Assessment of the impacts of climate change and …randd.defra.gov.uk/Document.aspx?Document=11705_DefraFFG...2013/12/05 · Defra’s Sustainable, Secure and Healthy Food Supply

Assessment of the impacts of climate change and changes in land use on future water requirement and availability for farming, and opportunities for adaptation

R&D Technical Report FFG1129/TR Produced: December 2013

www.defra.gov.uk

Defra’s Sustainable, Secure and Healthy Food Supply Research Programme

Assessment of the impacts of climate change and changes in land use on future water requirement and availability for farming, and opportunities for adaptation

R&D Technical Report FFG1129/TR Produced: December 2013 Author(s): Jerry Knox, Andre Daccache and Keith

Weatherhead (Cranfield University) Simon Groves and Adele Hulin (ADAS Ltd)

Statement of use This report is part of Defra’s Sustainable, Secure and Healthy Food Supply Research Programme. The project is design to explore a range of future scenarios in potential crop distribution and associated water use and abstraction requirements in England and Wales taking into account projected changes in climate and commodity prices. Dissemination status Internal: Released Internally External: Released to Public Domain Keywords: Adapting to Climate Change, Water resources, abstraction, regulation Research contractor: Cranfield University, ADAS Ltd Defra project officer: Henry Leveson-Gower Publishing organisation Department for Environment, Food and Rural Affairs Defra’s Sustainable, Secure and Healthy Food Supply Nobel House, 17 Smith Square London SW1P 3JR

© Crown copyright (Defra); 2013 Copyright in the typographical arrangement and design rests with the Crown. This publication (excluding the logo) may be reproduced free of charge in any format or medium provided that it is reproduced accurately and not used in a misleading context. The material must be acknowledged as Crown copyright with the title and source of the publication specified. The views expressed in this document are not necessarily those of Defra. Its officers, servants or agents accept no liability whatsoever for any loss or damage arising from the interpretation or use of the information, or reliance on views contained herein.

v1: Dec5th 2007

Assessment of the impacts of climate change and changes in land use on future water requirement and availability for farming, and opportunities for adaptation (FFG1129): (Phase I) Final Report

Jerry Knox, Andre Daccache, Keith Weatherhead (Cranfield University)

Simon Groves and Adele Hulin (ADAS Ltd)

15 November 2013

Cranfield University FFG1129 (Phase I)

Table of Contents EXECUTIVE SUMMARY ...................................................................................................... 4

1 INTRODUCTION........................................................................................................ 10

1.1 STUDY AIM AND OBJECTIVES .............................................................................................. 10 1.2 WATER RESOURCES FOR AGRICULTURAL IRRIGATION .............................................................. 11

2 CURRENT AGRICULTURAL WATER DEMAND ............................................................. 13

2.1 IRRIGATED CROPPING ....................................................................................................... 13 2.2 LIVESTOCK WATER DEMAND .............................................................................................. 17 2.3 TOTAL WATER DEMAND.................................................................................................... 18

3 FUTURE AGRICULTURAL WATER DEMAND ................................................................ 22

3.1 IRRIGATED CROPPING ....................................................................................................... 22 3.2 MEDIUM-TERM IRRIGATION DEMAND FORECASTS (2030S) ..................................................... 23 3.3 LONG-TERM IRRIGATION DEMAND FORECASTS (2050S) .......................................................... 28 3.4 FUTURE LIVESTOCK WATER DEMAND ................................................................................... 43

4 REFERENCES ............................................................................................................. 44

5 APPENDIX ................................................................................................................ 46

5.1 DEFINING A BASELINE FOR DEMAND ASSESSMENT .................................................................. 46 5.2 CROP MODELLING ........................................................................................................... 47

1


List of Figures Figure 1 Reported volumes of water applied for irrigation, by crop category in England, 1987 to

2010. ...................................................................................................................................... 11

Figure 2 Spatial distribution of agricultural spray irrigation abstraction licence points in England and Wales (a) and AHDB HDC reported horticultural holdings for relative to resource availability (b). ........................................................................................................................ 12

Figure 3 Proportion (%) of agricultural and horticultural holdings located in catchments with defined levels of water resource availability. Based on EA and AHDB data for England and Wales in 2008. ................................................................................................................. 13

Figure 4 Location of irrigation abstraction ‘hotspots’ in England and Wales, based on 2010 land use. ......................................................................................................................................... 16

Figure 5 Total volumetric water demand (Mm3) for agriculture, by sub-sector, by EA Region. ........... 19

Figure 6 Total volumetric water demand (m3) for irrigated cropping (a) and livestock (b) in England and Wales, based on 2010 land use. ........................................................................ 20

Figure 7 Total agricultural water demand in England and Wales, based on 2010 land use. ................ 21

Figure 8 Defining a baseline for demand forecasting using PSMDmax. .................................................. 22

Figure 9 Reported licensed and abstracted volumes (Ml) for spray irrigation, 1974 to 2010. ............. 24

Figure 10 Volumes abstracted (Ml) between 1990 and 2010, with fitted curves showing abstraction allowing for actual weather and the underlying dry year trend. ....................... 24

Figure 11 Comparison of the reported total volumes abstracted/applied annually between 1974 and 2010 based on EA NALD and Defra Irrigation Survey datasets....................................... 25

Figure 12 Cropped area (x 000 ha) trends, by crop category (potatoes, sugar beet, cereals, vegetables, soft and orchard fruit) from 1973 to 2010 where data available. ...................... 27

Figure 13 Linear regression correlation between average agroclimate (PSMDmax) for the region and the EA reported volume of water abstracted (m3) for irrigation, by region. .................. 46

Figure 14 Spatial variability in agroclimate across England and Wales, using PSMDmax as an aridity indicator; based on the long term average (1961-1990) (a), and data generated for 2010 (b). .......................................................................................................................................... 47

Figure 15 Correlation between Irriguide modelled design dry year irrigation needs (IN) and agroclimate (PSMDmax) for maincrop potatoes grown on three contrasting soil types (LMS, MSL, LP). ....................................................................................................................... 50

Figure 16 Total ‘design’ dry year volumetric irrigation water demand (m3) by crop category in England and Wales. ................................................................................................................ 51

Figure 17 Total annual average volumetric water demand (m3) for livestock, by sub-sector in England and Wales. ................................................................................................................ 54

2


List of Tables Table 1 Volumetric irrigation water demand (× 000 m3) by crop category, by EA Region, based on

2010 pattern of land use. ....................................................................................................... 14

Table 2 Comparison of irrigation volumes (× 000 m3) for a ‘design’ dry year with Defra Irrigation Survey (2010) data and EA abstraction data, by EA Region. * Excludes ‘other’ crops and orchards. ................................................................................................................................ 14

Table 3 Water requirements used for the baseline livestock analysis. ................................................. 17

Table 4 Summary volumetric livestock demand (× 000 m3), by sub-sector, by EA Region. .................. 18

Table 5 Regional distribution of medium-term trends in irrigation demand from 2010 to the 2030s, by EA Region. .............................................................................................................. 28

Table 6 Summary descriptors for the four EA defined socio economic scenarios. ............................... 29

Table 7 Expected direction and magnitude of change for selected micro-components of demand in the “Innovation” socio economic scenario. ....................................................................... 34

Table 8 Expected direction and magnitude of change for selected micro-components of demand in the “Uncontrolled demand” socio economic scenario. ..................................................... 36

Table 9 Expected direction and magnitude of change for selected micro-components of demand in the “Local resilience” socio economic scenario. ................................................................ 37

Table 10 Expected direction and magnitude of change for selected micro-components of demand in the “Sustainable behaviour” socio economic scenario. .................................................... 38

Table 11 Reported scores for each MCD, based on the EA scenario workshop. .................................. 39

Table 12 Estimated change factors (%) for each MCD based on the workshop outputs. ..................... 39

Table 13 Projected rates of change (%) in water use in arable cropping, by EA scenario, for the 2050s. ..................................................................................................................................... 40

Table 14 Overall change in water use (growth factors) for the 2050s, by socio economic scenario and crop category. ................................................................................................................. 40

Table 15 Projected total irrigation volumes (Ml year-1) for the 2050s, by socio economic scenario. .. 40

Table 16 Sensitivity analysis showing change (%) in total projected water use for a “one arrow” reduction in the rate of change identified by the workshop participants. ............................ 42

Table 17 Summary of total future volumetric demand (m3), by livestock sub-sector for the 2030s and 205s, low and high emissions scenario. .......................................................................... 43

Table 18 Crop characteristics used for Irriguide modelling (after Knox et al., 1997). .......................... 48

Table 19 Irrigation schedule for each crop by soil type determined by Available Water Capacity (AWC); mm water applied at mm soil moisture deficit (after Knox et al., 1997). ................. 49

Table 20 Soil types and characteristics ................................................................................................. 49

Table 21 Linear regression equations derived for each crop category on three different soil types. .. 49

3

Cranfield University

Executive Summary Context UK agriculture accounts for about 75% of the total land area (Angus et al., 2009), including both cropping (arable, horticulture) and livestock (beef, dairying, pigs, poultry). It is strategically important for food provision, providing over half of all food consumed in the UK (Defra, 2010). Although most crops in England and Wales are rainfed, supplemental irrigation could become more important and widespread if adequate water were available, not only for high value potato and vegetable crops, but also on cereals where germination and crop development can be affected by extreme temperatures and drought stress (Richter and Semenov, 2005). A changing climate will also increase production risks (Knox et al., 2012) and lead to greater uncertainty in future water demands for agriculture, from both irrigated cropping and livestock production. Since over half of all irrigated production is currently cropped in catchments defined as being ‘over-abstracted’ or ‘over-licensed’ (Hess et al., 2011), there are understandably concerns regarding the environmental impact that any future increases in water demand might have on water resources. Climate change will exacerbate the situation and is expected to make direct summer abstraction less reliable (Arnell et al., 2012).

This report provides a summary of the Phase I work undertaken as part of a larger project to assess the impacts of climate and land use change on water demands for agriculture and opportunities for adaptation (FFG1129). The objectives of this study (Phase I) were to (i) model and map current water demand (irrigated cropping and livestock) and (ii) assess the impacts of climate and socio-economic change on future water demand (2030s and 2050s). The outputs from Phase II which included reviewing a range of adaptations in agriculture and their technical feasibility, and development of a visualisation web-tool for farmers and stakeholders, will be reported on separately.

Current agricultural water demand Irrigated cropping

The demand for irrigation varies significantly from year to year depending on the weather and summer rainfall. Assessments of ‘theoretical’ demand are thus usually based on a statistically defined ‘design’ dry year equivalent to the 5th driest year in 20. This criterion matches the EA approach for setting ‘reasonable’ irrigation needs for abstraction licensing. In this study, the net volumetric demand in a ‘design’ dry year in England and Wales was calculated and mapped using a GIS procedure developed by Knox et al. (1996; 1997) but updated to account for more recent information on land use, climate and irrigation practices. Maps showing the spatial distribution of demand, by crop category, were produced. By combining information on abstractions with water resource stress, irrigation abstraction ‘hotspots’ were identified.

Total ‘theoretical’ dry year irrigation demand based on land use and irrigation practices in 2010 was estimated to be 85 Mm3. Over half (60%) is concentrated on potatoes, with a further fifth (20%) on field vegetables. Most is in EA Anglian Region (60%) although EA Midlands and Southern Regions are also important, accounting for a further 20% of demand. The split between individual crop categories and regions has remained broadly similar to previous assessments although there is evidence of a declining trend in potato irrigation. Improvements in management (better scheduling) coupled with rising costs of irrigation mean that potato irrigation seems to be consolidating across a smaller cropped area. Conversely, there is still strong growth in field-vegetable irrigation which is not surprising given the broad range of high value crops within this sector.

