicated by a series of “-” departures in the spatial homogeneity test. Where such missed entries are confidently identified, the missed values should be replaced by the estimates derived from neighbouring stations by spatial interpolation. This is also the approach for values confidently identified as erroneous. Where there is some doubt as to the interpretation, the



Page 30

value should be left unchanged but flagged as suspect. There are three analytical procedures for estimating rainfall by spatial interpolation: • Arithmetic average – applied if the average annual rainfall of the station under consideration

is within 10% of the average annual rainfall at the neighbouring stations. The missing or erroneous rainfall at the station under consideration is estimated as the simple average of neighbouring stations. Thus, if the estimate for the missing or erroneous rainfall at the station under consideration is Ptest and the rainfall at M adjoining stations is Pbase,i (i = 1 to M), then:

Usually, averaging of three or more neighbouring stations is considered to give a satisfactory estimate.

• Normal ratio – applied if the average annual rainfall of the station under consideration differs

from the average annual rainfall at the neighbouring stations by more than 10%. The erroneous or missing rainfall at the station under consideration is estimated as the weighted average of the data at the neighbouring stations. The rainfall at each of the neighbouring stations is weighted by the ratio of the average annual rainfall at the station under consideration and average annual rainfall of the neighbouring station. The rainfall for the missing or erroneous value at the station under consideration is estimated as:

Where: Ntest = annual average rainfall at the station under consideration

Nbase,i = annual average rainfall at the adjoining stations (for i = 1 to M)

A minimum of three neighbouring stations should generally be used for obtaining good estimates using the normal ratio method.

• Distance power – this is the approach used for the spatial homogeneity test in Section 4.3.4 of

this Handbook, and weights neighbouring stations on the basis of their distance from the station under consideration, on the assumption that closer stations are better correlated than those further away and that, beyond a certain distance, they are insufficiently correlated to be of use. Spatial interpolation is made by weighing the adjoining station rainfall as inversely proportional to some power (e.g. 2) of the distances from the station under consideration.

However, due to prevailing wind conditions or orographic effects, spatial heterogeneity may be present in which case, normalised values rather than actual values should be used in the interpolation. The observed rainfall values at the neighbouring stations are multiplied by the ratio of the average annual rainfall at the test station and the average annual rainfall at the neighbouring stations:

Where: Pcorr,i,j = for corrected rainfall value at the neighbouring station i at time j

Ntest = annual average rainfall at the station under consideration Nbase,i = annual average rainfall at the neighbouring stations (for i = 1 to Mbase)

4.4.3 Correcting accumulated data Should an observer miss one or more readings, he/she may make one of three choices for the missed period of record:



Page 31

• Enter the value of the accumulated rainfall on the day on which he/she returned from absence and indicate that the intervening values were accumulated (the correct approach)

• Enter the value of the accumulated rainfall on the day on which he/she returned and enter a zero (or no entry) in the intervening period

• Attempt to guess the distribution of the accumulated rainfall over the accumulated period and enter a positive value for each of the days

The third option is probably the more common as the observer may fear penalisation for missing a period of record, even for a legitimate reason. The second also occurs. Observers should be encouraged to follow the first option, as a more satisfactory interpolation can be made from adjacent stations than by the observer’s guess. By preparing a list of holidays and weekends, it is possibly to proactively screen for accumulated values. For unindicated accumulations with zeros in the missed values, during rainy periods, missed values will be anomalous in the multiple station tabulation and plot and will be indicated by a series of “-” departures in the spatial homogeneity test followed by a “+” on the day of accumulation. The user should inspect the rainfall time series for patterns of this type and flag such occurrences as suspect. Anomalies should initially be followed up by checking the field documents to check for unnoticed mistakes during data entry or primary validation, in which case the data can be corrected accordingly. If field data were entered and processed correctly, the user should search backwards from the date of the accumulated total to the first date on which a measurable rainfall has been entered to identify the period of accumulation. Any positively identified or suspect accumulation not correctly indicated should be communicated to the supervising field officer and observer to confirm data and/or rectify problems. Accumulations that are clearly marked by the observer, or identified during secondary validation, can be distributed over the period of absence, by comparison with the distribution over the same period at neighbouring stations. Firstly, given that the accumulated value of the rainfall and the number of days of accumulation are known, estimates of daily rainfall, for each day of the period of accumulation, at the station under consideration are made using spatial interpolation from the adjoining stations (in the first instance without reference to the accumulation total) using:

Where: Pest,j = estimated rainfall at the test station for jth day

Pij = observed rainfall at ith neighbour station on jth day Di = distance between the test and ith neighbouring station Nbase = number of neighbouring stations considered for spatial interpolation. b = power of distance used for weighting individual rainfall value e.g. 2

Secondly, the accumulated rainfall is then apportioned in the ratio of the estimated values on the respective days as:

Where: Ptot = accumulated rainfall as recorded

Nacc = number of days of accumulation Pappor,j = apportioned rainfall for jth day during the period of accumulation

Where it is not possible to adequately reason in favour or against such an accumulation, the value should be left unchanged but flagged as suspect.



Page 32

4.4.4 Correcting shifted data Shifts may result from rainfall values being entered against the wrong days. Because, rainfall data are interspersed with many entries having zero values, one or more zero entries may get omitted or repeated by mistake. For daily data, such mistakes are more likely when there are a few non-zero values in the middle of the month and most zero entries at the beginning and end of the month. Shifts tend to get automatically corrected with the start of a new month, because a new column or page is started in the field documents. Shift errors in rainfall series can usually be spotted in the tabulated or plotted multiple series, especially if they are repeated over several wet/dry spells. It is assumed that no more than one of the listed series will be shifted in the same direction in the same set. Correction for removing the shift in the data is done by either inserting the missing data and/or deleting the extra data points causing the shift (usually zero entries). Before inserting or deleting data points, it is important to establish the number of days of shift (e.g. a 2 day forward shift) and the extent of data affected by the shift (e.g. until the end of the month). Spatial homogeneity testing will generate a “+” at the beginning of a wet spell and a “-”- at the end (and possibly others in between) if the data are shifted forwards, and the reverse if the data are shifted backwards. A shift to coincide with the timing of adjacent stations and rerun of the spatial homogeneity test will generally result in the disappearance of the “+” and “-” flags, if the interpretation of the shift was correct. The re-shifted series is then adopted as the validated series for the station/period in question. 4.4.5 Correcting suspect data The practice of some observers may be to visit the gauge only when they know that rain has occurred. This results in zeros on a number of days on which a small amount of rain has occurred, though totals should be generally correct at the end of the month. Spatial homogeneity testing may not pick up such differences. However, the number of rain days may be anomalously low in comparison to neighbouring stations. The tolerance in the number of rain days between the stations should be based on the variability experienced in the region, and can easily be established using historical daily data belonging to a homogenous region to establish the expected maximum variation in the number of rainy days for each month of the year and for the year as a whole. For the group of rainfall stations being validated, the number of rain days at each station within the month(s) or year is obtained, and compared with every other station in the group. Graphical or tabular comparison of the difference in the number of rain days for the stations for the monthly or annual period should reveal instances when the expected variation is exceeded by the actual difference in the number of rain days. If the difference is more than expected, the field documents should first be checked for unnoticed mistakes during data entry or primary validation, in which case the data can be corrected accordingly. Comparison with other related data like temperature and humidity at the station, if available, can be made. If necessary, the anomaly should be communicated to the supervising field officer and observer to confirm data and/or rectify problems. Together with analytical comparison, feedback from the observer or supervisor will be of a great value in checking this validation. If related data corroborate the occurrence of a low number of rain days, the data can be accepted. Where there is strong evidence to support the view that the number of rainy days derived from the record is incorrect, then the total may be amended by reference to neighbouring stations. Such action implies that there are unreported errors remaining in the time series, which it has not been possible to identify and correct. A comment to this effect should be included with the data and provided to data users.



Page 33

4.5 Compilation 4.5.1 Overview Rainfall compilation is the process by which observed rainfall is transformed: • From one time interval to another • From one unit of measurement to another • From point to areal values • From non-equidistant to equidistant series Compilation is required for validation, reporting and analysis. Hence, some compilation is done prior to and during validation as required, but final compilation is carried out after correction and completion. The majority of correction and completion, and therefore the majority of final compilation, is carried out by State DPCs. Examples of many of the techniques described in this section are given in Hydro-Meteorology Training Module 11 “How to compile rainfall data”. 4.5.2 Aggregation of rainfall to longer durations Rainfall from different sources is observed at different time intervals, but these are generally daily or sub-daily. Aggregation to longer time intervals is required for validation and analysis e.g. hourly rainfall to daily rainfall for comparison of SRG and ARG/TBR data during primary validation. For secondary validation, small persistent errors may not be detected at the small time interval of observation but may more readily be detected at longer time intervals. • Hourly to other intervals - from rainfall data at hourly or lesser time intervals, it may be

necessary to obtain rainfall data for every 2 hours, 3 hours, 6 hours, 12 hours, etc. for any specific requirement. Such compilations are carried out by adding up the corresponding rainfall data at available smaller time intervals

• Daily to weekly – aggregation of daily to a weekly time interval is usually done by considering the first 51 weeks as equal length (i.e. 7 days) and the last (52nd) week of either 8 or 9 days according to whether the year is a non-leap year or a leap year, respectively. Rainfall data for such weekly periods are obtained by summing the consecutive sets of 7-day rainfalls, with the last week by summing the last 8 or 9 days of daily rainfalls. For applications where it is necessary to align with calendar weeks (Mon-Sun), the first week in any year will start from the first Monday in that year and there will be 51 or 52 full weeks in the year and one or more days left at the beginning and/or end of the year

• Daily to 10-day – aggregation of daily to a 10-day time interval is usually done by considering each month as three 10-day periods. Hence, every month will have the first two 10-day periods as 10 days each and the last 10-day period as either 8, 9, 10 or 11 days according to the month and the year. Rainfall data for such 10-day periods are obtained by summing the corresponding daily rainfall data

• Daily to monthly – monthly data are obtained from daily data by summing the daily rainfall data for the calendar months. Thus, the number of daily data to be summed up will be 28, 29, 30 or 31 according to the month and year under consideration

• Daily/monthly to annual – yearly data are obtained by summing the corresponding daily data or monthly data, as available

4.5.3 Estimation of areal rainfall Many hydrological applications (e.g. basin rainfall-runoff modelling) require the average depth of rainfall occurring over an area which can then be compared directly with runoff from that area. However, raingauges measure rainfall at individual points and, since rainfall is spatially variable and the spatial distribution varies between events, point rainfall does not provide a precise



Page 34

estimate or representation of areal rainfall. All methods for estimation of areal average rainfall compute the weighted average of the point rainfall values, and the difference between various methods is only in the assigning of weights to these individual point rainfall values, the weights being primarily based on the proportional area represented by a point gauge. The areal rainfall will always be an estimate and not the true rainfall depth irrespective of the method. SW8-OM(II) Chapter 4 includes five analytical procedures for estimating areal rainfall: • Arithmetic average – in the simplest method, the areal average rainfall depth is estimated by

averaging of all selected point rainfall values for the area under consideration:

Where: Pat= estimated average areal rainfall depth at time t Pit = individual point rainfall values considered for an area, at station i ( for i = 1,N) and time t N = total number of point rainfall stations considered

In this case, all point rainfall stations are allocated weights of equal magnitude, equal to the reciprocal of the total number of stations considered. Generally, stations located within the area under consideration are taken into account, but it is good practice also to include such stations which are outside the area but close to the areal boundary and represent some part of the areal rainfall within the boundary. This method gives satisfactory estimates and is recommended where the area under consideration is flat, the spatial distribution of rainfall is fairly uniform, and the variation of individual gauge records from the mean is not large.