This total theoretical ‘dry’ year demand (85 Mm3) (i.e. ‘modelled’) can be compared against two ‘observed’ datasets for 2010 – the Defra Irrigation Survey (70 Mm3) and data from the Defra ABSTAT database (104 Mm3). The differences between the two ‘observed’ irrigation demand datasets reflect

4


different data collection approaches, inherent survey limitations (e.g. dealing with non-responses), and definition and inclusion/exclusion of different types of irrigation abstraction (spray and trickle). The theoretical (‘modelled’) demand reflects unconstrained demand in a dry year assuming present cropping and irrigation practices. In contrast, actual abstractions are usually less, reflecting sub-optimal levels of irrigation management (e.g. not all irrigators use a scientific or objective approach to scheduling) and implicitly including constraints in resource availability during the growing season (e.g. some growers might experience abstraction licence restrictions during low flows). In this study, the theoretical ‘dry’ year demand (85 Mm3) for 2010 was used as the reference or ‘baseline’ against which future changes in unconstrained demand under contrasting socio-economic and climatic scenarios could be compared.

Livestock

Total water demand for livestock was estimated to be 147 Mm3. This was derived from Defra data for 2000, 2004, 2009 and 2010. Most (42%) is used in the dairying sector, with sheep (24%) and beef (20%) also being important. Maps showing the spatial distribution of water demand, by sub-sector, were produced. At a regional level, EA Wales, South West and Midlands represent the main regions where water demand is greatest.

Total demand

This study provided the most updated spatial assessment of water demand in agriculture. Combining the estimates for irrigated cropping with livestock gives a total water demand for outdoor agriculture of c241 Mm3. This can be compared against data from King et al (2005) who estimated total water use to be c247 Mm3. If previous estimates for protected and nursery cropping (c53 Mm3) are included then the total agricultural demand is around 284 Mm3.

Future agricultural water demand Irrigated cropping

Various approaches have been developed to simulate future irrigation demand (e.g. Weatherhead et al., 1994; Weatherhead and Knox, 1999; Downing et al., 2003). In this study, work by Weatherhead and Knox (2008) which was used to inform the EA water resources strategy (EA, 2008) was updated. For the longer-term forecasting, the focus was on demand sensitivity to different socio-economic scenarios, rather than critique the methodologies used for forecasting per se. Of relevance also is the impact that climate change might have on land suitability as this directly influences the future viability of rainfed cropping, which is important in areas where supplemental irrigation may not currently be significant. Two time-scales were used (i) a medium-term projection to the 2030s for a ‘business as usual’ type scenario based on underlying trends; and (ii) a longer-term projection, representing changes out to the 2050s using data from a stakeholder workshop framed within the latest EA socio-economic scenario.

The former is based on extrapolation from current trends and modelling, the latter is fundamentally founded on expert opinion within the industry, and hence the two do not necessarily agree.

(i) Current underlying trends

Current underlying trends, particularly for important irrigated crops, highlight quite major differences to those derived previously. They suggest that the strong upward growth until around 1990 in the volumes abstracted has halted and reversed. This at least partly reflects the increasing yield and hence decreasing cropped area of potatoes and some other major irrigated crops, together with increased efficiency and better scheduling. The increasing restrictions on water availability, and hence a greater appreciation of its value, are also likely to have contributed. The most complete national dataset on volumes abstracted for spray irrigation (SI) is the EA National Abstraction

5


Licensing Database (NALD) which informs the Defra ABSTAT data. Until around 1998, the volumes licensed for SI were growing steadily at around 10,000 Ml/year, equivalent to c3% of the 1998 value per annum. Since the late 1990s, however, the total licensed volume has declined slowly at c0.3% per annum. This date roughly coincides with many catchments becoming over-abstracted and changes in licensing policy. This steady decline in the national total does disguise local and regional variation and significant ‘churn’ with some new licences still being issued whilst others were being systematically reduced and/or revoked.

There is much more variation in the reported volumes abstracted, partly due to weather differences between years. Again, there has been a period of strong growth followed by a decline, though the change in the abstraction trend appears to have occurred earlier than the change in the licensed volume trend. Since around 1990, the volumes abstracted appear to have been declining at an average rate of 2% to 3% (of the 2010 value) per annum. However, this partly reflects a recent sequence of wetter summers. After allowing for annual weather variations, the underlying decline in dry year demand was estimated to be -1.4% per annum, averaged over the 1990 to 2010 period. Extrapolating forward as a compound decline would suggest a further reduction of around one quarter (-25%) to 2030; however, population growth, socio-economic change and climate change are likely to alter that.

To obtain trends for different crop categories, an analysis based on multiple linear regression using data from 9 Defra Irrigation Surveys was undertaken. This shows similar trends to the EA NALD data, but they do not match entirely. After consultation with statisticians, the ABSTAT trend was used for the total volume abstracted and the Defra Irrigation Survey used to distribute that overall trend between individual crop categories. The analysis of Defra Irrigation Survey data between 1990 and 2010 suggests the average depth applied (volume per unit area) has been growing slowly. There has been an underlying decline in both the area of potatoes irrigated and volumes applied. The picture is statistically less clear for vegetables, but tends to suggest an overall slow increase in both areas and volumes, but possibly a short-term decline more recently. In contrast, irrigation of sugar beet, grass, cereals and orchard fruit all appear to have been in longer-term decline, but have perhaps seen recent increases. Irrigation of soft fruit (e.g. strawberries) shows a steady decline in both areas and volumes, though this is one crop where depths applied have significantly increased.

In summary, these trends, particularly for the important irrigated crops, highlight quite major differences to those derived previously. Together they suggest that the strong upward growth until around 1990 in the volumes of water abstracted has halted and is now reversed. This at least partly reflects the increasing yield and hence decreasing cropped area of potatoes and some other major irrigated crops, together with increased efficiency and better scheduling, largely in response to rising energy costs for irrigation and demands for quality assurance. Emerging issues regarding water availability and reliability, and hence greater appreciation of its value, are also likely to have contributed to the change in trend. These forecasts do not include the impacts of climate change (other than any already occurring) on crop yield. It is likely therefore that they are a short-term downward trend which will level out before increasing in the long-term (see below). Furthermore, they do not yet reflect very recent changes due to higher food prices, particularly for cereals.

(i) Medium term forecasts (2030s)

Projecting current trends forwards, the Irrigrowth model outputs projects a continuing decline in the dry year demand for water for most crop categories. By 2030, total dry year demand has declined by -25% to around 71,000 ML. Potatoes remain the dominant irrigated crop (54%), but vegetables by then account for one third of the water use. There are also changes in the distribution regionally; Anglian region still uses 61% of the water, but the decline is fastest in Thames and slowest in the North West, North East and Welsh regions. It is cautioned however that these short-term projections do not include responses to the very recent increases in world food prices (particularly for cereals),

6


climate change impacts, or any relocation of crops due to increasing climate change or water resource pressures.

In the longer-term however, including climate change and population growth, demand is expected to increase under all the scenarios below. Combined with a probable decline in low-flow (summer) water availability, this indicates major future water resource issues. The point at which the current underlying decline meets the longer-term increase trend is not defined, and will likely depend on a host of factors, including actual weather variation. The projections also ignore impacts of any step-change genetic improvements, such as the introduction of genetic modification techniques and the effects of elevated CO2 concentrations on crop growth.

(ii) Long term forecasts (2050s)

Demand forecasts are highly sensitive to the prevailing socio-economic conditions. In this study, four EA scenarios were used (‘innovation’, ‘uncontrolled demand’, ‘sustainable behaviour’ and ‘local resilience’). These reflect updated scenarios developed for the EA water resource strategy (Burdett et al., 2006). In England where irrigation is supplemental to rainfall, many crops are not irrigated, and even for irrigated crops, not all farmers irrigate. Furthermore, many farmers apply less than agronomic demand, because of equipment and/or water resource constraints, or as a deliberate policy to maximise profit. The methodology for estimating future changes in demand is thus much more complex than for say an arid area, where demand is a more a function of agroclimate.

Future demand therefore depends on the area of each crop grown, the proportion of each crop irrigated and depth of water applied. Each of these in turn depends on agro-economic and technical conditions which will inevitably change, as well as the fundamental agronomic and agroclimatic conditions, which will themselves vary. Hence, the approach developed here integrated information on changes in climate with likely changes in socio- economic conditions and agro-economic policy. These ‘externalities’ are important as they influence the extent to which adaptation measures become economically viable, and geographically where they should be promoted/implemented.

Key informants at a stakeholder workshop discussed the EA scenarios and impacts on future water demand. The outputs then fed into a demand forecasting model. Participants were asked to prioritise a set of ‘drivers for change’, such as water use and availability, price and availability of resources (including energy and land), price and availability of staple crops, land use and productivity, global demand for food products, global food markets and environmental quality and biodiversity. These provided an agricultural narrative of how different ‘drivers’ might impact on the arable, potatoes and horticulture sectors. Data on the estimates of change (direction and magnitude) for five key micro-components of demand (MCD) were then quantified by the workshop participants – these included changes in national consumption, the proportion of the crop grown in the UK, yield, the cropped area and change in proportion irrigated. The MCD outputs were expressed as ‘% change factors’ and then mathematically combined to estimate future demand. The key findings are summarised below:

• ‘Innovation’ resulted in a +157% increase in irrigation demand, abstracted within environmental constraints. A high technology and knowledge-led society with consumption in a relatively resource intensive manner led to growth in potatoes (+44%), horticulture (+34%), arable (+14%) and other crops (+8%). The high increases for potatoes and horticulture was due to higher per capita consumption and population growth, but offset by substantial yield increases. The cropped areas were grown under intensive conditions predominantly managed by agribusiness with a higher proportion irrigated than at present and scheduled for high yield and quality under a more arid climate;

• ‘Uncontrolled demand’ represents a consumption based scenario that results in the largest increase in water demand (+167%). There is no significant change in overall food demand other than to supply the population increase. Potatoes remain the largest abstractor accounting for nearly half all water use (49%); there is also major growth for vegetable irrigation (27%) and nearly a fifth (15%) for arable irrigation to meet increased population food needs;

7


• ‘Sustainable Behaviour’ resulted in a gradual steady growth in demand, rising by around +42% by the 2050s, again concentrated on potatoes (+56%) and vegetables (+27%). This scenario featured low resource consumption and local food production using greener technologies, with lower yields and less emphasis on quality assurance. Diets are more vegetable based than meat, population growth is modest but the decline in imports and lower yields results in significantly more land needed for cultivation. There is a larger area under potato irrigation than at present, but mainly on family farm units rather than large scale agribusiness. Without demands for quality assurance, irrigation is widely used to boost yield;

• ‘Local Resilience’ resulted in a similar increase in demand to sustainable behaviour (+40%) due to population growth and climate change, as the population starts to accept lower quality and locally sourced fruit and vegetables. This scenario reflects a society more concerned about the environment than consumption. Consumers want naturally grown, sustainable food. Organic food and free range becomes more accepted. Farming becomes more extensive, non-organic fertilizers are expensive and yields decline. There is high concern for water application efficiency and low wastage, but efficiency falls in terms of productivity per unit of water – ‘crop per drop’ - due to the lower yields. Moderate population growth leads to significantly larger land areas under potatoes, horticulture and arable. The proportions irrigated decline slowly, and climate change is partly moderated by relocating crops. The increased areas and the impacts of climate change result in significant increases in water use on potatoes (+55%) and horticulture (+27%).

These forecasts are sensitive to model input values from the stakeholder workshop. For example, another workshop with different key informants could quite feasibly generate a different set of input values with consequent impacts on demand. A sensitivity analysis of the various input values used was therefore undertaken. The resulted in a change of ±5% for most of the input model variables. The most sensitive was ‘change in potato yield’ which could change demand by +20% for some scenarios (and vice versa). The forecasts were also sensitive to model inputs for cereals – current projections could be markedly different if there was a large-scale switch from rainfed to irrigated production (the current baseline irrigated area is very small so as to make any % changes unreliable). Given current water resource pressures, it is difficult to imagine a situation where sufficient water would be available, but changing economics to justify cereal irrigation could be a key driver.

The range in demand forecast reported in this study (+40% to +167%) is surprisingly similar to the previous range from the EA (2008) study (+27% to +167%) given that the forecasts have been completely reworked with new independent industry opinion on the micro-component changes, and have modified socio-economic scenarios.

Livestock

Assessing future livestock water demand is complex and a comprehensive assessment taking into account future livestock patterns, potential changes in diet, the moisture content of diet and water needs based on future agroclimatic conditions was beyond the scope of this project. In this study, a more simplistic approach was adopted, based on a review of existing literature coupled with published estimates of water needs for farm livestock, verified by expert opinion from animal nutrition specialists. The analysis focussed on the effect of future increases in temperature on water consumption. Projected changes in demand for the 2030s (low and high emissions) scenario were calculated by combining reported livestock temperature impacts with UKCP09 climate data on future temperature changes. The volumetric demand for each livestock sector, relative to 2010, was then derived. The forecasts result in future changes in demand of around +10% for sheep and beef cattle, <+5% for dairying and +10-20% for poultry. These equate to total increases in demand from a baseline of 147 Mm3 up to 160 Mm3. However, these are likely to under-estimate the true impact of climate change due to the many other indirect impacts which have been excluded, and provide no information on the uncertainty or sensitivity of climate or socio economic changes in demand.

8


Other factors influencing future demand The modelling approaches to assess agricultural water demand inevitably contain limitations which need to be recognised. For example, the GIS modelling is sensitive to the accuracy and resolution of input datasets (e.g. land use, soils and climate). At a regional or national scale, this is unlikely to pose a problem but more detailed catchment scale assessments would be sensitive to the relatively coarse level of spatial information on soils and land use. The demand forecasts highlighted the sensitivity in defining an appropriate ‘baseline’ for projection, and how the forecasts were also sensitive to some of the model input variables derived from the stakeholder workshop.

The forecasts are also dependent on the assumptions made regarding land use and the relative split between rainfed and irrigated cropping, and how these might change in future. For example, changes in land suitability, due to either increased aridity and/or increased flood risk, coupled with changes in agro-economic policy could have major consequences on agricultural water demand. The impacts of land use change, for example, in terms of movement of cropping in response to changing aridity or water stress, or the introduction of new crops, have not been explicitly considered. However, recent research (Daccache et al., 2012) has shown that large areas of land in England and Wales currently suited to potato production could become marginal due to increased droughtiness, and that expansion of potato cropping would be constrained by water availability. Similarly, for other important rainfed crops such as cereals, changes in land suitability could have a major impact on location of cropping and the need for supplemental irrigation. These issues regarding changes in land suitability will be considered in Phase II, set within the context of which adaptation options would then be most appropriate to mitigate the impacts.

9


1 Introduction Climate change will influence the way crops grow, develop and yield. UK grown crops will be directly sensitive to any future changes in sunshine, rainfall, temperature and CO2 concentration levels. There will also be indirect impacts on the agricultural potential of soils by modifying soil water balances, affecting moisture availability and land management practices including trafficability and workability. Reduced water availability for agricultural abstraction as a result of lower river flows will impact on supplemental irrigation, both for existing irrigated crops and new crops which may need watering to cope with increased droughtiness. Climate change may also lead to more frequent and extended periods of water logging, and more frequent and larger areas of inundation of high grade (floodplain) agricultural land, with consequences for productivity. There will inevitably be many other indirect effects – including, for example, changes in the range of native/non-native pests and crop diseases (e.g. potato blight, fusarium), increased crop damage (e.g. grapes and salads) at extreme temperatures, crop diversification and introduction of new or novel crops (e.g. maize, sunflowers). Many of these are inextricably linked and could have a positive or negative impact, or a combination, depending on the assumed future socio-economic scenario and farmer perceptions to climate risk.

For outdoor livestock and animal farming, the effects of climate change will be complex and variable. Grass production may be enhanced by increases in the length of the growing season especially in upland areas, although future soil water shortages could limit production in some years in lowland areas. A changing climate would impact on factors as diverse as livestock health, forage yields, feedstuff quality, availability and cost, water availability, thermal stress and related welfare issues, including disease spread and control measures. The main impacts are likely to relate to changes in CO2 levels (impacting on grass productivity and dietary quality), temperature (causing heat stress, influencing reproductive capacity and increasing pathogen and fly problems), soil water availability (for grassland production) and weather extremes (changing housing and supplementary feed needs).

1.1 Study aim and objectives The project aim was to assess the current and future water requirements for agriculture (cropping and livestock) and identify where it would be most cost-effective to provide supplementary water conservation and management mechanisms to meet projected increases in water. The project was deliberately split into two phases:

Phase I:

1. Model and map the current spatial distribution of agricultural water demand, including irrigated cropping and livestock; and

2. Assess the impacts of climate and socio economic change (and their sensitivities) on future agricultural water demand, in both cropping and livestock sectors.

Phase II:

1. Identify potential adaptations to water availability, particularly in water stressed regions; 2. Assess the costs and benefits of adaptation measures and impacts to farm businesses; and 3. Develop a web-based impact and adaptation tool for farmers.

This report provides a summary of the research conducted in Phase I. The assessment considers likely changes in demand out to the 2030s and 2050s, to help identify demand impacts, and to assess where there are likely to be production constraints and/or new crop opportunities. A brief introduction to water use in agriculture is present to provide some context to the drivers and pressures facing water allocations for irrigated crop production and livestock.

10


1.2 Water resources for agricultural irrigation Most agricultural cropping in England and Wales is rainfed and, even in a dry year, only a very small proportion of cropped land (<0.5%) is typically irrigated. Compared with total national freshwater abstractions, including those for public water supply, industry and electrical generation agricultural irrigation constitutes only a minor (1 to 2%) use. However, irrigation is a consumptive use (that is, the water is not returned to the environment in the short term) and is concentrated in the driest areas in the driest years and driest months when resources are most constrained (Knox et al., 2009). As a consequence, irrigation can be the largest abstractor in some catchments in dry summers. Concerns have been raised over the potential impacts on the environment, particularly in catchments where irrigation abstractions are concentrated and where water resources are under pressure. In many catchments, summer water resources are already over-committed and additional summer licences for surface and groundwater irrigation abstraction are unobtainable. Furthermore, existing summer sources are increasingly unreliable.

Since 1955, the government has published statistics on agricultural irrigation in England and Wales, based on surveys carried out roughly triennially, and most recently in 2010. These provide statistics on the areas irrigated, volumes applied and water sources used for irrigation. Data from the last 7 Irrigation Surveys (1987-2010) are summarised in Figure 1.

Figure 1 Reported volumes of water applied for irrigation, by crop category in England, 1987 to 2010.

Over the last 25 years, there have been significant changes in the composition of crops irrigated. The proportion of irrigation on grass, sugar beet and cereals has declined steadily. In contrast, there has been a marked increase in irrigation of high value crops, particularly potatoes and field vegetables. This trend is driven by supermarket demands for quality, consistency and continuity of supply, which can only be guaranteed by irrigation. The total area of potatoes grown in England fell by almost 19% between 2001 and 2005 as yields increased. Nevertheless, potatoes continue to be the dominant irrigated crop, accounting for 43% of the total irrigated area and 54% of water use. Irrigated vegetables have increased slowly in relative share to 25% of the area and 26% of water use. Cereals show an increase in irrigated area, but much less change in water use.

11


Information on the spatial distribution of agricultural and horticultural holdings across England and Wales are collected annually by the agricultural levy board (Agriculture and Horticulture Development Board, AHDB) as part of their statutory duty. The type of information and level of detail available in the public domain depends on its commercial sensitivity, but baseline data can be used to map the spatial distribution of growers. For this study, those involved in the production of the three most important irrigated crop categories have been considered, namely potatoes, field vegetables and soft fruit. Collectively, these account for 83% of the total volume of irrigation water abstracted annually. In England and Wales, the water regulatory authority, the Environment Agency (EA), has assessed the availability of water resources for abstraction at a catchment level. Each catchment has been defined according to its resource status and allocated to one of four categories, ‘water available’, ‘no water available’, ‘over-licensed’ and ‘over-abstracted’, in order of increasing water stress. The spatial distribution of agricultural holdings involved in potato, field vegetable and soft fruit production in 2008 has been mapped, and then compared with water resource availability, by catchment, using a GIS (Figure 2). The aggregated data by crop sector and water resource category are summarised in Figure 3.

Figure 2 Spatial distribution of agricultural spray irrigation abstraction licence points in England and Wales (a) and AHDB HDC reported horticultural holdings for relative to resource availability (b).

The analysis shows that on average only 10-15% of agricultural holdings are located in catchments where additional water abstraction would be available during summer low-flow periods (“water available”). About half of all holdings are located in catchments defined as either having ‘no (more) water available’ or already ‘over- licensed’. Nearly a fifth are in catchments defined as already being ‘over- abstracted’, and where abstraction may actually have to be reduced, at least at times of low flow. Hence there are clearly major constraints on any increase in irrigation water use. Conversely, reducing irrigation abstraction would release water resources to sustain environmental flows or support other uses.

12


Figure 3 Proportion (%) of agricultural and horticultural holdings located in catchments with defined levels of water resource availability. Based on EA and AHDB data for England and Wales in 2008.

2 Current agricultural water demand This section briefly describes the approaches used to model and map current water demand for irrigated cropping and livestock in England and Wales.

2.1 Irrigated cropping To model and map the current spatial distribution of agricultural water demand across England and Wales an updated procedure developed by Knox et al (1996; 1997) was used. In summary, the net volumetric (m3) irrigation water requirements for the main crop categories currently irrigated in England and Wales were calculated and mapped using a geographic information system (GIS). This involved simulating the annual irrigation needs (mm) for the major irrigated crop categories (potatoes, field vegetables, sugar beet, soft fruit, orchard fruit, grass, and cereals), grown on a contrasting range of soil types at agroclimatically contrasting locations using a daily time-step water balance model. For this, the Irriguide model was used, an irrigation scheduling water balance model developed by ADAS and Met Office (Bailey and Spackman, 1996; Silgram et al., 2007). The Irriguide results were then correlated using linear regression analyses to existing national spatial datasets relating to climate (rainfall and evapotranspiration), land use, soils and irrigation practices to generate volumetric (m3) irrigation water requirement maps (for each crop category, and in total) at approximately 2 km2 resolution (the lowest resolution for the digital datasets).

A detailed description of the methodology adopted is given in Knox et al. (1996; 1997). In this study, that procedure was updated using soil datasets from the National Soil Resources Institute (NSRI), data from the UKCP09 climatology to generate an agroclimate baseline, catchment boundary and abstraction data from the Environment Agency, irrigation data from the 2010 Defra Irrigation Survey, and rasterised land use datasets from the Defra cropping census produced by the Edinburgh University Data Library (EDINA).

13


The derived maps show the current ‘theoretical’ irrigation demand, by crop type and in total varies spatially across in England and Wales. The total volumetric water demand for outdoor irrigated crops in England and Wales was estimated to be 85 Mm3 based on the 2010 cropping pattern (Table 1). The irrigation demands by crop category have been mapped and summarised in the Appendix (Figure 16).

Table 1 Volumetric irrigation water demand (× 000 m3) by crop category, by EA Region, based on 2010 pattern of land use.