• Weighted average using user-defined weights – a variation of the arithmetic average

method where different stations have different weights to account for orographic effects and especially where raingauges are predominantly located in the lower rainfall valleys. User-defined weights can be assigned to the stations under consideration. The estimation of areal average rainfall depth is:

Where: ci = weight assigned to individual raingauge station i (i = 1,N) To account for under-representation by gauges located in valleys the weights do not necessarily need to add up to 1.

• Theissen polygon method – this method accounts for the variability in spatial distribution of

rain gauges and the consequent variable area which each gauge represents. The areas representing each gauge are defined by drawing lines between adjacent stations on a map. The perpendicular bisectors of these lines form a pattern of polygons (the Thiessen polygons) with one station in each polygon (Figure 4.3). Stations outside the basin boundary should be included in the analysis as they may have polygons which extend into the basin area. The area of a polygon for an individual station as a proportion of the total basin area represents the Thiessen weight for that station. Areal rainfall is estimated by first multiplying individual station totals by their Thiessen weights and then summing the weighted totals:

Where: Ai = the area of Thiessen polygon for station i

A = total area under consideration



Page 35

Figure 4.3 Example of basin area divided into Theissen polygons (after: NIH. Roorkee)

Figure 4.4 Example of drawing of isohyets using linear interpolation

The Thiessen method is objective and readily computerised but is not ideal for mountainous areas where orographic effects are significant or where raingauges are predominantly located at lower elevations of the basin. Altitude-weighted polygons (including altitude as well as areal effects) have been devised but are not widely used.

• Isohyetal method – this method overcomes the Theissen polygons’ inability to deal with

orographic effects, by interpolating between point rainfalls, taking into account orographic effects, to draw lines of lines of equal rainfall (isohyets). The locations of the raingauges within and outside near the basin boundary are plotted on a map and connected with their neighbouring stations by straight lines. Selected isohyetal rainfall depths are marked on the connecting lines, and smooth curves drawn between connecting lines connecting equal rainfall depths, taking into account topography and/or storm orientation (Figure 4.4). The inter-isohyet area between each two adjacent isohyets and the basin boundary is assumed to have received the average rainfall from the two adjacent isohyets. If the isohyets are indicated by



Page 36

P1, P2, …, Pn with inter-isohyet areas a1, a2, …, an-1 the mean precipitation over the basin is computed from:

If the maximum and/or minimum point rainfall value(s) are within the basin boundary then P1 and/or Pn is replaced by the highest and/or lowest point rainfall value(s). A variant of the isohyetal method is the isopercentile method. For more information, see SW8-OM(II) Chapter 4.3.5.

• Kriging method – an interpolation method which provides rainfall estimates at points (point-

kriging) or blocks (block-kriging) based on a weighted average of observations made at neighbouring stations. In the application of the point kriging method for areal rainfall estimation and drawing of isohyets a dense grid is put over basin. By estimating the rainfall for the gridpoints, the areal rainfall is determined as the average rainfall of all grid points within the basin. In addition, in view of the dense grid, isohyets based on the rainfall values at the grid points can readily be drawn. At each gridpoint the rainfall is estimated from:

Where: Pe0 = rainfall estimate at some gridpoint “0”

w0,k = weight of station k in the estimate of the rainfall at point “0” Pk = rainfall observed at station k N = number of stations considered in the estimation of Pe0

The weights are different for each grid point and observation station. The weight given to a particular observation station k in estimating the rainfall at gridpoint “0” depends on the gridpoint-station distance and the spatial correlation structure of the rainfall field. For more information see SW8-OM(II) Chapter 4.3.6.

4.5.4 Transformation of non-equidistant to equidistant time series Digital data from TBRs can take two forms: time mode where the number of tips in a pre-set time interval (e.g. 1 hour, 15 minutes, etc) are recorded, or in event mode where the times of every tip are recorded, thereby producing a more flexible record for subsequent analysis. Time mode data are imported to an appropriate equidistant time series, whilst event mode data are imported to a non-equidistant time series. The eHYMOS module of e-SWIS provides a means of transforming such non-equidistant series to equidistant series by accumulating each unit tip measurement to the corresponding time interval. Time intervals for which no tip has been recorded are filled with zero values. 4.5.5 Compilation of mean, maximum and minimum time series The annual, seasonal or monthly maximum series of rainfall is frequently required for flood analysis, whilst minimum series may be required for drought analysis. The eHYMOS module of e-SWIS provides options for the extraction of minimum, maximum, mean, median values for any defined period within the year (i.e. season) or for the complete year e.g. minimum, maximum, mean, median rainfalls of 10-day duration during the Jul-Oct monsoon months, between a specified start and end date.



Page 37

4.6 Analysis 4.6.1 Overview Some analysis of rainfall data is required for validation and further analysis may be needed for data presentation and reporting. The majority of analysis (see Table 3.1) is carried out by State DPCs. Examples of many of the techniques described in this section are given in Hydro-Meteorology Training Module 12 “How to analyse rainfall data”. It is only possible to include an overview of the techniques below, but more detailed information is provided in SW8-OM(III) Chapter 4. For statistical analysis, rainfall data from a single series should ideally be homogenous i.e. the characteristics of different portions of the data series do not vary significantly, and rainfall data for multiple series at neighbouring stations should ideally possess spatial homogeneity. Tests of homogeneity are required for validation purposes and are described in the appropriate sections of this Handbook: • Spatial homogeneity testing in Section 4.3.4 • Double mass analysis for consistency testing in Section 4.3.5 • Multiple station validation in Section 6.3.2 • Single site homogeneity testing in Section 6.3.3 4.6.2 Computing basic statistics Basic statistics are widely required for validation and reporting including: • Arithmetic mean

• Median - the median value of a ranked series Xi • Mode - the value of X which occurs with greatest frequency or the middle value of the class

with greatest frequency • Standard deviation - the root mean squared deviation Sx:

• Skewness – the extent to which the data deviate from a symmetrical distribution • Kurtosis – the peakedness of a distribution In addition, empirical frequency distributions can be presented as a graphical representation of the number of data per class and as a cumulative frequency distribution. From these, selected values of exceedance probability or non-exceedance probability can be extracted e.g. the daily rainfall which has been exceeded 1%, 5% or 10% of the time. 4.6.3 Annual maximum and annual exceedance rainfall series The following are widely used for reporting or for subsequent use in frequency analysis of extreme rainfalls: • Maximum series. The maximum rainfall value of an annual series, or of a month or season.

Most commonly for rainfall, daily maxima per year are used, but hourly maxima, N-hourly and N-day maxima may also be selected



Page 38

• Exceedance series (also known as peaks-over-threshold series and partial duration series). All daily rainfall values over a specified threshold may also be selected

• Minimum series. As the minimum daily value with respect to rainfall is frequently zero, this is useful for aggregated data only

4.6.4 Fitting frequency distributions A common use of rainfall data is in the assessment of probabilities or return periods of a given rainfall at a given location. Such information can be used in assessing flood discharges of given return periods through modelling, for use in the design of flood alleviation schemes, bridges and culverts, and in flood forecasting. Frequency analysis usually involves the fitting of a theoretical frequency distribution using a selected fitting method, although empirical graphical methods can also be applied. The fitting of a particular distribution implies that the rainfall sample of annual maxima or annual exceedance were drawn from a population of that distribution. For the purposes of application in design, it is assumed that future probabilities will be the same as past probabilities. However there is nothing inherent in the series to indicate whether one distribution is more likely to be appropriate than another, and a wide variety of distributions and fitting procedures has been recommended for application in different countries and by different agencies. See the e-SWIS/eHYMOS manual for details about which frequency distributions and fitting procedures are available. Different distributions can give widely different estimates, especially when extrapolated or when an outlier (an exceptional value, well in excess of the second largest value) occurs in the dataset. Although the methods are themselves objective, a degree of subjectivity is introduced in the selection of which distribution to apply. Output typically includes: • Estimation of parameters of the distribution • A table of rainfalls of specified probabilities or return periods with confidence limits • Results of goodness of fit tests • A graphical plot of the data fitted to the distribution When fitting frequency distributions, graphical, as well as numerical output, should always be inspected. 4.6.5 Other statistical analysis techniques for rainfall data Other techniques for analysing rainfall involve advanced statistical analysis and are likely to be of interest only to experienced hydro-meteorologists at State level and at IMD. Three techniques are: • Frequency and duration curves Frequency curves provide a way to show the variation of

rainfall through the year, where each frequency curve indicates the rainfall depth for a specific probability of non-exceedance. Duration curves are ranked representations of frequency curves. The average duration curve gives the average number of occasions a given rainfall depth was not exceeded in the years considered. See SW8-OM(III) Chapter 4.6.

• Intensity-frequency-duration analysis (IDF) (also known as depth-duration-frequency

DDF analysis) If rainfall data from a recording raingauge are available for long periods such as 25 years or more, the frequency of rainfalls of given depths and durations can be determined. Intensity-frequency-duration relationships may be established for a full year, a season or a month. Analysis is carried out for a number of stations in a region enables development of isopluvial maps, where each map depicts maximum rainfall depths for a specified duration and frequency, including at locations where there is no raingauge. See SW8-OM(III) Chapter 4.7.



Page 39

• Depth-area-duration analysis (DAD) In most of the design applications, the maximum depth of rainfall that is likely to occur over a given area for a given duration is required. A storm of given duration over a certain area rarely produces uniform rainfall depth over the entire area. The storm usually has a centre, where the rainfall Po is maximum which is always larger than the average depth of rainfall P for the area as a whole. Generally, the difference between these two values (Po – P) increases with increase in area and decreases with increase in the duration. If rainfall data from recording raingauges in a region are available for long periods such as 25 years or more, quantitative relationships between Po and P can be developed. See SW8-OM(III) Chapter 4.8.



Page 40

5. Snow Data Processing and Analysis 5.1 Snow in the Hydrology Project 5.1.1 Overview Snow is an important form of precipitation that occurs only in the Himalayan regions of India, where most snowfall occurs between November and February. Hence, snow data were not considered in HPI which did not cover any Himalayan States or agencies, and there is no guidance to snow data in the HIS Manual. Whilst this Handbook provides an overview, users are referred to the WMO Report 168 “Guide to Hydrological Practices” and WMO Report 749 “Snow Cover Measurements and Areal Assessment of Precipitation and Soil Moisture” for more detailed information on snow monitoring sites and on snow sampling equipment and procedures, and also to the e-SWIS manual for instruction on data entry, validation and subsequent actions. Accurate quantification of the coverage, depth and density of snow is important because, unlike other hydro-meteorological variables, snow can be “stored” in a snowpack for many months before melting and re-entering the hydrological cycle. Sudden snowmelt, triggered by a rapid rise in temperature or heavy rain on snow, can cause flooding. Hence, snow data provide useful information for river forecasting and flood studies. In the Himalayas, snowmelt starts between March and May. Singh & Singh (2001, after Linacre, 1992) present a maximum observed snowmelt for the Himalayas of 25 mm/day, but this should be regarded as a general rate given the temporal and spatial variability in atmospheric conditions and snowpack conditions across the Himalayas. 5.1.2 Snow data At snow stations, observations are made of: • Snowfall since the last observation - snowfall is the depth of fresh snow deposited over a

specified period i.e. the snowfall since the last observation. Snowfall on flat ground (sloping surfaces should be avoided where possible), is measured using a graduated ruler or scale, taking an average of several vertical measurements in places where the effects of wind are minimised and there is an absence of drifting snow. To avoid measuring old snow, a suitable patch of ground should be swept clear beforehand or the top of the snow surface should be covered with a piece of suitable material (such as wood, with a slightly rough surface, painted white). Snowfall may be measured daily, twice-daily, four-time a day, or hourly.

• Total depth of snow on the ground (i.e. the depth of the snowpack) - the total depth of snow

on the ground is combined depth of new and old snow on the ground at observation time. To measure the total depth of the snow pack, calibrated stakes, painted white and graduated in metres and centimetres, are fixed at representative sites that can be inspected easily from a distance (e.g. using binoculars). Snowstake readings are usually made daily.

• Snow-water equivalent (SWE i.e. the depth of liquid precipitation contained in that snowfall

and/or the snowpack). Snowfall SWE is usually measured using a storage snow gauge, or a tipping bucket snow gauge, or a combined gauge that records both rainfall and snowfall (also known as a double bucket collector). For a non-recording gauge, the snow in the container is melted and poured into a standard precipitation measuring cylinder to read the depth of melted snow. A tipping bucket snow gauge collects snow, ideally in a heated container, until a critical weight (equivalent to a specific depth of melted snow) is reached, at which time it tips and empties, sending a signal to an autographic chart recorder or digital data logger. Precipitation in the form of snow is much more subject to adverse wind conditions than snowfall and, in



Page 41

exceptionally windy locations, the catch in a gauge may be a significant under-estimate, or less common over-estimate.

Should the water equivalent of the snowpack be required, this and the density of the snowpack, may be estimated from physical measurements of cylindrical samples of snow (also known as snow pillars). Sampling points should be located by measuring from a reference mark, as indicated on the map of the snow station/course. Snow sampling equipment typically consists of a metal or plastic tube with a snow cutter fixed at its lower end and with a length scale stamped on its exterior surface throughout its length, a spring or level balance for determining the weight of the snow pillars and their snow-water equivalent, a wire cradle for supporting the tube while it is being weighed, and tools for operating the snow sampler. Results should be recorded on site in the field notebook, together with other relevant observations e.g. weather conditions.

The water equivalent of the snowpack may also be recorded automatically using a snow pillow. A snow pillow is a flat circular bag, usually made of a rubberised material and containing an environmentally-safe antifreeze. Snow pillows of various dimensions and material exist. The pillow is installed on the surface of the ground, flush with the ground, or buried under a thin layer of soil or sand. Hydrostatic pressure is an indicator of the weight, and therefore the water equivalent, of the snow on the pillow. Pressure is measured using a pressure transducer inside the pillow, and recorded on a digital data logger. The snow-water equivalent determined by snow pillow typically differs by 5-10% from measurements by the weighing method.

5.2 Data entry 5.2.1 Overview Entry of data to computer is primarily done at a Sub-Divisional office level where staff are in close contact to field staff who have made the observations and/or collected the physical, chart or digital data. Data entry is carried out using e-SWIS, the data entry module of which is referred to as eSWDES. The eSWDES snow module provides a variety of options for entering snow data, including snowfall, snow depth (referred to as snow stake) and SWE. Where climate data (sunshine, temperature, humidity, wind speed and direction, and atmospheric pressure) are also collected at a snow station, they are entered against the snow station using the appropriate form. However, see the relevant parts of Section 6 for guidance on data entry, validation, correction and completion and analysis of climate data. Prior to entry to computer, two manual activities are essential: registration of receipt of the data, and manual inspection of charts, forms and notebooks from the field, for complete information and obvious errors. Data entry (see Table 3.1) and primary validation of field data from observational stations is required to be completed at Sub-Divisional/Divisional office level by the 10th working day of the following month (e.g. for December data by 10th working day in January), ready for subsequent actions by State offices. 5.2.2 Manual inspection of field records Prior to data entry to computer an initial inspection of field records is required. This is done in conjunction with notes received from the observation station on equipment problems and faults, missing records or exceptional snowfall. Snowfall sheets and charts are inspected for the following: • Is the station name and code and month and year recorded? • Does the number of record days correspond with the number of days in the month?



Page 42

• Are there some missing values or periods for which snowfall has been accumulated or snow stake readings missed during absence of the observer?

• Have monthly totals of snowfall and snow days been entered? • Have any autographic hourly data been extracted? • Is the record written clearly and with no ambiguity in digits or decimal points? Any queries arising from such inspection should be communicated to the observer to confirm ambiguous data before data entry. Any unresolved problems should be noted and the information sent forward with the digital data to Divisional/State offices to assist in initial secondary validation. Any equipment failure or observer problem should be communicated to the supervising field officer for rectification. 5.2.3 Entry of snow data Daily and sub-daily snowfall, snow stake and SWE data are entered manually using the eSWDES snow module in e-SWIS. The user selects the correct station and data series type, and the screen for entry (or editing) of the data is displayed, along with pre-set physical limits of data values for quality control during data entry e.g. for snowfall, the minimum value is 0.0 mm, and a snow day is defined as that day on which the snowfall is more than 0.0 mm. Negative and non-numerical entries are automatically rejected, and the user should add comments where appropriate. Various data entry checks are performed: • Entered snow data are compared against pre-set physical limits for that data series. This

identifies potentially suspect values to the user who can refer back to the field documents to see if there was some error in entering the data. If values which exceed the limits are actually reported in the field documents, the user should add an appropriate comment.

• For daily snowfall, the total monthly snowfall and maximum snowfall in the month entered by the user are compared with the values calculated by the software as the user enters the data. For sub-daily snowfall, the total daily snowfall entered by the user is compared with the value calculated by the software. In the case of a mismatch the user is prompted by colour highlighting and can refer back to the field documents to see if there was some error in entering the data. If cumulated values are also available in the field documents, it becomes quicker to isolate the error.

• For snow stake data, the average monthly snow stake depth and the maximum snowfall in the month entered by the user are compared with the values calculated by the software as the user enters the data.

• For SWE data, the average monthly SWE and the maximum SWE in the month entered by the user are compared with the values calculated by the software as the user enters the data.

Any mismatch remaining after thorough checking of the field documents must be due to incorrect field computations by the observer and should be communicated to the supervising field officer. After entering snow data, the user should also view entered data graphically to identify potentially suspect data not apparent in tabular form, which may reflect an error in data entry. Missing data When data are missing, the corresponding cell is left as -999 (not zero) and a comment entered against that day. Accumulated data Where the observer has missed readings over a period of days and an accumulated total is subsequently measured, the cells corresponding to the missed days are left as -999 (not zero) and a comment entered against the date of the accumulation to specify the period over which the accumulation has occurred (e.g. Accumulated from 23 to 27 Dec). There are occasions when the observer is legitimately absent from her/his station, for example on account of sickness. The observer should be encouraged to leave such spaces “Missing” or “Accumulated” rather than guess the missing values.



Page 43

Digital snow data can be imported directly should an appropriate import interface be available (bespoke to each type of data logger), and can undergo entry checks and be viewed graphically as described above. 5.3 Primary validation 5.3.1 Overview Primary validation is primarily done at a Sub-Divisional office level where staff are in close contact to field staff who have made the observations and/or collected the physical, chart or digital data. Primary validation is carried out using e-SWIS, the data entry module of which replicates the SWDES software from HPI, and is referred to as eSWDES. Primary validation (see Table 3.1)of field data from observational stations is required to be completed at Sub-Divisional office level by the 10th working day of the following month (e.g. for December data by 10th working day in January), ready for secondary validation by State offices. This time schedule ensures that any obvious problems (e.g. indicating an instrument malfunction, observer error, etc) are spotted at the earliest opportunity and resolved. Other problems may not become apparent until more data have been collected, and data can be viewed in a longer-term context. 5.3.2 Typical errors Staff should be aware of typical errors in measurement of snow data, listed in Table 5.1, and these should be considered when interpreting data and possible discrepancies. 5.3.3 Error detection Primary validation of snowfall, snow stake and snow water equivalent (SWE) data focuses on validation within a single data series by making comparisons between individual observations and pre-set physical limits which screen spurious, and potentially suspect, values. Graphical inspection permits detection of errors which are more difficult to identify in tabular form. When a doubtful or incorrect value is identified, and there is any uncertainty as to the correct action, this should be marked with an appropriate flag to indicate that it is suspect. The data flagged as suspect are reviewed at the time of secondary validation. 5.4 Secondary validation Secondary validation of snow data is primarily carried out at State DPCs, to take advantage of the information available from a larger area. Data entering secondary validation have already received primary validation on the basis of knowledge of the station and instrumentation and field documents. Data may have been flagged as missing, accumulated or suspect for some other reason. However, the large spatial variation in snow cover makes quantitative inter-station comparison, a standard secondary validation technique, difficult as data on snow depth and SWE demand much manual validation and verification by integrating data from snow stations, snow gauges and conventional raingauges, beyond the scope of this Handbook. For the Hydrology Project, secondary validation (see Table 3.1) done at State level should be completed by the end of the following month (e.g. for December data by 31st January). Some secondary validation (including comparison with IMD data) will not be possible until the end of the hydrological year when the entire year’s data can be reviewed in a long-term context, so data should be regarded as provisional approved data until then (e.g. for December data by the end of