Crop category Anglian EA

Wales Midlands North East

North West

South West Southern Thames Total

Earies 1288 140 153 64 52 15 54 41 1807 Maincrop 34647 3053 6998 6570 1247 434 1633 1009 55591 Cereals 948 22 120 29 2 1 7 81 1208 Sugar beet 9803 4 393 198 0 0 0 28 10426 Vegetables 3300 203 513 177 77 79 698 122 5169 Soft fruit 247 48 83 38 0 78 945 35 1474 Orchards* 0 0 0 0 0 0 0 0 1 Grass 5406 28 130 22 4 357 1461 1692 9100

Total 55640 3498 8389 7097 1382 964 4798 3008 84776

* Estimated The analysis shows that over half (60%) the theoretical demand is concentrated on potatoes, with a further fifth (20%) on field vegetables. Most irrigation demand is concentrated in EA Anglian Region (60%); EA Midlands and North East are also important, accounting for a further 20% of total demand. The analysis excludes ‘other crops grown in the open’ which would typically add another 5% to the overall total. The analysis also excludes protected cropping and irrigation for ornamental or nursery stock. Previous estimates using scheduling models and large agroclimatic areas were 222 Mm3 (Weatherhead et al., 1994) and 140 Mm3 (Knox et al., 1997). The demand derived in this study can also be compared against the Defra reported irrigation water use for England and Wales in 2010 of 70 Mm3 (Defra, 2011) and the EA reported direct abstractions for spray irrigation of 104 Mm3 (EA, 2012) (Table 2). It is noted, however, that the scope of the various datasets are all slightly different and all figures contain inaccuracies. In 2010, there may have also been abstraction restrictions in force which supressed the volumes of actual abstraction reported in the EA NALD data.

Table 2 Comparison of irrigation volumes (× 000 m3) for a ‘design’ dry year with Defra Irrigation Survey (2010) data and EA abstraction data, by EA Region. * Excludes ‘other’ crops and orchards.

EA Region ‘Design’ dry year based on 2010 land use (this study)

Defra Irrigation Survey (2010)*

EA volumes abstracted spray irrigation (2010)

Anglian 55640 48286 63875 EA Wales 3498 2272 1825 Midlands 8389 7467 17885 North East 7097 4784 6570 North West 1382 1015 1825 South West 964 706 1460 Southern 4798 3914 6570 Thames 3008 1621 4015 Total 84776 70065 104025

14


These datasets provide the reference ‘baseline’ from which the impacts of climate and socio-economic change were then investigated (Section 3). Outputs were produced for each crop type and in total, by EA CAMS sub catchment, for England and Wales. The crop datasets used and more detail on the methodology are given in the Appendix.

2.1.1 Irrigation ‘hotspots’ The demand maps and datasets were then combined with historical records of spray irrigation abstraction from the EA National Abstraction Licensing Database (NALD) and with the latest estimates of EA water resource availability (at sub-catchment level) to identify irrigation abstraction ‘hotspots’. These were defined as being areas where high ‘theoretical’ and ‘actual’ irrigation demand coincided with sub-catchments which were most constrained (over-abstracted and/or over-licensed) in terms of resource availability (Knox et al., 2009).

Figure 4 highlights the spatial distribution of irrigation abstraction ‘hotspots’ and summary data for each CAMS catchment where irrigation pressures are most likely to emerge.

In a parallel study (Weatherhead et al., 2013) investigating opportunities for water resource collaboration and uptake of on-farm storage reservoirs, these ‘hotspots have been correlated against the location of abstraction licenses for irrigation storage reservoirs.

15


Figure 4 Location of irrigation abstraction ‘hotspots’ in England and Wales, based on 2010 land use.

CAMS WRMU No

licences Licensed

volume (%) Abstracted volume (%)

Arun & Western Streams 280 1.71 2.41 Broadland Rivers 357 4.38 5.96 Cam & Ely Ouse (incl S.Level) 562 16.77 25.08 Derwent 33 0.23 0.45 East Suffolk 239 4.15 7.40 Idle & Torne 337 6.06 5.41 Louth Grimsby and Ancholme 51 0.41 0.61 Lower Trent & Erewash 287 3.33 2.57 Medway 36 0.97 0.50 Nene 50 0.47 0.43 North Essex 617 5.87 4.52 North Norfolk 55 1.08 2.05 North West Norfolk 41 0.97 1.32 Old Bedford incl Middle Level 379 4.16 3.87 Rother 20 0.18 0.42 Shropshire Middle Severn 116 1.42 1.04 South Essex 7 0.22 0.48 Stour 158 0.95 0.67 Tame Anker and Mease 79 0.64 0.52 Tees 122 0.75 0.88 Warwickshire Avon 82 0.62 0.63 Welland 71 0.43 0.45 Wharfe and Lower Ouse 64 0.76 0.97 Witham 198 1.94 1.88 Worcs Middle Severn 170 1.45 1.79 Wye 452 3.54 2.75 Total 4863 63.48 75.04

16


2.2 Livestock water demand The aim of this part of the study was to model and map livestock water demand for the baseline (2010) across England and Wales. The process involved agricultural census data being coupled with estimates of livestock drinking water demand by category, based on those recently derived for Defra projects WU0101 and WU0132. The estimates of livestock water demand take into account the age and size of animals, the composition of their diets, production levels and ambient temperatures all of which are known to influence water requirements. By combining spatial datasets on livestock population with the estimated water requirement, GIS outputs were produced showing the spatial variation in livestock water demand, expressed as average volumetric (m3) water demand for each of the main livestock sectors (beef, dairy, pigs, poultry, sheep).

2.2.1 Methodology Livestock numbers from the Defra Census figures for the years 2000, 2004, 2009 and 2010 were aggregated into the ADAS 1 km grid square livestock database. The 1km dataset is derived using established peer reviewed methodologies that combine pycnophylactic interpolation with dasymetric mapping to generate statistical estimates of the area of crop and number of livestock head per km² (Comber et al., 2008). In summary, the approach begins with holding level agricultural census data, aggregates it to parish group and then distributes it back to the 1 km² cell based on known land uses from third party datasets. The resulting data is commonly recognised as the best resolution agricultural census data in England and Wales. Cattle categories in the census data changed between survey years. This is not uncommon as livestock management practices change or size categories alter, however in order to ensure a common set for the purposes of comparison, it is necessary to derive harmonised groupings that provide as close as possible inter year comparability. All categories were assessed by ADAS experts and a common set was derived. Specifically, this meant assuming that for years 2000 and 2004, the categories of K1 (all dairy cows and heifers that have calved) and K5 (other females (1-2 years) intended for dairy herd replacement) were the dairy herd, and for 2009 it was K203 (female dairy cattle < 1 year old), K206 (female dairy cattle 1-2 year old), K209 (female dairy cattle > 2 year old (not calved yet)) and K 211 (female dairy cattle > 2 year old (have calved)).

The census data for livestock numbers were multiplied by estimates of water demand per head of livestock (Table 3) to give a total water demand per 1 km grid square for each survey year and livestock category. The water demand was based on drinking water requirement and wash water requirement based on those derived for Defra WU0101 and WU0132 (University of Warwick and ADAS, 2006; ADAS, 2012). These projects developed estimates of livestock water demand by taking into account the age and size of animals, the composition of their diets, production levels and ambient temperatures all of which are known to influence water requirements.

The water demand by livestock type (beef, dairy, pigs, poultry and sheep) was calculated by summing the water requirement in each livestock category for each year. The baseline water demand by livestock type and category was then calculated as the mean of individual data generated for the four survey years (2000, 2004, 2009 and 2010). These data were then used to create the 2 km resolution spatial datasets for England and Wales.

Table 3 Water requirements used for the baseline livestock analysis.

Livestock type Livestock category

Cycle duration

(days)

Drinking water per head per

day (l)

Wash water per head per

year (l) Cattle Dairy cow herd 365 90.61 29 Beef cows & heifers 365 20 0 Dairy & beef bulls 365 20 0

17


Cattle <1yr 365 12.5 0 Poultry Broilers 133 0.09 1.14 Ducks, geese & other birds 56 0.2 2.71 Turkeys 406 0.2 0.24 Pullets 406 0.22 0.47 Laying hens - caged 322 0.19 0.94 Laying hens - non caged 63 1.22 4.13 Broiler breeders, layer breeders, cocks 140 0.58 4.37 Pigs Sows 365 13.73 453.22 Maiden gilts 365 5.5 0 Barren sows 365 5.5 0 Weaners (20kg) 365 1.8 104.39 Growers (50kg) 365 4.2 135.42 Finishers 365 5.6 0 Boars 365 10 0 Sheep Ewes 365 4.56 0.75 Lambs 365 2.65 0.75 Rams and other adult sheep 365 3.3 0.75

The total volumetric water demand (m3) for livestock has been modelled and mapped. The outputs are summarised by category in Table 4.

Table 4 Summary volumetric livestock demand (× 000 m3), by sub-sector, by EA Region.

Sub sector Anglian

EA Wales Midlands

North East

North West

South West Southern Thames Total

Sheep 1499 12254 5474 5922 4151 3864 1220 781 35166 Poultry 4727 1537 1821 1245 992 1457 629 438 12847 Pigs 2703 150 906 2382 312 845 213 400 7911 Dairy 1981 10928 11921 4944 11348 16652 2311 1882 61967 Beef 2368 5705 5057 4317 3109 6012 1142 1194 28904

Total 13279 30574 25179 18810 19913 28829 5516 4694 146795

In this study, the current (baseline) total water demand for livestock is estimated to be 147 Mm3. This is based on averaged Defra livestock data from 2000, 2004, 2009 and 2010. Most (42%) is used in dairying, with sheep (24%) and beef (20%) also important sectors for water demand. EA Wales, South West and Midlands constitute the areas where water demand for livestock is greatest.

Defra (2011) reported that livestock water use for drinking in 2010 was 75 Mm3, similar to that reported for irrigated cropping (70 Mm3). Other sub sectors where water use is important include spraying (4 Mm3) washing down (13 Mm3) and ‘other’ agricultural uses (18 Mm3).

2.3 Total water demand A summary of the total volumetric water demand, for irrigated cropping and livestock is summarised by EA Region (Figure 5) and spatially for England and Wales (Figure 6).

Combining these estimates for irrigated cropping and livestock gives an overall water demand for agriculture of 231 Mm3. This figure can be compared against King et al (2005) who estimated total on-farm water abstraction to be in excess of 300 Mm3 per year, although included in that estimate

18


was protected and nursery cropping (53 Mm3). When this figure is included the overall figures are similar; however, the split between sectors is quite different. King et al. (2005) estimated that approximately 60% of total water demand was used for irrigation of outdoor field-scale agricultural and horticultural crops, equivalent to 128 Mm3 mainly on potatoes and field vegetables. This is consistent with previous estimates for agricultural irrigation in England and Wales by Knox et al (1997) of 140 Mm3, but clearly much higher than the latest estimate in this study of 85 Mm3.

Figure 5 Total volumetric water demand (Mm3) for agriculture, by sub-sector, by EA Region.

19


Figure 6 Total volumetric water demand (m3) for irrigated cropping (a) and livestock (b) in England and Wales, based on 2010 land use.

20


Figure 7 Total agricultural water demand in England and Wales, based on 2010 land use.

21


3 Future agricultural water demand 3.1 Irrigated cropping Over the last two decades a number of contrasting approaches have been developed to estimate future spatial and temporal changes in irrigation water demand (e.g. Weatherhead et al., 1994; Weatherhead and Knox, 1999; Downing et al., 2003). In England where irrigation is supplemental to rainfall, many crops are not irrigated, and even for irrigated crops, not all farmers irrigate. Furthermore, many farmers apply less than agronomic demand, because of equipment or water resource limitations, or as a deliberate policy to maximise profit. The methodology for estimating future changes in demand is thus more complex than for an arid area, where demand is more directly a function of agroclimate. For irrigated cropping, water demand depends on the area of each crop grown, the proportion of each crop irrigated, and depth of water applied. Each of these in turn depends on agro-economic and technical conditions which will inevitably change, as well as the fundamental agronomic and agroclimatic conditions, which will themselves vary.