Page 44

Table 5.1 Measurement errors for snow data Snowfall measurement errors • Observer reads measurement ruler incorrectly • Observer enters snowfall depth incorrectly in the field sheet (e.g. misplacement of decimal point) • Observer enters snowfall depth to the wrong date or time • Observer fabricates readings • Observer did not clear ground/board after previous reading • Observer only takes reading in one place rather than an average so depth is unrepresentative • Observer does not hold measurement ruler vertically whilst taking readings Snow stake measurement errors • Observer reads stow stake incorrectly • Observer enters snow stake depth incorrectly in the field sheet (e.g. misplacement of decimal

point) • Observer enters snow snake depth to the wrong date or time • Observer fabricates readings • Observer cannot see snow stake clearly (e.g. forgets binoculars) • Instrument fault – damaged or broken snow stake SWE measurement errors • For non-recording gauges:

Observer reads measuring glass incorrectly Observer enters amount incorrectly in the field sheet Observer reads gauge at the wrong time (i.e. the correct amount may tgus be allocated to the

wrong day) Observer enters amount to the wrong day Observed total exceeds the capacity of the gauge Catch in snow gauge is under-reported or over-reported due to wind effects

• For recording gauges: Reed switch fails to register tips Reed switch double registers rainfall tips as bucket bounces after tip. (better equipment

includes a debounce filter to eliminate double registration) Failure of electronics due to lightning strike etc. (though lightning protection usually provided) Incorrect set up of measurement parameters by the observer or field supervisor

the hydrological year plus 3 months), after which data should be formally approved and made available for dissemination to external users. 5.5 Analysis Snow data, whether quantitative or qualitative, are important validation data for a wide range of other hydrological variables e.g. anomalous water level data may be explained, and possibly corrected, if background data indicated the nature and extent of snow, and snowmelt, conditions. Factors affecting the contribution of snowmelt to floods include accumulated snow depth at time of melt, ice jamming, basin storage and the return period of the event in question. There are techniques that can be used to estimate the statistical reliability of snow station observations under conditions of melting snow. Degree-day factors (the degree-day approach generates a fixed amount of snowmelt for each degree above a defined temperature) are widely used for correlation purposes. Where snowmelt represents a significant proportion of river flow, established relationships between runoff and SWEs may be used. Air temperature relationships are valuable for the computation of degree-day factors.



Page 45

After the depth of snowmelt has been estimated, it can be treated like rainfall and converted into flow by application of a hydrological model (Water Level, Stage-Discharge and Flow Handbook, Section 6.3.3). There are several operational models that have a snowmelt routine, including HEC-HMS (http://www.hec.usace.army.mil/software/hec-hms/). .

http://www.hec.usace.army.mil/software/hec-hms/�



Page 46

6. Climate Data Processing and Analysis 6.1 Data entry 6.1.1 Overview Entry of data to computer is primarily done at a Sub-Divisional office level where staff are in close contact to field staff who have made the observations and/or collected the chart or digital data. Data entry is carried out using e-SWIS, the data entry module of which replicates the SWDES software from HPI, and is referred to as eSWDES. Prior to entry to computer, two manual activities are essential: registration of receipt of the data, and manual inspection of the climate charts, forms and notebooks from the field, for complete information and obvious errors. Data entry (see Table 3.1) and primary validation of field data from observational stations is required to be completed at Sub-Divisional office level by the 10th working day of the following month (e.g. for June data by 10th working day in July), ready for secondary validation by State offices. 6.1.2 Manual inspection of field records Prior to data entry to computer an initial inspection of field records is required. This is done in conjunction with notes received from the observation station on equipment problems and faults, missing records or exceptional climatic events. Climate sheets and charts are inspected for the following: • Is the station name and code and month and year recorded? • Does the number of record days correspond with the number of days in the month? • Are there some missing values or periods for which values of variables have been

accumulated during absence of the observer? • Have monthly totals or averages of variables been entered? • Have the autographic hourly totals been extracted? Do check manual readings at the

beginning and end agree with the chart values and, if not, has a correction been applied? • Is the record written clearly and with no ambiguity in digits or decimal points? Any queries arising from such inspection should be communicated to the observer to confirm ambiguous data before data entry. Any unresolved problems should be noted and the information sent forward with the digital data to Divisional/State offices to assist in initial secondary validation. Any equipment failure or observer problem should be communicated to the supervising field officer for rectification. 6.1.3 Entry of daily climate data Using the eSWDES module in e-SWIS, the user selects the correct station and daily series. The screen for entry (or editing) of daily climate data is displayed, along with the upper and lower warning levels for each variable, used to flag suspect values (which can be altered for different seasons), and the maximum and minimum values for each variable, for that station. The user selects the correct year and month, and enters each daily climate data value recorded at 08:30 for each date, adding comments where appropriate. Each column is for a different variable and columns can be turned off if that variable is not recorded at a station. Non-numerical entries are automatically rejected. For each month, the user also enters the monthly total or monthly average for each variable, as appropriate, except for wind directions. The software also calculates the monthly totals and averages as the user enters the data. Two types of data entry checks are performed for this case of daily climate data:



Page 47

• The monthly totals and averages entered by the user are compared with the values calculated by the software. In the case of a mismatch the user is prompted by colour highlighting and can refer back to the field documents to see if there was some error in entering the data.

• The entered daily data are compared against the upper and lower warning levels and the maximum and minimum limits for each variable. This identifies potentially suspect values to the user who can refer back to the field documents to see if there was some error in entering the data. If values which exceed the warning levels and the limits are actually reported in the field documents, the user should add an appropriate comment.

Any mismatch remaining after thorough checking of the field documents must be due to incorrect field computations by the observer and should be communicated to the supervising field officer. The user should also view entered data graphically to identify potentially suspect data not apparent in tabular form, which may reflect an error in data entry. Both individual climate variables and combinations of climate variables can be plotted, as bar charts or line graphs as appropriate or, in the case of wind direction, as a rose diagram. Missing data When data are missing, the corresponding cell is left as -999 (not zero) and a comment entered against that day. Accumulated data Where the observer has missed readings over a period of days and an accumulated total is subsequently measured, the cells corresponding to the missed days are left as -999 (not zero) and a comment entered against the date of the accumulation to specify the period over which the accumulation has occurred (e.g. Accumulated from 23 to 27 Sep). There are occasions when the observer is legitimately absent from her/his station, for example on account of sickness. The observer should be encouraged to leave such spaces “Missing” or “Accumulated” rather than guess the missing values. The completion procedures (Section 6.4), based on adjoining information, are better able to estimate such missing values. 6.1.4 Entry of climate data at twice daily interval Using the eSWDES module in e-SWIS, the user selects the correct station and twice-daily series. The screen for entry (or editing) of twice-daily climate data is displayed, along with the upper and lower warning levels for each variable, used to flag suspect values (which can be altered for different seasons), and the maximum and minimum values for each variable, for that station. The user selects the correct year and month, and enters each twice-daily climate data value recorded at 08:30 and 17:30 for each date, adding comments where appropriate. There are two rows for each date, corresponding to the different times, and each column is for a different variable and columns can be turned off if that variable is not recorded at a station. Non-numerical entries are automatically rejected. For each month, the user also enters the monthly total or monthly average for each variable, as appropriate, except for wind directions. The software also calculates the monthly totals and averages as the user enters the data. Two types of data entry checks are performed for this case of twice-daily climate data: • The monthly totals and averages entered by the user are compared with the values calculated

by the software. In the case of a mismatch the user is prompted by colour highlighting and can refer back to the field documents to see if there was some error in entering the data.

• The entered daily data are compared against the upper and lower warning levels and the maximum and minimum limits for each variable. This identifies potentially suspect values to the user who can refer back to the field documents to see if there was some error in entering the data. If values which exceed the warning levels and the limits are actually reported in the field documents, the user should add an appropriate comment.



Page 48

Any mismatch remaining after thorough checking of the field documents must be due to incorrect field computations by the observer and should be communicated to the supervising field officer. The user should also view entered data graphically to identify potentially suspect data not apparent in tabular form, which may reflect an error in data entry. Both individual climate variables and combinations of climate variables can be plotted, as bar charts or line graphs as appropriate or, in the case of wind direction, as a rose diagram. Missing and accumulated data are handled in the same way as for entry of daily climate data (Section 6.1.3). 6.1.5 Entry of hourly climate data Hourly climate data are obtained either from the chart records of thermographs, hydrographs and barographs, and Campbell-Stokes sunshine recorders, or from the digital data of AWSs. Digital data can also be imported directly, but can undergo entry checks and be viewed graphically using this option. Using the eSWDES module in e-SWIS, the user selects the correct station, variable and hourly series. The screen for entry (or editing) of hourly rainfall is displayed, along with the upper and lower warning levels for the variable used to flag suspect values (which can be altered for different seasons), and the maximum and minimum values for the variable for that station. The user selects the correct year and month, and enters the hourly variable values, with each row corresponding to a different day and each column to a different time, adding comments where appropriate. Non-numerical entries are automatically rejected. For each day, the user enters the daily minimum, maximum and average. For each month, the user also enters the minimum, maximum and average for that variable. The software also calculates the daily and monthly minima, maxima and averages as the user enters the data. The entry screen for sunshine data is slightly different in that there are only columns for times between 05:00 and 19:00, as beyond these hours there is no possibility of getting sunshine anywhere in the country. The sunshine value is entered against the time following the hour in which the sunshine occurred e.g. sunshine between 11:30 and 12:30 is recorded against 1230. Two types of data entry checks are performed for this case of hourly data: • The daily and monthly minima, maxima and average values entered by the user are compared

with the values calculated by the software. In the case of a mismatch the user is prompted by colour highlighting and can refer back to the field documents to see if there was some error in entering the data.

• The entered hourly data are compared against the upper and lower warning levels and the maximum and minimum limits. This identifies potentially suspect values to the user who can refer back to the field documents to see if there was some error in entering the data. If values which exceed the warning levels and the limits are actually reported in the field documents, the user should add an appropriate comment.

Any mismatch remaining after thorough checking of the field documents must be due to incorrect field computations by the observer and should be communicated to the supervising field officer. The user should also view entered hourly data for the day and month graphically to identify potentially suspect data not apparent in tabular form, which may reflect an error in data entry. For sunshine, graphs include hourly variation in sunshine during the day and daily variation in sunshine during the month.