Irrigation demand varies from year to year depending on summer weather. For comparability therefore, all irrigation demand forecasts relate to a ‘design dry year’. This is defined as a year where the unconstrained demand has an 80% probability of non-exceedance. This criterion is widely used in planning irrigation water resources, e.g. for abstraction licencing and reservoir sizing; the farmer would then have adequate water for 80 years in 100, but would have to under-irrigate some crops in the other 20. This level of planned headroom is considerably less than for most other sectors.

A baseline year is also required for modelling forwards. Calculations using maximum potential soil moisture deficit (PSMDmax) as a climate indicator have shown that 2010 was in fact fairly close to a ‘design’ dry year (Figure 8). The cropping patterns in 2010 have therefore been used for the baseline, with data adjusted where necessary to the dry year equivalent.

Figure 8 Defining a baseline for demand forecasting using PSMDmax.

22


It is emphasised that all the projections here are for ‘demand’; actual water use is already constrained in dry years and will be significantly reduced by any increased restrictions on water availability and/or reduced reliability, which may then lead to a relocation of irrigated cropping and hence water demand. Such relocation is not considered in these calculations.

Forecasts were made for two periods; firstly, for the short to medium term (i.e. 2010 to 2030s) for a ‘business as usual’ scenario under current economic and water policy conditions based on underlying trends, and secondly, for the medium to long-term (up to 2050s) with projections framed within four EA defined socio-economic scenarios. The former is based on extrapolation from current trends and modelling, the latter is fundamentally founded on expert opinion within the industry, and hence the two do not necessarily agree. All demand projections are relative to the 2010 baseline. The approaches used and key findings are summarised below.

3.2 Medium-term irrigation demand forecasts (2030s) For the medium-term projection, an extrapolation based on underlying trends was developed. This approach builds directly on work by Weatherhead and Knox (2008) which formed part of the EA water resources strategy (EA, 2008b). That study considered eight crop categories which matched those in the Defra Irrigation Surveys, namely; early and main crop potatoes, sugar beet, vegetables (grown in the open), soft fruit, orchard fruit, cereals, and grass. Climate impacts were not modelled.

In England, the irrigated areas and the volumes of irrigation water applied each year vary considerably depending on the summer weather, and distribution of rainfall. The data published in government irrigation surveys and reported by the EA on irrigation abstractions therefore partly reflect the weather in each year, and do not directly show the dry year demand in a particular year nor indicate the underlying trends in dry year demand. However, Weatherhead and Knox (2008) developed a statistical approach to analyse these datasets using calculated theoretical irrigation needs (depths) for selected crops as the independent climate variable in a multiple linear regression analysis. This can be used to calculate the underlying growth rate in the areas irrigated and volumes of water applied. Two complementary datasets were combined; the first is annual data from the EA National Abstraction Licensing Database (NALD), which informs Defra ABSTAT data; the second is the periodic but crop-specific data from the Defra Irrigation Surveys.

3.2.1 Trends based on NALD/ABSTAT data Data on the total volumes of water licensed and abstracted for spray irrigation are available from the EA NALD or in processed format from the ABSTAT files published by Defra. Almost all irrigators abstracting > 20m3day-1 are required to have an abstraction licence and flow meter(s) and to return data to the EA on their volumes abstracted. After statistical correction for non-returns and missing data, the aggregated results are published as ABSTAT data, available from the Defra website. Until around 1998, the volumes licensed for spray irrigation were growing steadily at around 10,000 Ml/year, equivalent to around 3% of the 1998 value per annum (Figure 9).

Since the late 1990s, however, the total licensed volume has declined slowly, at about 0.3% per annum. This date roughly coincides with many catchments becoming over-abstracted and changes in licensing policy. This steady decline in the national total does disguise local and regional variation and significant ‘churn’ with some new licences still being issued and others being systematically reduced and/or relinquished.

There is much more variation in the volumes reported as actually abstracted, partly due to the weather differences between years (Figure 10).

Again, there has been a period of strong growth followed by a decline, though the change in abstraction trend appears to have occurred earlier than the change in licensed volume trend. Since around 1990, the volume abstracted appears to have been declining at an average rate of around 2% to 3% (of the 2010 value) per annum.

23


Figure 9 Reported licensed and abstracted volumes (Ml) for spray irrigation, 1974 to 2010.

However, this partly reflects a recent sequence of wetter summers. After allowing for the annual weather variation, the underlying decline in dry year demand was estimated to be around 1.4% per annum, averaged over the 1990 to 2010 period (Figure 10). Extrapolating that forward alone as a compound decline would suggest a further reduction of around one quarter from 2010 to 2030.

Figure 10 Volumes abstracted (Ml) between 1990 and 2010, with fitted curves showing abstraction allowing for actual weather and the underlying dry year trend.

3.2.2 Crop based trends based on MAFF/Defra Irrigation survey data More detailed data on irrigation use for individual crop types is available from the periodic Defra Surveys of “Irrigation of Outdoor Crops”. Since 1982 the main questions have been kept consistent, giving now nine sets of directly comparable data, for 1982, 1984, 1987, 1990, 1992, 1995 (all by MAFF), then 2001 and 2005 (for Defra by Cranfield University), and most recently in 2010 (by Defra). The next survey is planned for 2013. The data is broken down between 8 crop categories, by irrigated

24


area and volume. Other data includes irrigation method, water source, and scheduling method. This data is much richer in content but is less complete, due to its intermittent nature and the lower return rate. The address list is obtained from a trigger question in the annual Defra "Agricultural and Horticultural Cropping" census (“June Census”). That is only a full census in selected years (including 2010). Statistical corrections therefore have to be made for those not included in the June Census, non-returns to the June Census, and non-returns to the irrigation survey. This can be a major source of error for calculating national totals. In 2001, for example, the final returns were estimated to cover only about 40% of the total irrigated area. The survey also only refers to outdoor crops grown on registered agricultural holdings; it therefore excludes glasshouse crops, and landscape and other non-crop irrigation. It does include other water sources outside the NALD/ABSTAT dataset such as mains supply, rainwater harvesting, water re-use, trickle irrigation and abstractions less than 20 m3

day-1, though these are relatively minor compared to direct abstraction from surface water or groundwater.

The total annual water use data from the Defra Irrigation Survey data shows similar trends to the EA NALD data, but they do not match entirely (Figure 11). This may be partly due to the different ranges of water sources covered and businesses surveyed, the different year end dates (affecting when reservoir refill is counted), and/or inaccurate returns from water users, but is probably mainly due to the difficulty of correcting for non-returns in both datasets. A detailed discussion of the various survey differences is given in Weatherhead et al. (1997). Analyses based on the 1995 to 2010 Irrigation Surveys data appear to show very much faster rates of decline than the ABSTAT abstraction returns; however, their statistical reliability is lower, due to the limited number of surveys. After discussions with Defra statisticians, the ABSTAT trend has therefore been accepted for the total volume abstracted, and the Defra Irrigation Survey data used to distribute that overall trend between the individual crop categories.

Figure 11 Comparison of the reported total volumes abstracted/applied annually between 1974 and 2010 based on EA NALD and Defra Irrigation Survey datasets.

The resulting analysis suggests the average depths applied (volume per unit area) have been falling slowly. There has been an underlying decline in both the area of potatoes irrigated and volumes applied. The picture is statistically less clear for vegetables, but tends to suggest an overall slow increase in both area and volumes, but possibly a short-term decline more recently. In contrast,

25


irrigation of sugar beet, grass, cereals and orchard fruit all appear to have been in longer-term decline, but have perhaps seen recent increases. Irrigation of soft fruit (e.g. strawberries) shows a steady decline in both areas and volumes, though this is one crop where depths applied have increased.

These trends, particularly for the important irrigated crops, highlight quite major differences to those derived previously. They suggest that the strong upward growth until around 1990 in the total volumes of water abstracted has reversed. This at least partly reflects the increasing yield and hence decreasing cropped areas needed (particularly for potatoes and some other major irrigated crops), together with increased efficiency and better scheduling. The increasing problems relating to reduced water availability and reliability, and hence a greater appreciation of its value, are also likely to have contributed to water conservation. It is noted however that these short term trends do not include climate change (other than any already occurring), and may therefore be temporary.

The trends in total cropped areas, for each crop category, were calculated from the Defra Agricultural Census data. Figure 12 shows the crop trends at UK level. As well as a steady underlying decline, a distinct break is apparent around 1996 for several crops, and trends have changed again recently (notably for potatoes) following the recent increases in crop prices and energy costs.

26


Figure 12 Cropped area (x 000 ha) trends, by crop category (potatoes, sugar beet, cereals, vegetables, soft and orchard fruit) from 1973 to 2010 where data available.

27


These underlying trends have then been used in the Irrigrowth model developed previously to forecast future demand (Weatherhead et al., 2008). This superimposes the national trends onto the baseline 2010 values, spatially distributed by EA region, and models annual changes in crop areas, proportions irrigated, irrigation depths, yields and efficiency.

Projecting forwards, the Irrigrowth model outputs confirm a continuing decline in the dry year demand for water for most crop categories. By 2030, total dry year demand has declined by -25% to around 71,000 ML. Potatoes remain the dominant irrigated crop (54%), but vegetables by then account for one third of the water use. There are also changes in the distribution regionally; Anglian region still uses 61% of the water, but the decline is fastest in Thames and slowest in the North West, North East and Welsh regions (Table 5). It is cautioned however that these short-term projections do not include responses to the very recent increases in world food prices (particularly for cereals), climate change impacts, or any relocation of crops due to increasing climate change or water resource pressures.

Table 5 Regional distribution of medium-term trends in irrigation demand from 2010 to the 2030s, by EA Region.

EA Region 2010 dry year demand (ML)

2030 dry year demand (ML)

Change (%)

North East 5950 4748 -20% North West 1653 1353 -18% Midlands 16196 12621 -22% Anglian 57845 43419 -25% Thames 3636 2231 -39% Southern 5950 4364 -27% South West 1322 1016 -23% Welsh 1653 1350 -18%

Total (E&W) 94204 71102 -25%

3.3 Long-term irrigation demand forecasts (2050s) The methodology adopted for longer-term irrigation demand is based on that developed with the Environment Agency previously for the EA Water Resource Strategy (EA, 2008), utilising a basic spread sheet approach at national level. However, rates of change were obtained from independent industry estimates collected using a stakeholder workshop format, rather than directly estimated by researchers. Water demand forecasts are highly sensitive to the assumed prevailing socio economic conditions. In this study, the four new EA future socio-economic scenarios were used termed ‘innovation’, ‘uncontrolled demand’, ‘sustainable behaviour’ and ‘local resilience’. These reflect updated scenarios developed previously for the EA water resource strategy (EA, 2008) termed ‘Alchemy’, ‘Jeopardy’, ‘Survivor’ and ‘Restoration’, respectively (Burdett et al., 2006). The scenarios are summarised in Table 6.

A stakeholder workshop was organised in July 2012 to discuss the four socio-economic scenarios and their impacts on irrigation demand. Several ‘givens’ were highlighted to the participants prior to discussion – these are drivers where we assume just one possible future development across all scenarios. These included (i) climate change, where it was assumed that recent developments would continue without any trend breaks (ii) population demographics, which were already defined under each socio-economic scenario, (iii) consumption patterns and behaviour, which were already part of the ‘axis of uncertainty’ so their status is already given in the EA scenario, and (iv) government interventions in EU and national agricultural policy, which were similarly already part of the ‘axis of uncertainty’ so their status is already given in the EA scenario.

28


Table 6 Summary descriptors for the four EA defined socio economic scenarios.