Page 49

Missing and accumulated data are handled in the same way as for entry of daily climate data (Section 6.1.3). 6.1.6 Import/entry of digital data Digital data from AWSs takes the form of climate values at a pre-set time interval (e.g. 1 hour, 15 minutes, etc). Data can be imported directly should an appropriate import interface be available (bespoke to each type of data logger), and can undergo entry checks and be viewed graphically as described in Section 6.1.5. 6.2 Primary validation 6.2.1 Overview Primary validation is primarily done at a Sub-Divisional office level where staff are in close contact to field staff who have made the observations and/or collected the chart or digital data. Primary validation is carried out using e-SWIS, the data entry module of which replicates the SWDES software from HPI, and is referred to as eSWDES. Primary validation (See Table 3.1) of field data from observational stations is required to be completed at Sub-Divisional office level by the 10th working day of the following month (e.g. for June data by 10th working day in July), ready for secondary validation by State offices. This time schedule ensures that any obvious problems (e.g. indicating an instrument malfunction, observer error, etc) are spotted at the earliest opportunity and resolved. Other problems may not become apparent until more data have been collected, and data can be viewed in a longer-term context during secondary validation. Primary validation of climate data focuses on validation within a single data series by making comparisons between individual observations and pre-set physical limits, within a single series by making comparisons between sequential observations to detect unacceptable rates of change and deviations from acceptable hydrological behaviour, and between two measurements of a climate variable at a single station (e.g. dry bulb thermometer and thermograph recorder). Examples of many of the techniques described in this section are given in Hydro-Meteorology Training Module 16 “How to carry out primary validation for climatic data”. The final check by comparison with climate variables at neighbouring stations is carried out during secondary validation where more gauges are available for comparison (Section 6.3). 6.2.2 Typical errors Staff should be aware of typical errors in measurement of climate variables, listed in Table 6.1, and these should be considered when interpreting data and possible discrepancies (SW8-OM(I) Chapter 7). 6.2.3 Error detection For primary validation of climate data, graphical inspection permits detection of errors which are more difficult to identify in tabular form. For pan evaporation, upper warning levels and maximum limits are allocated to screen spurious values arising from observer error, leakage, animal interference or dirty water. Where leakage has been identified and is recorded in the field documents, the records for a period preceding the discovery should be inspected and flagged as suspect and for review under secondary validation. For sunshine, maximum and minimum limits will ensure that values of hourly sunshine greater than 1.0 or less than 0.0 are rejected. Sunshine records before 05:00 and after 19:00 are not permitted and, hence, daily totals greater than 14.0 hours are also not permitted. Upper warning



Page 50

Table 6.1 Measurement errors for climate data Pan evaporation measurement errors • Observer errors - the observer over- or underfills the pan - such values should be compensated for the

following day • Instrument errors:

Leakage - this is the most serious problem and it occurs usually at the joint between the base and the side wall. Small leaks are often difficult to detect in the field but may have a significant systematic effect on measured evaporation totals

Animals may gain access to the pan, especially if the wire mesh is damaged Algae and dirt in the water will reduce the measured rate of evaporation Errors arise in periods of high rainfall when the depth caught by the raingauge is different in depth

from the depth caught in the pan as a result of splash or wind eddies round the gauges Sunshine measurement errors • Observer uses wrong chart for time of year which may result in the burn reaching the edge of the chart,

beyond which it is not registered • Observer extracts data from the chart incorrectly Temperature measurement errors • Observer error in reading the thermometer, often error of 1oC (difficult to detect) but sometimes 5oC or

10oC. Such errors are made more common in thermometers with faint graduation etchings • Observer error in registering the thermometer reading • Observer reading meniscus level in minimum thermometer rather than index • Thermometer faults which will result in individual or persistent systematic errors in temperature:

Thermometer fault - breaks in the mercury thread of the dry, wet or maximum thermometer Thermometer fault - failure of constriction of the maximum thermometer Thermometer fault - break in the spirit column of minimum thermometer or spirit lodged at the top or

bubble in the bulb • Thermograph out of calibration and no correction made • Observer fails to correct thermograph for manually observed values at the beginning and end of the

chart period Humidity measurement errors • Measurement errors using dry and wet bulb thermometers are the same as those for temperature • Thermometer faults which will result in individual or persistent systematic errors in temperature e.g. too

high a reading of wet bulb temperature and, therefore, relative humidity: Thermometer fault - the muslin and wick of the wet bulb are not adequately saturated Thermometer fault - the muslin of the wet bulb becomes dirty or covered by grease

• Hygrograph out of calibration and no correction made • Observer fails to correct hygrograph for manually observed values at the beginning and end of the chart

period Wind speed measurement errors • Observer reads or records counter total incorrectly • Observer makes arithmetic errors in the calculation of wind run or average wind speed • Instrument faults caused by poor maintenance • Instrument faults caused by damage to the spindle which might thus result in reduced revolutions for

given wind speed Atmospheric pressure measurement errors • Observer reads or records barometer reading incorrectly (observation problems can result from the use

of the Vernier scale) • Observer makes arithmetic errors in corrections for temperature or reduction to sea level. • Barometer faults from entry of air into the space above the mercury • Barometer faults from mechanical defects in the Vernier head • Barograph out of calibration and no correction made • Observer fails to correct barograph for manually observed values at the beginning and end of the chart

period



Page 51

levels may be set seasonally based on the maximum possible sunshine for the location and time of year. For temperature, upper and lower warning levels and maximum and minimum limits appropriate to the location and season, and each thermometer, are allocated to screen spurious values. In addition: • Dry bulb temperature should be greater than or (rarely) equal to the wet bulb temperature • Maximum temperature should be greater than minimum temperature • Maximum temperature measured using the maximum thermometer should be greater than or

equal to the maximum temperature recorded by the dry bulb during the interval, including the time of maximum observation. The value of the maximum should be set to the observed maximum on the dry bulb if this is greater

• Minimum temperature measured using the minimum thermometer should be less than or equal to the minimum temperature recorded by the dry bulb during the interval, including the time of minimum observation. The value of the minimum should be set to the observed minimum on the dry bulb if this is lower

• Thermograph readings should agree with dry bulb thermometer readings at corresponding times.

For humidity, upper and lower warning levels appropriate to the location and season are allocated to screen spurious values. The maximum limit is 100%. Hygrograph readings should agree with calculated values of humidity at corresponding times. Where an observer has to calculate relative humidity to calibrate the hygrograph, the observer-calculated values may also be compared with those calculated by the software. Because of extreme variability in wind speed in space and time, it is difficult to set up convincing rules to detect suspect values. Nevertheless simple checks include: wind speeds should be zero where the direction is reported as “0” (i.e. calm); wind speeds cannot exceed 5 kmh-1 when the wind speed is reported as variable; and wind speeds in excess of 200 kmh-1 should be considered suspect. For atmospheric pressure, upper and lower warning levels and maximum and minimum limits are allocated to screen spurious values. Barograph readings should agree with barometer readings at corresponding times. Where a doubtful or incorrect value is identified, and there is any uncertainty as to the correct action, this should be marked with an appropriate flag to indicate that it is suspect. The data flagged as suspect are reviewed at the time of secondary validation. 6.3 Secondary validation 6.3.1 Overview Secondary validation of climate data is primarily carried out at State DPCs, to take advantage of the information available from a larger area. Secondary validation is carried out using e-SWIS, the validation module of which replicates the HYMOS software from HPI, and is referred to as eHYMOS. Data may also be exported to Excel for secondary validation. For the Hydrology Project, secondary validation (see Table 3.1) done at State level should be completed by the end of the following month (e.g. for June data by 31st July). Some secondary validation (including comparison with IMD data) will not be possible until the end of the hydrological year when the entire year’s data can be reviewed in a long-term context, so data should be regarded as provisional approved data until then (e.g. for June data by the end of the hydrological year plus 3 months), after which data should be formally approved and made available for dissemination to external users.



Page 52

Data entering secondary validation have already received primary validation on the basis of knowledge of the station and instrumentation and field documents. Data may have been flagged as missing, accumulated, shifted or suspect for some other reason. Secondary validation focuses on comparisons with neighbouring stations to identify suspect values. Although secondary validation is designed to detect anomalies in time series, such as trends or changes in spatial relationships, these may result from genuine environmental change as well as data error. Data processing staff should not adjust climate data unless they are confident that the data are incorrect and not reflecting a changed climate or micro-climate. Spatial validation of climate data is not so much concerned with individual values as with the generality of values received from a station, often best illustrated by the use of aggregate values. However, data processing staff should continue to be aware of field practice and instrumentation and the associated errors which can arise in data. Some of the secondary validation checks are oriented towards the specific types of errors just mentioned, whilst others are general in nature and lead to identification of spatial inconsistencies in the data. Many of the techniques for the secondary validation of climate data are the same as those for the secondary validation of rainfall data. Examples of the techniques described in this section are given in Hydro-Meteorology Training Module 17 “How to carry out secondary validation of climatic data” and Training Module 09 “How to carry out secondary validation of rainfall data”. 6.3.2 Multiple station validation methods • Comparison plots - Comparative time series plots are an effective visual method for

identifying potential anomalies between stations and should be the first validation test that is carried out. For climate data, the same variable at the same time from two or more stations should be displayed as a line graph, and the plot should also include data from at least the previous month to ensure that there are no discontinuities between one batch of data received from the station and the next, which would be a possible indication that the data had not been allocated to the correct station. For climate variables with a strong spatial correlation, series from neighbouring stations should generally run along closely parallel, with the mean separation representing some location factor e.g. altitude. Abrupt or progressive straying from this pattern, evident from the comparative plot, would not necessarily have been perceived at primary validation from inspection of a single station. Comparison of series may also permit the acceptance of values flagged as suspect at primary validation because they fell outside the warning range. Where two or more stations display the same behaviour, there is strong evidence to suggest that the values are correct.

Comparison plots provide a simple means of identifying anomalies but not of correcting them, which should be done through regression analysis, spatial homogeneity testing (nearest neighbour analysis) or double mass analysis.

• Balance series - An alternative method of displaying comparative time series is to plot the

difference between two or more time series, to detect anomalies and to flag suspect values or sequences. Anomalous values are displayed as departures from the mean difference line. Considering Zi as the balance series of the two series Xi and Yi , the computations can be done as:

Zi = Xi - Yi

• Derivative series - A method of assessing temporal consistency is to plot the magnitude of

the rate of change between consecutive time steps (i.e. a series derived from the original time series), to detect anomalies and to flag suspect values or sequences. Anomalous values are displayed as exceedances of the maximum rate of rise and maximum rate of fall. Considering Zi as the derivative series of the series Xi, the computations can be done as:



Page 53

Zi = Xi - Xi-1

• Regression analysis - For climate variables where individual or short sequences of anomalous values are present in a spatially conservative time series, regression analysis, a commonly used statistical technique, may provide a sufficient basis for interpolation through a simple linear relationship with a neighbouring station of the form:

Yi = a Xi + c

In a graphical plot, any suspect values will generally show up as outliers but, in contrast to the comparison plots, regression analysis provides no indication of the time sequencing of the suspect values and whether the outliers were scattered or contained in one block. Previously identified suspect values should be removed before deriving the relationship, which may then be applied to compute corrected values to replace the suspect ones. For more information see SW8-OM(III) Chapter 2.