Innovation “Our scientists and technologists can solve the problems of environmental damage through their ideas and innovation”

Uncontrolled demand “The rich shall inherit the earth – because we’re worth it”

• Government chooses to drive the UK economy and avoid resource shortage by large-scale investment in sustainable technologies

• Strict UK regulation is seen as a driver of innovation • Closed loop systems attempt to ensure nothing is wasted by using

‘cradle to cradle’ manufacturing systems • The purpose of innovation is to create sustainability which doesn’t

inconvenience consumers or expect them to change their behaviour • GM food technology booms – with consumer support - and there is a

15% increase in crop yield from 2035 to 2050. • Water and energy grids have been upgraded, and most homes have

been upgraded with new resource management technology systems.

• Apparent national prosperity masks a polarisation in incomes. People are broadly aware of environmental issues, but because the main impacts are outside of the UK they remain of little concern.

• Resources remain cheap due to new methods of extraction and manufacture, regardless of the polluting nature of these methods.

• GDP is amongst the world’s highest, and free trade and open market trading systems continue.

• The UK is potentially self-sufficient in food production, and operates high intensity agricultural practices to feed the growing population with some new crops to be grown in the UK for the first time, e.g. soy.

• More chemicals are used in agriculture, but these are less polluting than previously. Agriculture is growth sector for the less-skilled labour force.

• Water and energy demand becomes self-regulating due to price. Local resilience “It is better to have fewer wants than greater resources” Sustainable behavior “We can cut out resource use through new ways of

managing our societies and our relationships” • Food and energy costs increase due to continued demand from

emerging economies. Increasing energy costs and a reliance on imported power – with subsequent blackouts – leads to rationing. Environmental concerns are low due to protectionist economic policies

• Consumerism and consumption are focused on services, not products, as people seek richer, local experiences that are not financed on debt

• People adapt their houses to reduce energy and water use, and there is more shared living to reduce utility costs

• Food is seasonal, local and there is more urban agriculture with allotments, school plots and other ‘unused’ land used for growing food

• Goods are built to last and be repaired, rather than replaced and there is more shared ownership of expensive items such as vehicles and tools

• There is a heightened sense of social awareness and ‘being seen to be green’

• Initially driven by regulation, consumers now make sustainable choices voluntarily

• The price of products reflects their full environmental costs – including pollution

• Consumers are happy for government to regulate for the collective good

• People choose diets which reflect the carbon used to produce the food they eat

• There is a shift to public ownership or mutualisation in the water, energy and waste industries

29


Workshop participants were first asked to prioritise a set of ‘drivers for change’ that would most influence future water demand. Those given highest priority were water use and availability; price and availability of resources including energy and land; price and availability of staple crops; land use and productivity; global demand for food products; global food markets; and environmental quality and biodiversity.

Two exercises were then conducted. The first was to develop a qualitative description of how these ‘drivers of change’ might impact on agriculture within each socio-economic scenario. This helped to provide a consistent agricultural narrative for each scenario (Boxes 1 to 4).

Boxes 1-4 Qualitative impacts of the prioritised drivers for change on agriculture under each socio-economic scenario

Innovation Major drivers under this scenario

Scientific and technological advances drive innovation in sustainable technologies, with strong government support and regulation

Environmental quality and biodiversity • The UK will still demand high levels of biodiversity and landscape • Will keep grazing on untilled land – its landscape value will be recognised • Increased support for ecosystem services – better valuation of ESS • Higher risk from disease will require more technology and regulation – this relies on higher levels of

surveillance and response. • There will be a mix of factory (for bulk crops such as cereals) and small high tech farms (for e.g.

horticulture)

Water use and availability • Water consumption per unit production decreased • Have developed water use technology (world leaders) to reduce water use in agriculture • More water available due to changes in water use in other sectors (more closed loop technologies) • Technological solutions -recharging groundwater using e.g. SUDS; and cheap/low energy (renewable)

water from desalination

Price and availability of staple crops • Possibly changes in what are considered ‘staple’ foods or crops • Leguminous wheat or cheap/low energy way of fixing nitrogen • More C4 plants e.g. sorghum, millet, with better water use efficiency • Measure output in terms of nutrient per resource, as production methods change, e.g. per t carbohydrate • All food prices higher, particularly protein (due to pressures from increased population, rise in demand

from developing world etc.)

Land use and land productivity • Better management of marine environment - as potential for resource production • Vertical farming/aquaponics/hi-tech urban agriculture • Concentrating production onto a smaller area • Sustainable intensification – taking an integrated approach • Optimal scale achieved - at a lower area “return of the family farm”

Price and availability of resources (land, water, energy) • Energy prices expected to increase, • Increasing pressures on land will increase land prices • Higher cost of fossil fuels will drive development/use of renewable resources/technologies

o More biofuels (potentially using GM?) will also cause increased pressures on land availability • Pressures on land use for materials, food and fuel, will drive massive land use change

30


• Closed loop use of food waste throughout the supply chain • Reductions in food waste due to biosensors • Precision agriculture driving resource use efficiency

Global food markets • All food prices higher due to global food demand • Rest of the food market is also innovating driving flow of technology ideas • We (the UK) are best at producing so will manage and remain that way

Global demand for food products:

• Global demand for food high, populations high

Uncontrolled demand

Major drivers under this scenario

Consumption and short-time view lead management and consumptions of resources. Price is the main driver of consumer choices. Short-term gains in productivity drive agriculture and industrial activities. Sustainability and protection of environmental resources are not a priority.

Environmental quality and biodiversity • The environment is not a priority. There a significant discrepancy between beliefs and discourse on the

environment and their actions to support it. • Resource management is ruled by a narrow perspective that favours short term gains against long term

viability. Excessive use of resources may result in irreparable damage to the environment • Deregulation is likely to result in negative environmental impacts (pollution and vulnerability to undesired

natural events.

Water use and availability • There is a lack of investment in an efficient water abstraction structure leading to waste and ill

management of water resources. • High-level consumption from both domestic and industry drives increase in water abstraction levels. • Domestic use of water takes preference over agriculture use. • There is water stress which leads to exposure to drought and shortages

Price and availability of staple crops • Staple crop production is less reliant in state funded subsidies • There is an incremental increase in the price of staple crops • Price increase lead to solution to improve efficiency in production increasing cooperation for sharing

resources • Market drives the vertical integration of the production chain (increasing agricultural production of food

with end use in sight and reducing the open marketing of agricultural goods. Land use and land productivity • Land prices, especially for high quality soil go up • Food demand, stress in resources and inequality of incomes results in the development of two-tier

agriculture (high values goods using higher quality land and extensive agriculture (low input) in marginal and less valuable land. Two-tier agriculture divides the agriculture work force - high value goods attract a highly skilled work force.

• There is incentive to produce legume plants (soya and bean)

Price and availability of resources (e.g. land, energy, etc.) • Price of resources incrementally increase due to management with short term foresight • Increased prices for land, water, fertilizer, seeds, energy • Land prices, especially those with access to resources increase significantly. • Strain in resource lead to vertically integrated farming

Global food markets

31


• UK follows the global market flow trends (does not set trends) • UK has a services driven economy (agriculture is not the main economical driver) • Importation of high value product (according to seasonality) for high income bands of the population • Feed price for livestock will be high which is associated with an increasing demand for meat

overseas • UK focuses on producing high quality products, a portion of which is for the export market • China and India are main import markets • UK import low quality, affordable food goods to meet demand of the lower income band • Migration is high especially within lower income bands

Global demand for food products • UK relies on imports to support (feed) its population; low income band relies on low quality goods; global

demand for food goods increases, driving prices up and increasing stress for resources

Local Resilience

Major drivers under this scenario This scenario was most likely to come about as a result of global rises in food prices. Consumers and government respond by attempting to grow and supply locally. Whilst we still export for economic benefit, much UK stock is used in the UK. As the focus is on economic growth and local resilience, concern for the environment is likely to suffer, with individuals and Governments focused on short term improvements. Environmental quality and biodiversity

• Possible negative implications for the environment, as people cannot afford to care. However, the scenario may drive a move towards local food production and consumption, which may have environmental benefits if attitudes change and individuals are willing to only eat seasonal goods.

Water use and availability • Self sufficient water use becomes more prominent, as people try to cut costs and save on bills • This means more rainwater harvesting, a possible change in attitudes towards water quality (people are

happy to accept slightly lower standards if it is free), and water recycling. • Water scarcity pressures may also drive water sharing • Changes in cropping patterns to meet local production may change the level and location of water

demand pressures in the UK • Opportunities may arise for water reservoirs to become multifunctional and provide services to other

sectors • There may be less competition for water from environmental schemes (due to reduced environmental

concern), and therefore there may be more water available. However, with increased populations and climate change, increases in availability may not translate to an increase per capita

• The amount of regulation regarding water use and quality Price and availability of staple crops

• Prices are possibly higher, but efforts are all focused on reducing costs • ‘Raw’ staple crops are cheaper locally than in global markets/ then processed food Land use and land productivity

• Land prices continue to rise, as such, land ownership may change, as loans increasingly difficult to obtain • This may lead to a higher number of small local units, which may in turn decrease productivity as farmers

are less able to bulk-buy fertilizers etc. • To combat this, farmers may need to work in cooperatives • To meet local demand for different product offerings, more diverse local production may be needed.

However, this will have detrimental effects for the environment. • A “Local UK” national strategy would therefore be more appropriate, whereby farmers grow produce

close to its intended users wherever feasible and sustainable. Elsewhere, land use is optimized to ensure the right food is grown on the optimal land, and transported through the UK.

Price and availability of resources e.g. land, energy etc • People adapt homes to decrease water and energy use due to increased prices. Land prices rise

32


Global food markets

• We are in the scenario because of global markets and high food prices; we are less responsive to global demand. As we become more internally focused, we need to produce a greater mix of crops to meet the wide UK demand (which is no longer met through imports). As such, cropping patterns will change.

• Consumers become less interested in quality of food (aesthetics etc.) and focus more on affordability Global demand for food products

• There is high global demand However, we don’t engage in trying to meet it Consumer attitudes and behaviour

• To reduce costs, more people grow veg at home • Urban farming takes off, and there is more creative use of urban space • Consumers accept seasonal growth • Less food processing occurs (to reduce energy usage) • Local supermarkets offer local foods, or more markets emerge Other

• There is an urban vs. rural tension in this scenario, as demand in rural areas may be more easily met than demand in urban areas

• There is also a polarization of education in urban areas • Children generally become more engaged with farming and food education for the youngest occurs • The next generation may therefore be more positive towards environmental concerns, even though their

parents were focused on economic recovery • There may be the creation of more local jobs • Aging populations are likely to affect • Food security and resilience is a key issue, as pest/disease induction, major floods, or other disasters

could wipe out the food stock

Sustainable Behaviour

Major drivers under this scenario:

Education; consumers willing to pay for sustainable goods; consumers demanding sustainable practices, and strong government regulation of agricultural practices and markets Environmental quality and biodiversity

• Multi-use of land e.g. agriculture plus habitat preservation and maintenance of ecosystem services. Integrated approach

• Consumer demand for farms to operate within their environmental limits. Water use and availability

• Climate change may bring extreme weather events so there is potential for farmers to build reservoirs and store water during periods of heavy rain for periods of drought.

• Stored water used for collective good so those who contribute have some form of water security in return.