• Spatial homogeneity testing – Although there are generally more rainfall stations in the

vicinity of a target station than there are climate stations, spatial homogeneity testing (Section 4.3.4) is still a useful approach because there is less spatial variability for climate variables so that the area over which comparison is permitted (the maximum correlation distance Rmax) may be increased. Spatial interpolation is made by weighing the adjoining station rainfall as inversely proportional to some power (e.g. 2) of the distances from the station under consideration. However, although there is strong spatial correlation, there may be a systematic difference due, for example, to altitude for temperature, in which case, normalised values rather than actual values should be used in the interpolation. The observed temperature values at the neighbouring stations are multiplied by the ratio of the average annual temperature at the test station and the average annual temperature at the neighbouring stations:

Where: Tci = for corrected temperature value at the neighbouring station i

Ntest = annual average temperature at the station under consideration Ni = annual average temperature at the neighbouring station

Spatial homogeneity testing provides a list of those values (flagged values “+” or “-”) which are outside a specified range (mean + standard deviation times a multiplier), and provides the option of replacing observed values with estimated values.

• Double mass analysis - Double mass analysis (Section 4.3.5) may also be used to show

trends or inconsistencies between climate stations, and it is usually used with longer, aggregated series. In the case of a leaking evaporation pan, a double mass plot of daily values for a period commencing some time before leakage commenced, will reveal the leakage as a curvature in the plot. However, double mass analysis cannot be used to correct for such a progressive departure from previous behaviour, and may only be used to correct suspect values where there has been a systematic but constant shift in the variable at the station in question.

6.3.3 Single station validation methods Single series testing for homogeneity involves statistical analysis and are likely to be of interest only to experienced hydro-meteorologists at State level and at IMD. Furthermore, it will normally only be used with long datasets and, therefore, has to await the data entry of historical data. Time series may be inspected graphically for evidence of trend, but statistical hypothesis testing can be more discriminative in distinguishing between expected variation in a random series and real trend or more abrupt changes in the characteristics of the time series:



Page 54

• Trend analysis - A climate time series can be considered homogeneous if there is no

significant linear or curvilinear trend. Generally, trend does not become evident for a number of years and so tests should be carried out on long data series, often aggregated into monthly or annual series. Trend may result from a wide variety of factors including changes in station location, instrumentation, observer or observation practice, growth of vegetation or nearby new buildings affecting exposure of the station, effects of new irrigation in the vicinity of the station (affecting humidity, temperature and pan evaporation), effects of the urban heat island with growing urbanisation, and global climate change.

The presence of trend may be examined by graphical display and statistical tests, where the data are plotted on a linear or semi-logarithmic scale with the climate variable on the Y-axis and time on the X-axis. The coefficients of the fitted regression line can be tested for statistical significance. The presence of trend does not necessarily mean that part of the data are faulty but that the environmental conditions have changed. Unless there is reason to believe that the trend is due to a systematic change, the data should not generally be altered but the existence of trend noted in the station record.

• Residual mass curve - A residual mass curve represents accumulative departures from the

mean. It is an effective visual method of detecting climatic variabilities or other inhomogeneities. An upward curve indicates an above average sequence, a horizontal curve an about average sequence, and a downward curve indicates a below average sequence. The residual mass curve is derived as:

Where: N = number of elements in the series

mx = mean value of Xi , i=1,N • Hypothesis testing - Hypothesis testing forms a framework for many statistical tests. An

assumption about the distribution of a statistical parameter (e.g. the mean of a climate time series) is stated in the null-hypothesis H0 and is tested against an alternative formulated in the H1 hypothesis. The statistical parameter under investigation is called the test statistic. Under the null-hypothesis, the test statistic has some standardised sampling distribution e.g. standard normal distribution. For the null hypothesis to be true, the value of the test statistic should be within the acceptance limits of the sampling distribution of the parameters under the null-hypothesis. If the test statistic does not lie within the acceptance limits, expressed as a significance level, the null-hypothesis is rejected and the alternative is assumed to be true. Some risk is involved however in being wrong in accepting or rejecting the hypothesis. Common tests include Student’s t-test, Wilcoxon W-test, and Wilcoxon-Mann-Whitney U-test. See the e-SWIS/eHYMOS manual for details about which tests are available. For more information see SW8-OM(II) Chapters 5.5.4 to 5.5.6.

6.4 Correction and completion 6.4.1 Overview Completion – the processing of filling in missing values and correcting erroneous values – is done as a continuous process with primary and secondary validation. Although the HIS Manual SW separates correction and completion in SW8-OM(II) Chapter 6 from secondary validation in SW8-OM(II) Chapter 5, and from primary validation in SW8-OM(I) Chapter 7, there is substantial overlap between the techniques used. In this Handbook, some correction and completion techniques have



Page 55

been included in the appropriate parts of Sections 6.2 and 6.3, and others are described below. The techniques should be applied by experienced staff with appropriate training. The majority of secondary validation, and therefore the majority of correction and completion, is carried out by State DPCs to take advantage of the information available from a larger area. For the Hydrology Project, correction and completion (see Table 3.1) should be completed by the end of the following month (e.g. for June data by 31st July). Some secondary validation, correction and completion will not be possible until the end of the hydrological year when the entire year’s data can be reviewed in a long-term context, so data should be regarded as provisional approved data until then (e.g. for June data by the end of the hydrological year plus 3 months), after which data should be formally approved and made available for dissemination to external users. Procedures for correction and completion of climate data are generally based on regression analysis, as the spatial correlation of climate variables is considerable in many instances, though may vary with season. Correlation with nearby stations should always be checked. The regression estimate may be used either to adjust or replace an erroneous entry or to fill in a gap. It should be recognised that values estimated from other gauges are inherently less reliable than values properly measured. There will be circumstances where no suitable neighbouring observations or stations are available, such that missing values should be left as -999 and incorrect values should be set to -999, and suspect original values should be given the benefit of the doubt and retained in the record with an appropriate flag. In all cases, the climate data should be revalidated after correction and/or completion, to ensure that the series remains consistent with the other climate variables observed at the site and with the same variable measured at nearby stations. 6.4.2 Correcting pan evaporation data Pan evaporation data are corrected or infilled using evaporation estimates obtained from the Penman method computed from other climate variables. The pan coefficient for the season has to be derived first, using the available reliable records on climate data and pan evaporation. 6.4.3 Correcting sunshine data A comparison of records at neighbouring stations should reveal whether the number of sunshine hours shows good correlation or not. If a good correlation is available, regression analysis is the appropriate tool to adjust/replace erroneous values and to fill in missing ones. Otherwise, the records from ARGs and TBRs are particularly important as they can put an upper limit on the number of sunshine hours. Data corrected by either technique should always be rechecked and, if necessary, readjusted against the rainfall record at the same station and/or neighbouring stations. 6.4.4 Correcting temperature data The correction of temperature data depends on the type of error. If the thermometer has been misread, generally by a full number of oC (e.g. 1 or 5 or 10 oC), the correction to be applied may be deduced from a combination of: • Time series graph of the temperature variable, observing surrounding values in the graph • Multiple time series graph of related temperature parameters • Multiple time series graph of the same temperature variable observed at neighbouring stations If the thermograph has been out of calibration for some time, the correction to be applied may be obtained from a graphical comparison of daily maximum and minimum maximum temperatures from the thermograph with daily maximum and minimum thermometer readings at the station (or at neighbouring stations), to detect any sudden deviation, followed by a double mass analysis to establish the start date of malfunctioning and the type and size of error. Corrections may be



Page 56

applied in a similar fashion (step-trend or linear trend is most common). It is important to check whether any deviations are a function of temperature. If temperature data are missing, the gap(s) may be infilled by: • In the case that a number of data are missing for maximum, minimum, dry and/or wet bulb

temperatures and no thermograph record is available, regression analysis with neighbouring stations’ temperature records is used to match with the last value before the gap and the first value after.

• In the case that a number of data are missing from the wet bulb temperature record, a relationship between wet bulb temperature and dry bulb temperature, relative humidity and psychometric constants is developed from regression analysis with relative humidity records. The equation has to be solved iteratively. Neighbouring stations may be used if necessary.

• In the case that one or more data are missing from the thermograph record, interpolation within the time series may be used if neighbouring stations do not show any abnormal changes with time.

• In the case that a number of data are missing from the thermograph record, a relationship developed from regression analysis with neighbouring stations may be adjusted to match with the maximum and minimum temperatures measured at the thermograph location.

The corrected and/or infilled record should be subjected to consistency tests such as: the maximum temperature should be greater than the minimum temperature, and dry bulb temperature should be greater than or (rarely) equal to the wet bulb temperature, though on wet days their difference should be small as the humidity is likely to be high. 6.4.5 Correcting humidity data Historical records of relative humidity can be corrected by recomputing the variable from the dry and wet bulb temperatures, if available. Otherwise, if a good correlation between humidity at nearby stations is available, regression analysis is the appropriate tool to adjust/replace erroneous values and to fill in missing ones. Hygrometer records can be corrected or infilled by means of regression on nearby stations, adjusted by local instantaneous values from dry and wet bulb temperatures when available. The infilled record should always be rechecked for consistency and continuity at the start and end of the gap. 6.4.6 Correcting wind data A good correlation between winds speed and direction at neighbouring FCSs is expected and, hence, regression analysis is the appropriate tool to adjust/replace erroneous values and to fill in missing ones. 6.4.7 Correcting atmospheric pressure data A good correlation between atmospheric pressure at neighbouring FCSs is expected and, hence, regression analysis is the appropriate tool to adjust/replace erroneous values and to fill in missing ones. 6.5 Analysis 6.5.1 Overview The principal climate variables of interest in hydrology are evaporation from a free water surface and potential evapotranspiration i.e. the water loss which will occur from a surface fully covered by green vegetation if there is at no time a deficiency of water in the soil for use by the vegetation.



Page 57

Evaporation estimates may be based on measurement of losses from an evaporation pan (for open water evaporation), or on theoretical and empirical methods based on climatological data (for potential evapotranspiration). The methods range from regression type equations to more detailed approaches such as those representing water budget, energy budget and mass transfer approaches, as in the Penman method which is widely used in India. The majority of estimation (see Table 3.1) is carried out by State DPCs. For more information see Hydro-Meteorology Training Module 19 “How to analyse climate data” and SW8-OM(III) Chapter 5. 6.5.2 Analysis of pan evaporation Evaporation measured by pans does not represent the evaporation from large water bodies such as lakes and reservoirs because pans have the following limitations: • Pans differ from lakes and reservoirs in the heat storage characteristics and heat transfer.

Pans exposed above ground are subject to heat exchange through the sides • The height of rim in an evaporation pan affects the wind action over the surface • The heat transfer characteristics of the pan material is different from that of the reservoir Since heat storage in pans is small, pan evaporation is nearly in phase with climate, but in the case of very large and/or deep lakes, the time lag in lake evaporation may be up to several months. Estimates of lake evaporation can be obtained by application of the appropriate pan coefficient to measured pan evaporation. The pan coefficient is given by El / Ep where El is the evaporation from the lake and Ep is the evaporation from the pan. Pan coefficients show considerable variation from place to place and from month to month for the same location, the latter precluding the use of a constant pan coefficient. Monthly pan coefficients depend on climate and on lake size and depth, and generally range from 0.6 to 0.8 (with the average value used generally 0.7). For dry seasons and arid climates, the pan water temperature is less than the air temperature and pan coefficients may be 0.60 or less, whilst for humid seasons and climates where the pan water temperature is higher than air temperature, pan coefficients may be 0.80 or higher. The average pan coefficient for the Indian Standard pan is 0.8, ranging from 0.65 to 1.10. A further correction (commonly 1.144) is applied to take into account the reduction in pan evaporation caused by the mesh screen over evaporation pans in India. However, it is preferable to retain the original measured values and leave mesh corrections to data users to allow for the possibility that future amendments may be made to the correction factor. Pan evaporation may also be used to estimate reference crop evapotranspiration using a pan coefficient based on relative humidity, daily wind run and fetch. 6.5.3 Estimation of Penman evapotranspiration The Penman method, in wide use in India for estimation of potential evapotranspiration, arose from earlier studies of methods to estimate open water evaporation. In turn, both depend on the combination of two physical approaches: • The mass transfer method, sometimes called the vapour flux method, which calculates the

upward flux of water vapour from the evaporating surface • The energy budget method which considers the heat sources and sinks of the water body and

air and isolates the energy required for the evaporating process The disadvantage of these methods is the requirement for data not normally measured at standard climatological stations. To overcome this difficulty, Penman (1948) developed a formula for calculating open water evaporation, combining the physical principles of the mass transfer and energy budget methods with some empirical concepts, to enable standard meteorological



Page 58

observations to be used. The Penman method was subsequently adapted to estimate potential evapotranspiration and to substitute alternative more commonly measured climate variables for those less commonly measured. For more information see SW3-DM Chapter 2.3 and SW8-OM(III) Chapter 5.3



Page 59

7. Data Dissemination and Publication 7.1 Hydro-meteorological products The traditional primary visible output of hydrological data archives is published reports, usually in the form of annual hydrological yearbooks. However, this is not generally the most convenient format of hydro-meteorological data for data users who often require long-term records for a single station or a group of stations i.e. data by station rather than by year. For data users in the past, this necessitated the collation of data from a set of annual reports and the keying in of the data for the required analysis. In many countries, recent advances in IT, combined with well-established links between data suppliers and data users, mean that annual reports are no longer published in print, with the same information being provided online, and data requests met with a rapid and bespoke response. A further consequence is that data suppliers have more time to focus on data analysis, periodic reports and short-term operational reports of interest to key data users e.g. reports on unusual rainfall events, snow reports, rainfall and evaporation data in bulletins for agricultural and irrigation design and operation, real-time rainfall data for flood forecasting and for hydropower and reservoir operation, etc. A combination of digital and hardcopy hydrological products and online dissemination provide an effective means of demonstrating the capability of the HIS, in particular: • Providing information on availability of data for use in planning and design, and making

reporting and use of data more efficient by reducing the amount of published data and cost of annual reports

• Advertising the work of the HIS and its capability, and to create interest and awareness amongst potential data users

• Providing tangible evidence to policymakers of a return on substantial investment • Providing feedback to data producers, and acknowledging the contribution of observers and

co-operating agencies • Providing a clear incentive to keep archives up to data and a focus for an annual hydrometric

audit Hence, the long-term goal of the HIS is web-based dissemination of user guidance and station metadata (additional datasets that include items that could assist users of the data to understand the data, their accuracy and any major influencing factor), which is usefully complemented by the publication of catalogues or registers of hydrometric stations (e.g. Marsh & Hannaford, 2008) and occasional reports, and by a dedicated enquiry and data retrieval service. 7.2 Annual reports 7.2.1 Hydrological yearbooks Hydrological yearbooks should report over the hydrological year from 1 June to 31 May. The hydrological year corresponds to a complete cycle of replenishment and depletion, so it is appropriate to report on that basis rather than over the calendar year. Annual flow, precipitation and climate data may be presented in a single combined report. The hydro-meteorological elements of such reports incorporate a summary of information on the pattern of rainfall, snow (where relevant) and evaporation over the year, and information on the long-term spatial and temporal pattern of rainfall, snow and evaporation in the region and how the recent year compares with past statistics (see Table 3.1). Whilst a broad range of climate variables are measured at FCSs, the majority are used for computing evaporation and Penman evapotranspiration, and it is the statistics of these products that are included in the reports, not the statistics of the individual



Page 60

climate variables. Annual reports are produced at the State DPC and should be published within 12 months of the end of the hydrological year covered. SW8-OM(III) Annex I and GW8-OM(V) present templates for Surface Water and Groundwater Yearbooks published at State level. The following are typical contents: • Introduction – The report introduction should describe the administrative organisation of the

rainfall, snow and climate networks and the steps involved in the collection, data entry, processing, validation, analysis and storage of data, including any agencies contributing to the included data. Standard climatic observation practice should be summarised. The report should explain how the work is linked with other agencies collecting or using rainfall, snow and climate data including IMD. It should also set out how data may be requested and under what terms and conditions they are supplied. The report introduction may change little from year to year.

• Observational network – Maps and tables should be used to summarise the salient features

of the observational network. The rainfall station map should also show major rivers and basin boundaries and distinguish each site by symbol between SRG, ARG and TBR, and whether rainfall alone is observed or the raingauge is sited at an FCS or AWS. The snow station map should, similarly, show each site with, if appropriate, some distinction for different instrumentation or monitoring. The climate station map should also show major rivers and basin boundaries and distinguish each site by symbol between the combinations of instruments in use at each station e.g. FCS, AWS, the type(s) of raingauge, etc. Mapped stations should be numbered so that they can be related to information contained in tabular listings. Tables of current stations should be listed by named basin and sub-basin, as well as the latitude, longitude, altitude, responsible agency, the full period of observational record and the period of observation which is available in digital format. A similar listing of closed stations (or a selection of closed stations with long records) may also be provided. All additions and closures of stations should be highlighted in the yearly report. Similarly, station upgrading and the nature of the upgrading should be reported.

• Descriptive account of rainfall occurrence during the report year - An account of rainfall

occurrence in the region in the year should be given in the form of a concise commentary for each month, placed in its meteorological context. Significant stretches of dry or wet periods in the parts of the region should be highlighted. A summary of the hydrological impact of rainfall with particular reference to floods and droughts should also be included.

• Thematic maps of monthly, seasonal and annual rainfall – Such maps provide a summary

of the rainfall pattern in space and time. Basin or administrative boundaries may also be shown to illustrate variations between districts or basins. The rainfall should be mapped as the actual value at each station for the specified period, or by the drawing of isohyets of equal rainfall over the region using a technique such as point kriging.

• Graphical and mapped comparisons with average patterns – Such maps present the

season’s and/or year’s rainfall, snow and evaporation as a percentage of the long-term average. The period over which the long-term average is taken should be noted. For a few representative stations, a graphical comparison of the monthly rainfall, snow and evaporation amounts for the whole year can be made with the long-term average patterns by plotting the actual monthly distribution against the long-term average for minimum, maximum and average monthly amounts. This kind of plot also facilitates comprehension of the temporal distribution of these variables during the hydrological year.

• Basic rainfall statistics – This section forms the core of the report. Full reporting of daily or

hourly data is no longer required, though sample tabulations of daily and hourly data may be provided for a few representative rainfall stations to illustrate the format of information available. Instead, summary statistics of monthly rainfall for the report year provide a means of



Page 61

making comparisons between stations and between months, and should satisfy the needs of general data users. Tables of current stations should be listed by named basin and sub-basin, and include monthly rainfall totals, the maximum daily amount in the year and the date of its occurrence. Any daily, monthly or annual totals which exceed previous maxima of record should be highlighted. For ARG/TBR stations, the maximum observed amount for selected durations including 1 hour, 2, 3, 6, 12 and 24 hours with dates of occurrence should be included.

• Snow reporting where relevant – This could also be published as a separate snow report.

For an example, see http://www.metoffice.gov.uk/archive/snow-survey. Snow reporting should include: Descriptive account of snow observations during the report year in the form of a summary

of the number of days with snow falling and the number of days with snow lying at key stations, supported by a.concise commentary for each month, placed in its meteorological context. A summary of the hydrological impact of snowfall and snowmelt with particular reference to floods should also be included.

Basic snow statistics – Maps may include monthly, seasonal and annual snow cover or extent. Tables of current stations should be listed by named basin and sub-basin, and should include: the number of days with snow falling and the number of days with snow lying at key stations for the current year and previous years to place snow observations in a historical context; the daily total depth of snow of key stations for each day and month; and for each station and month, the number of days with snow falling, the number of days with snow lying, the maximum accumulation of undrifted snow in the year and the date of its occurrence. Any figures which exceed previous maxima of record should be highlighted. Graphs should include: plots of the snowline for key stations for the current year.

• Basic evaporation statistics – Full reporting of daily data is no longer required, though

sample tabulations of daily data may be provided for a few representative climate stations to illustrate the format of information available. Tables of current stations should be listed by named basin and sub-basin, and include for the current year, monthly and annual pan evaporation and Penman evapotranspiration. Summary statistics for previous years should include: mean monthly and annual pan evapotranspiration; lowest monthly mean in period; highest monthly mean in period; various percentile values; mean monthly and annual Penman evapotranspiration; lowest monthly mean in period; highest monthly mean in period; and various percentile values. Values of evaporation should be reported to no more than 1 decimal place.

• Description and statistical summaries of major storms - Major storms which are known to have had an impact on flooding or operation of water resources should be described in more detail. Selection of events for description may be made in terms of impact or on an objective basis of areal amount and distribution. Storms should be described with respect to their meteorological context, centre of concentration, movement across the river basins and also the characteristics of the time distribution of rainfall within the storm. The description may be combined with a report of the flooding consequences for the storm.

• Data validation and quality - The limitations of the data should be made known to data users.

The validation process not only provides a means of checking the quality of the raw data but also a means of reporting. For evaporation, the general limitations of pan evaporation as a measure of open water evaporation should be explained. The number of values corrected or infilled as a total or a percentage may be noted for individual stations, by basin or by agency. The types of anomaly typically detected by data validation and remedial actions may be described.