• A change in attitude toward risk (risk adverse) • A reduction in domestic water use in households and precision water-use practices in agriculture may

increase water availability • Use water multiple times e.g. have farms at the base of catchments so water is applied to crops last

before going into the sea and after having been used many times further up catchment Price and availability of staple crops • Move crops e.g. potatoes to less suitable areas where produce is not currently grown • Standards of produce would fall but so would price and acceptance e.g. deformed potatoes would be

saleable and therefore less waste. Better for farmers too to be able to sell ‘substandard’ produce • Socially unacceptable to have pre-packed food etc. Land use and land productivity

33


• People open to government imposed zoning to ensure agriculturally productive land is used for farming. Price and availability of resources e.g. Land, energy etc:

• Vulnerable to energy costs • Would be acceptable to consume ‘substandard’ produces as conserves energy and less wasteful. • Waste reduction to feed the world i.e. increased food available, as food is no longer wasted. • Ethical and regulatory commitment to the utilization of waste e.g. Production of biofuel. Global food markets • UK small player in the global food market and has a high cost economy • UK has an emphasis on fresh veg (niche market) and the market is close by e.g. South eastern England

and Europe. • In order for the UK to take part in the global market as a sustainable producer the UK would have to

convince the rest of the world to follow suit in order to be competitive. Global demand for food products

• More people to feed, but we might not always be able to import, so there is a need to look at what we (UK) can produce. UK strong with livestock production

• Change in diet both UK and global, reducing dependence on meat but still there. Change in agriculture to have more alternative proteins.

• Commitment to ethical food and fair trade. • GM may prove more sustainable and therefore be acceptable. • Out of season produce and niche products reduced as resource intensive • High prices but consumers are aware of costs and sustainable production is a driver of consumer demand

i.e. People are prepared to pay for sustainably produced goods.

Working in small groups, workshop participants were then asked to consider the impacts of the seven ‘drivers of change’ on five key micro-components of water demand, namely, national consumption, the proportion of the crop grown in the UK, yield, the cropped area and proportion irrigated. For each EA socio economic scenario, the key findings are summarised below, highlighting the group’s consensus regarding (i) the direction of change and (ii) the magnitude of change for, each micro-component of demand, and in each of four defined sub sectors (arable, potatoes, horticulture, livestock, and biofuels).

3.3.1 Innovation

In the EA briefing document (EA, 2012) this scenario specified that yield was set to double, which therefore constrained some discussion of potential impacts on the micro-components of demand.

Table 7 Expected direction and magnitude of change for selected micro-components of demand in the “Innovation” socio economic scenario.

Innovation

Sector National consumption

Proportion UK-grown

Yield Cropped area Proportion irrigated

Arable Potatoes Horticulture Livestock Biofuels (Unlikely to

be irrigated)

34


Arable Provides basics and livestock feed. Land competes with biofuels. National consumption would increase moderately due to population increase, mitigated by reducing waste.). A crossover between arable and biofuels was recognised, with 2nd generation biofuels using waste products from arable production.

Potatoes National consumption will increase (net), but less per capita. Seasonality and low cost frozen products will drive our imports. A longer growing season due to changing climate, and yield increases mean the UK can be more self-sufficient. The irrigation proportion is likely to be higher, but the volume of water received likely to be less to sustain each irrigated hectare. Rain-fed production is deemed too risky due to unsuitable soil quality in high rainfall areas and lack of control in terms of water holding capacity of the soil and unpredictability of rainfall.

Horticulture Demand for novel and more horticultural crops will increase. Diversity of cropping and demand for this will increase. National consumption was expected to increase across an increasingly diverse range of crops. Changes in seasonality will increase the crops we are able to grow. Technology such as CHP will increase yield and crop type potential. Cropped area was expected to increase moderately due to our ability to produce more, import substitution, increases in yield compared to current reliance on imports in this sector. Climate change may mean there is a shrinking area for production.

Livestock National consumption is likely to increase, though per capita production is likely to go down. Will population be more conscious of how much we eat? Healthier diets and will buy less meat in future, and therefore less waste? In terms of yield it was questioned whether this was production per head of livestock or number of heads of livestock increasing (yield per unit water resource?). Livestock will drink less per unit of meat produced; the yield of feed per unit land will increase. Reliance on area will go down, but the need for water to support that area will increase. It was thought that national consumption was dependent on population and uptake of synthetic meat/protein substitution. Proportion home grown was thought increase moderately with technology and quality of UK grass, also higher welfare attracts premium. Cost of labour reduced by technological advances. Yield will only slightly increase – e.g. intensively housed livestock will drink more due to concentrate feed, but will have quicker fattening/have higher yield. In terms of cropped area – if yields of grain feeds set to double, pig and poultry feed land requirements would be reduced. Cropped area may be reduced due to pressures on land, will feed livestock more on by-products. Forestry on uplands may also affect production. Proportion irrigated may increase moderately due to technology growth to maximise water use, but will avoid more irrigated feed.

Biofuels International aspect – may be comparative advantage to produce (and therefore source) elsewhere -electricity and fuel much cheaper to transport than perishable foods. National consumption will be driven by price rises in energy price. There will be an improvement in efficiency of extracting energy from biomass. Proportion home grown will be globally driven, by volatility in global markets. Yields will increase moderately due to improvements in technology to be more efficient at extracting energy, mass yields and combustion efficiency improvements. Cropped area may increase moderately from a low current baseline. It was thought biofuels were unlikely to be irrigated.

3.3.2 Uncontrolled demand Arable Increase in population generate an increase in demand, arable land is used for cereals and legumes. Domestic consumption for national product remains the same, and for demands are met through imports. Technological and management advances improve yield and adaptation to climate change drive an increase in water consumption.

Potatoes Global population will increase by c40% associated with increases in consumption in lower income bands (affordable replacement for rice and pasta as low cost food) resulting in demand increasing by +50%. Demand remains high (currently UK produces 90% of domestic consumed

35


potatoes. Incremental increase in yield amounts to 20% by 2050 through technology, management techniques and increasing land area used for producing potatoes. Water consumption increases in response to climate change.

Table 8 Expected direction and magnitude of change for selected micro-components of demand in the “Uncontrolled demand” socio economic scenario.

Uncontrolled demand


Proportion UK-grown


Arable Potatoes Horticulture Livestock Biofuels

Horticulture Increase in population associated with a sight for a better diet and affordability of horticultural goods results in a significant increase in demand and proportion of UK land dedicated to its production. Yield increase based on technological advances, management techniques and better quality seeds produced. Currently UK dedicates minimal land use of production of horticultural products, so increasing demand will drive a significant increase in horticulture dedicated cropped land. Increase in cropland proportion associated with the effects of climate change result in the significant increase in irrigated area and water use.

Livestock Global population increases of 40% drive an increase in demand for meat products, which is offset by a significant decrease in meat demand by lower income bands of the population (affordability). Thus, there is little change to current level. Water use increases in response to climate change.

Biofuels The UK has an energy system dependent of fossil fuels where energy from bio-fuel is not significant, used mainly to meet local demands. Land use stress favour food production reducing the proportion and land use for UK produced energy crops. Water use for energy crop remain the same to the negative effect of climate change

3.3.3 Local resilience Arable Almost everything needs to increase for arable production in this scenario, as we need carbohydrates and potatoes are deemed too water intensive to fully meet this demand. In addition, arable crops are easier to store, and can be grown on a wide variety of land. Whilst productivity per hectare is likely o to decrease due to climate change, new technologies (such as drought-resistant crops) lead to a marginal increase in yield. Cropped area must also increase. However, it was noted that this would be achieved mainly through removing the environmental borders around farms (implemented for ELS funding). We therefore maximise the current land, as opposed to using new land. This will have negative environmental implications. Whilst irrigation might increase locally, there is unlikely to be an overall national increase in arable irrigation.

Potatoes Whilst demand per capita decreases, the overall increase in population is likely to mean national consumption stays roughly level. As government focus on meeting local demand from local supplies, the proportion of UK grown potatoes starts to increase. As a society we begin to grow more potatoes at home, meaning the cropped areas may stay the same. Incremental advances in technology may increase yield per hectare, however, balanced with climate change effects, yield

36


stays relatively constant. Due to climate change, more potatoes will need to be irrigated, although this will be directly influenced by water availability.

Table 9 Expected direction and magnitude of change for selected micro-components of demand in the “Local resilience” socio economic scenario.

Local resilience


Proportion UK-grown



Horticulture Would increase, specifically for crops that require less irrigation (e.g. Brussels sprouts). Individuals eat more fruit and veg and start to do more growing at home. Less fruit/vegetables are imported, and UK consumers adopt behaviours in line with seasonality.

Livestock Although the amount consumed per person decreases, population increase means that consumption levels stay the same overall. Whilst meat is possibly less sustainable, land suitability in some areas means that livestock is all you will get. If the current subsidy to prevent overstocking is removed, stocking density increases will lead to higher yield per hectare. Therefore, yield will increase, but area used will stay the same. However, this will have negative environmental implications. Increased yield due to overstocking means that we need to import less meat, and less poor quality processed food.

Biofuels It was felt that under this scenario, we couldn’t afford to use land to produce biofuels extensively in the UK. Biofuels come second to food. Government would need to remove the tariff/financial support for biofuel production. However, we did note that in this scenario, energy security would be an issue in the UK (due to high global prices). As such, we will need to invest more heavily in other types of local energy generation such as wind, wave, anaerobic digestion, or waste energy programs.

3.3.4 Sustainable behaviour Arable Legumes may increase for nitrogen fixation and as an alternative protein source. Return to crop rotation rather than fertilizer. UK grown will increase due to government regulation. Technology/breeding/better practice will improve yield. As yields increase not so much extra land will be needed for arable cropping. Water use will decrease as currently water is being used for these crops just because they can, but if pushed to reduce irrigation this would be possible. Resource constraints may challenge yield increase but use of animal waste etc. as fertilizer may bridge the gap.

Potatoes Possible backyard production and shift toward increased potato consumption with less importation and subsidies. Water use is high but there would be more water available for agriculture due to reduced domestic water use. Water to be used on potatoes rather than other crops as potatoes grow well here and are an important national crop. More area planted in potatoes and a shift in view to a valued crop. Perhaps a move for this crop back to wetter areas. Less fertilizer water etc. going into the system due to technological advancements and better cropping practices. Open to international trade especially in Europe as transport and carbon costs not too extreme.

Livestock Decrease in red meat consumption but more chicken and fish eaten as a substitute. Ethically driven change. Meat mostly grown in UK as it is something that is produced well here and

37


has a lower carbon footprint if not imported. Backing local producers (farming is historic and valued here). Fewer animals on the land but higher yield e.g. less cattle but each head producing more meat. Little change in area in livestock feed or irrigation.

Table 10 Expected direction and magnitude of change for selected micro-components of demand in the “Sustainable behaviour” socio economic scenario.

Sustainable behaviour


Proportion UK-grown



Biofuels High consumer demand. Little land in biofuel production now but will increase drastically although still be a small part of land use compared to food crops. Sub-irrigation, no irrigation or use of lower quality water as irrigation for this crop is not a priority. Use of marginal land for this crop although it may still cause a water burden in these marginal areas.

Horticulture Vegetarian-based diets due to a change in food production so increased national consumption. Seasonal food choices and increased yield due to technology, plant breeding etc. Increased consumption but smaller increase in yield compensated for by eating ‘substandard’ (i.e. imperfect fruit etc.) food that would otherwise be wasted.

3.3.5 Quantitative impacts

The scores from the workshop, expressed as arrows, were converted to numeric values from +3 (strong growth) to -3 (strong decline) (Table 11). It should be noted that demand forecasts were generated here for the arable, potatoes and horticulture sub-sectors only; livestock is considered later as a separate category using a different approach, and no estimates for biofuels have been derived either because of the difficulty in developing a baseline demand for this sub-sector.

These values were then converted to ‘% change factors’. Weightings were informed by a review of historical rates of change, assuming they are indicative of possible future rates of change. For example, crop yields have doubled over the last 40 years, so it would be reasonable to assume that they could double (+100%) again over the next 40 years. In contrast, the future change in national consumption for a particular crop commodity is likely to be tied to the population growth rate so more conservative increases is more realistic (e.g. +10 to +20%). The change factors (%) for each MCD based on the reported magnitude and direction of change from the EA scenario workshop (Table 11) are summarised below (Table 12).

38


Table 11 Reported scores for each MCD, based on the EA scenario workshop.