• Bibliography - Data users may be interested to know of other sources of rainfall, snow and

climate data, or of other hydrological data, including: concurrent annual reports from the HIS of

http://www.metoffice.gov.uk/archive/snow-survey�



Page 62

other hydro-meteorological or hydrological data, and previous annual reports (with dates) from the HIS or other agencies; any periodic reports of hydro-meteorological station metadata and time series statistics produced by the HIS or other agencies; and any special reports produced by the HIS or other agencies. A brief note on the administrative context of previous reports, methods of data compilation, and previous report formats may be helpful.

7.2.2 Annual hydrological reviews Shorter than hydrological yearbooks, annual reviews of the hydrological year provide users with published assessments of the key elements of the hydrological cycle. Hence, the reports combine rainfall, snow (where relevant), climate, flow, reservoir stocks and groundwater, and possibly also water quality. Annual reviews are produced at the State DPC and should be published within 12 months of the end of the hydrological year covered. For an example, see www.ceh.ac.uk/data/nrfa/nhmp/annual_review.html. 7.3 Periodic reports 7.3.1 Metadata catalogues Periodic reports of raingauge, snow station (where relevant) and climate station metadata and time series statistics may be published by the State DPC at 5-year or 10-year intervals. The reports should incorporate spatial as well as temporal analysis and provide statistical summaries in tabular and graphical to make the information accessible and interesting to data users. The following are typical contents of such a periodic report: • Introduction • Data availability - maps and tabulations • Descriptive account of annual rainfall, snow, measured pan evaporation and computed

Penman evapotranspiration since last periodic report • Thematic maps of mean monthly and seasonal rainfall, snow and evaporation • Basic rainfall statistics - monthly and annual means, maxima and minima: for the standard

climate normal period (1961-90) where available; for the updated decade; and for the available period of record

• Basic snow statistics where relevant - monthly and annual means, maxima and minima: for the standard climate normal period (1961-90) where available; for the updated decade; and for the available period of record

• Basic evaporation statistics - monthly and annual means, maxima and minima: for the standard climate normal period (1961-90) where available; for the updated decade; and for the available period of record

• Additional point rainfall statistics e.g. daily maximum rainfall, persistence of dry or wet spells during the monsoon, dates of onset or termination of the monsoon

• Additional areal mean rainfall statistics for administrative or drainage areas for periods of a month or year

• Analysis of temporal variability of rainfall using moving averages or residual mass curves to identify major wet and dry periods for a number of representative stations

• Analysis of snowfall and snowmelt, and its contribution to flow • Analysis of periodicity and trend in evaporation data • Frequency analysis of rainfall data - the primary responsibility for making such maps lies with

IMD

http://www.ceh.ac.uk/data/nrfa/nhmp/annual_review.html�



Page 63

7.3.2 Monthly hydrological summaries Routine monthly reports and statistics on the current state hydrological situation, including assessments of rainfall, snow (where relevant), evaporation, river flow, groundwater and reservoir stocks, provide users with a snapshot of the current situation and its historical context, and the future outlook. Such information may provide a vital input for planning domestic or industrial water supply, agricultural planning, hydropower and other water use sectors. Monthly summaries are produced at the State DPC and should be published within 10 working days of the month covered. For an example, see www.ceh.ac.uk/data/NRFA/nhmp/monthly_hs.html. 7.4 Special reports Occasional special reports should also be published by the State DPC providing reactive analysis in the aftermath of notable or significant extreme storms or monsoon events. As these may also have unusual hydrological consequences, the reports are normally combined with reports of the resulting streamflow and flooding within the affected area. For an example see, www.ceh.ac.uk/data/nrfa/nhmp/other_reports.html. 7.5 Dissemination to hydrological data users Final (approved) hydro-meteorology datasets are provided by Central/State hydrometric agencies on a request basis. The online HIS data catalogue in e-SWIS, which shows the availability of fully validated (approved) data, supports hydrometric agencies in disseminating their data, and also helps hydrological data users to search available data and formulate their data requests and the formats required and direct them to the appropriate agency. The more comprehensive the information a data catalogue provides, the easier for users to identify the monitoring stations of interest to them, and be aware of any limitations to exploiting the data effectively. Users should be informed of the quality of any data supplied indicated by the data flag (e.g. observed, estimated, suspect, etc). There may be a charge for data which is the product of significant investment in equipment and staff time. Data requests from users should be processed promptly: at least 95% of queries should be dealt with within 5 working days, and the remaining up to 5% of queries, which should be the more complex ones, within 20 working days.

http://www.ceh.ac.uk/data/NRFA/nhmp/monthly_hs.htmlx�

http://www.ceh.ac.uk/data/nrfa/nhmp/other_reports.html�



Page 64

References Hannaford, J., Holmes, M.G.R., Laize, C.L.R., Marsh, T.J. & Young, A.R. 2013. Evaluating hydrometric networks for prediction in ungauged basins: a new methodology and its application to England and Wales. Hydrology Research 44 (3), 401-418. Institute of Hydrology. 1999. Flood Estimation Handbook, 5 Volumes. Institute of Hydrology, Wallingford, UK. Linacre, E. 1992. Climate Data and Resources, 366pp. Routledge, London, UK. Marsh, T.J. 2002. Capitalising on river flow data to meet changing national needs — a UK perspective. Flow Measurement and Instrumentation 13, 291–298. Marsh, T.J. & Hannaford, J. (eds). 2008. UK Hydrometric Register. Centre for Ecology & Hydrology, Wallingford, UK. www.ceh.ac.uk/products/publications/documents/hydrometricregister_final_withcovers.pdf Singh, P. & Singh, V.P. 2001. Snow and Glacier Hydrology, 742pp. Water Science and Technology Library, Kluwer Academic Publishers, Netherlands. World Meteorological Organisation. 2008. Guide to Hydrological Practices Volume I: Hydrology – From Measurement to Hydrological Information. Report No 168. www.whycos.org/hwrp/guide/ World Meteorological Organisation. 2009. Guide to Hydrological Practices Volume II: Management of Water Resources and Application of Hydrological Practices. Report No 168. www.whycos.org/hwrp/guide/ World Meteorological Organisation. 1992. Snow Cover Measurements and Areal Assessment of Precipitation and Soil Moisture. Report No 749.

http://www.ceh.ac.uk/products/publications/documents/hydrometricregister_final_withcovers.pdf�

http://www.whycos.org/hwrp/guide/�

http://www.whycos.org/hwrp/guide/�



Page 65

Annex I States and Agencies participating in the Hydrology Project Phase I (1996-2003) Phase II (2006-2014) States States Andhra Pradesh Andhra Pradesh Chhattisgarh Chhattisgarh Goa Gujurat Gujurat Himachal Pradesh Kerala Kerala Karnataka Karnataka Madhya Pradesh Madhya Pradesh Maharastra Maharastra Orissa Orissa Pondicherry Punjab Tamil Nadu Tamil Nadu Agencies Agencies Bhakra-Beas Management Board (BBMB) Central Ground Water Board (CGWB) Central Ground Water Board (CGWB) Central Pollution Control Board (CPCB) Central Water and Power Research Station

(CWPRS) Central Water and Power Research Station

(CWPRS) Central Water Commission (CWC) Central Water Commission (CWC) Indian Meteorological Department (IMD) Indian Meteorological Department (IMD) Ministry of Water Resources (MoWR) Ministry of Water Resources (MoWR) National Institute of Hydrology (NIH) National Institute of Hydrology (NIH)



Page 66

Annex II Summary of Distribution of Hard Copy of HPI HIS Manual Surface Water

Volume Manual Part Training State

DSC State DPC

Div DPC

Sub-Div DPC

Station/ Lab

1 HIS Design + + + + + Field I Job description + + + + +

II ToR for HDUG + + + + III Data needs assessment + + + +

Reference + + + 2 Sampling Principles

Design + + + + + Reference + + +

3 Hydro-meteorology

Design + + + + + Field I Network design & site

selection + + + + +

II SRG operation & maintenance

+ + + + + 1 SRG

III ARG/TBR/SRG operation & maintenance

+ + + + + 1 ARG

IV FCS operation & maintenance

+ + + + + 1 FCS

V Field inspections, audits, maintenance & calibration

+ + + + +

Reference + + + 4 Hydrometry Design + + + + +

Field I Network design & site selection

+ + + + +

II River stage observation + + + + + + III Float measurements + + + + + + IV Current meter gauging + + + + + + V Field application of ADCP + + + (+) + (+) VI Slope-are method + + + + + + VII Field inspection & audits + + + + + VIII Maintenance & calibration + + + + +

Reference + + + 5 Sediment Transport

Design + + + + + Field + + + + + + Reference + + +

6 WQ Sampling

Design + + + + + Field + + + + + +

7 WQ Analysis Design + + + + Operation + + + +

8 Data Processing & Analysis

Operation I Data entry & primary validation

+ + + + +

II Secondary validation + + + + III Final processing & analysis + + + IV Data management + + + + +

9 Data Transfer, Storage & Dissemination

Design + + + Operation + + +

10 SW Protocols

Operation + + + + + + Forms + + + + + +



Page 67

Annex III Summary of Distribution of Hard Copy of HPI HIS Manual Groundwater

Volume Manual Part Training State

DSC State DPC

Div DPC

Sub-Div DPC

Station/ Lab

1 HIS Design + + + + + Field I Job description + + + + +

II ToR for HDUG + + + + III Data needs assessment + + + +

Reference + + + 2 Sampling Principles

Design + + + + + Reference + + +

3 Hydro-meteorology


selection + + + + +

II SRG operation & maintenance

+ + + + + 1 SRG

III ARG/TBR/SRG operation & maintenance

+ + + + + 1 ARG

IV FCS operation & maintenance

+ + + + + 1 FCS

V Field inspections, audits, maintenance & calibration

+ + + + +

Reference + + + 4 Geo-Hydrology


selection + + + + +

II Drilling litho-specific piezometers

+ + + + + +

III Aquifer tests + + + + + + IV DWLR testing + + + + + + V Reduced levels of wells + + + + + + VI Manual water level collection

+ + + + + +

VII DWLR water level collection

+ + + + + +

VIII Inspection & maintenance + + + + + + Reference + + +

5 GIS Operation + + + + + + 6 WQ Sampling

Design + + + + + Field + + + + + +

7 WQ Analysis Design + + + + Operation + + + +

8 Data Processing & Analysis

Operation I Data entry & data validation – water level

+ + + + +

II Data entry and primary validation – water quality

+ + + + + +

III Data processing & analysis + + + + + IV GW resource assessment + + + + + IV GW yearbook + + + + +

9 Data Transfer, Storage & Dissemination

Design + + + Operation + + +

10 HIS Activities - GW Domain

Operation + + + + + +

Documents

Managing hydrometric data: from ... - The Hydrology Projecthydrology-project.gov.in/PDF/MET_Handbook_180514.pdf · This Hydrology Project Phase II (HPII) Handbook provides guidance