MCD, by sub sector Innovation

Uncontrolled demand

Sustainable behaviour Local resilience

(Alchemy) (Jeopardy) (Restoration) (Survivor)

Arable Nat consumption 2.00 2.00 2.00 2.00 Propn UK grown 0.00 0.00 1.00 1.00 Yield 0.00 1.00 1.00 1.00 Crop area 0.00 0.00 1.00 1.00 Propn irrigated 0.00 2.00 1.00 1.00

Potatoes

Nat consumption 2.00 3.00 0.00 0.00 Propn UK grown 3.00 0.00 2.00 2.00 Yield 3.00 2.00 0.00 0.00 Crop area -2.00 2.00 0.00 0.00 Propn irrigated 3.00 2.00 1.00 1.00

Horticulture

Nat consumption 3.00 2.00 1.00 1.00 Propn UK grown 2.00 1.00 1.00 1.00 Yield 3.00 1.00 1.00 1.00 Crop area 2.00 2.00 1.00 1.00 Propn irrigated 2.00 2.00 1.00 1.00

Table 12 Estimated change factors (%) for each MCD based on the workshop outputs.

Conversion factor (workshop arrow)

National consumption

Proportion UK grown Yield Crop area Proportion

irrigated

Change (%) Change (%) Change (%) Change (%) Change (%)

-2.00 () 0.80 0.90 0.50 0.60 0.60 -1.00 () 0.90 0.95 0.75 0.90 0.80 0.00 () 1.00 1.00 1.00 1.00 1.00 +1.00 () 1.10 1.05 1.25 1.05 1.20 +2.00 () 1.20 1.10 1.50 1.20 1.40 +3.00 () 1.30 1.20 2.00 1.30 1.60

These change factors were then mathematically combined to estimate overall ‘change in water use’ or growth factors, for each agricultural sub-sector (arable, horticulture, potatoes).

An example for arable cropping is shown overleaf in Table 13. For Innovation, the estimated change in water use is +332% (recognising of course that the baseline demand for arable is negligible).

The change in water use (growth factors) for the 2050s, by socio economic scenario and crop category are summarised in Table 14. The Innovation and Uncontrolled Demand scenarios suggest

39


growth rates in irrigation demand of +157 and +167%, respectively. Much lower values for Sustainable Behaviour and Local Resilience are estimated (+42 and +40%).

By applying these growth rates (%) to the baseline (current) demand, the projected irrigation volumes (Ml year-1) for the 2050s, by socio economic scenario were estimated (Table 15). Two sets of demand forecasts are presented here for comparison based on two alternative baselines – the actual Defra Irrigation Survey 2010 baseline, and the derived ‘design’ dry year 2010 baseline. Table 13 Projected rates of change (%) in water use in arable cropping, by EA scenario, for the 2050s.

ARABLE Baseline 2010 Innovation

Uncontrolled demand


Local resilience

National consumption; factor

1.20 1.20 1.20 1.20 so consumption per head

0.91 0.85 0.99 1.02

Change in % grown in UK (asymptotic)

1.00 1.00 1.05 1.05 so % grown in UK 0.25 0.25 0.25 0.26 0.26 so change factor % grown in UK

1.00 1.00 1.05 1.05

change in yield factor

1.00 1.25 1.25 1.25 so area cropped change

1.20 0.96 1.01 1.01

cf expected area cropped change

1.00 1.00 1.05 1.05 Change in % irrigated (asymptotic) 1.00 1.40 1.20 1.20 so proportion irrigated 0.02 0.02 0.03 0.02 0.02 so change factor proportion irrigated

1.00 1.40 1.20 1.20

Climate change impact on OWU

1.20 1.20 1.20 1.20 Change in % of OWU applied

3.00 3.00 1.00 1.00

Overall change in water use 1.00 4.32 4.84 1.45 1.45

Table 14 Overall change in water use (growth factors) for the 2050s, by socio economic scenario and crop category.

Sector growth factors to 2050

Baseline Innovation Uncontrolled demand


Local resilience

Arable 1.00 4.32 4.84 1.45 1.45 Potatoes 1.00 2.06 2.41 1.46 1.43 Horticulture 1.00 3.00 2.54 1.33 1.33 Weighted across the 3 sectors:

1.00 2.57 2.67 1.42 1.40

Table 15 Projected total irrigation volumes (Ml year-1) for the 2050s, by socio economic scenario.

Projected irrigation demand Innovation Uncontrolled

demand Sustainable

behaviour Local

resilience

Based on 2010 Irrigation Survey 179771 187194 99313 98211

Based on 2010 ‘design’ dry year 241706 251687 133529 132047

40


A brief narrative summary of these demand forecasts for each EA scenario is given below:

Sustainability-led governance

Sustainable behavior “Cut out resource use through new ways of managing societies and relationships”

+42% Gradual steady growth, concentrated on potatoes (+56%) and vegetables (+27%). This scenario featured low resource consumption and local food production using greener technologies, with lower yields and less emphasis on quality assurance. Diets more vegetable based than meat, population growth is modest but the decline in imports and lower yields results in significantly more land needed for cultivation. There is little emphasis on GM products and new technologies due to the lack of past investment. There is a larger area under potato irrigation than present, mainly on family farm units rather than large scale agribusiness. Without demands for quality assurance, irrigation is widely used to boost yield.

Innovation “Scientists solve environmental damage through ideas and innovation”

+157% Major increase in irrigation, abstracted within environmental constraints. A high technology and knowledge-led society with consumption in a relatively resource intensive manner led to growth in potatoes (+44%), horticulture (+34%), arable (+14%) and other (+8%). High increases for potatoes and horticulture due to higher per capita consumption and population growth, but offset by substantial yield increases. Crops grown under intensive conditions predominantly managed by agribusiness with a higher proportion irrigated than present and scheduled for high yield and quality under a more arid climate.

Dematerialised consumption Materialised consumption

Local resilience “Better to have fewer wants than greater resources”

+40% Population starts to accept lower quality and locally sourced produce. Reflects society more concerned about environment than consumption. Consumers want naturally grown, sustainable food. Organic and free range becomes more accepted. Farming becomes more extensive, non-organic fertilizers expensive and yields decline. There is concern for water efficiency and low wastage, but efficiency falls in terms of productivity per unit of water due to lower yields. Proportions irrigated decline slowly, and climate change is partly moderated by relocating crops. The increased areas and impacts of climate change result in significant increases in water use on potatoes (+55%) and horticulture (+27%).

Uncontrolled demand “The rich shall inherit the earth – because we’re worth it”

+167% Represents a consumption based scenario that results in the largest increase in water demand (+167%). There is no significant change in overall food demand other than to supply the population increase. Potatoes remain the largest abstractor accounting for nearly half all water use (49%); there is also major growth for vegetable irrigation (27%) and nearly a fifth (15%) for arable irrigation to meet increased population food needs.

Growth-led governance

3.3.6 Sensitivity analysis The forecasts are of course sensitive to the model input values from the stakeholder workshop. For example, another workshop with different key informants could quite feasibly generate a quite different set of model input values with consequent impacts on the demand forecasts.

A sensitivity analysis of the various input values used in the demand forecast model from the stakeholder workshop was therefore undertaken (Table 16).

For most variables, the change is under ±10%. The most sensitive variables are the changes in potato yield, where a single arrow reduction in yield can increase total water demand by +20% in some scenarios (and vice versa).

41


The forecasts are also highly sensitive to assumptions around irrigation of cereal crops – projections would be markedly different if there was large-scale switch from rainfed to irrigated cereal production (note the current baseline irrigated area is so small as to make any % changes unreliable), though it is hard to believe that sufficient water would be available for that. The risk there is all on the upside.

Table 16 Sensitivity analysis showing change (%) in total projected water use for a “one arrow” reduction in the rate of change identified by the workshop participants.

3.3.7 Comparison with previous forecasts In summary, the current trends, particularly for the important irrigated crops, highlight quite major differences to those experienced previously. They suggest that the strong upward growth until around 1990 in the volumes of water abstracted has halted and reversed. This at least partly reflects the increasing yield and hence decreasing cropped area of potatoes and some other major irrigated crops, together with increased efficiency and better scheduling. The increasing restrictions on water availability and reliability, and hence a greater appreciation of its value, are also likely to have contributed. Projecting these trends forward in the Irrigrowth model, and allowing for climate change (but not population growth), suggested a -25% reduction in irrigation water use by the 2030s.

However, the range in the longer-term (2050s) demand forecasts reported in this study (+40% to +167%) is surprisingly similar to the range in estimates in the earlier EA (2008) study (+27% to +167%), especially given that the forecasts have been completely reworked with new independent industry opinion on the micro-component changes, modified socio-economic scenarios and climate change data. In the longer term, demand is forecast to increase under all the scenarios. Combined with a probable decline in low-flow (summer) water availability, this indicates major future water resource issues.

The point at which the shorter term decline changes to meet the longer term increase is not defined, and will likely depend on a host of factors, including actual weather variation.

The projections also ignore any impacts of step-change genetic improvements, such as the introduction of genetic modification techniques and the uncertainties around the effects of elevated CO2 concentrations on crop growth.

InnovationUncontrolled demand


Local resilience

Arable Alchemy Jeopardy Restoration SurvivorNational consumption -1% -1% -1% -1%Proportion UK grown -1% -1% 0% 0%Yield 5% 4% 2% 2%Crop areaProportion irrigated -3% -2% -2% -2%

PotatoesNational consumption -4% -4% -6% -6%Proportion UK grown -1% -1% -1% -1%Yield 10% 13% 20% 20%Crop areaProportion irrigated -3% -5% -9% -9%

HorticultureNational consumption -3% -2% -3% -3%Proportion UK grown -2% -1% -1% -1%Yield 3% 7% 7% 7%Crop areaProportion irrigated -5% -4% -5% -5%

42


3.4 Future livestock water demand A summary of the modelled future water demand for livestock, by sub-sector is given in Table 17. The demand forecasts, which are based only on simple relationships between temperature change and water use, result in future increases in demand of around c10% for sheep and beef cattle, <5% for dairying and 10-20% for poultry. However, these figures are likely to under-estimate the impact of climate change due to the many other indirect impacts, and provide no information on the uncertainty or variability around mean changes in demand. They do not consider changes in diets and/or the numbers of animals.

The CCRA Agriculture Sector Report (Knox et al., 2012) attempted to develop a risk metric to assess climate impacts on water abstraction for livestock, but without success. To examine the impact on the abstraction of water for livestock systems it would be necessary to monitor each category of water utilisation in the system (drinking, animal cooling, plant cooling) and link these to biological models of water requirements by animals. Despite these methodological constraints, the figures presented here at least provide an indicative estimate of climate impact on livestock water demand, but their limitations should be recognised and care taken when extrapolating the data to other situations.

Table 17 Summary of total future volumetric demand (m3), by livestock sub-sector for the 2030s and 205s, low and high emissions scenario.

Sector 2010 2030L 2030H 2050L 2050H

Beef cattle 28904 31342 31468 31973 32780 Dairy cattle 61967 63774 63867 64241 64839 Pigs 7911 7911 7911 7911 7911 Poultry 12847 14357 14435 14747 15247 Sheep 35166 37646 37774 38288 39109

Total 146795 155030 155454 157161 159885

43


4 References ADAS. 2012. Defra project WU0132: Sustainable water for livestock. http://randd.defra.gov.uk/Default.aspx?Menu=Menu&Module=More&Location=None&Completed=2&ProjectID=17363. Accessed on 30th July 2012.

ARC (1980). The Nutrient Requirements of Ruminant Livestock. CAB, Farnham Royal

Bailey, R.J. and Spackman, E. (1996). A model for estimating soil moisture changes as an aid to irrigation scheduling and crop water use studies. I. Operational details and description. Soil Use and Management 12: 122-128.

Bailey, R. J., Groves, S. J. and Spackman, E. (1996). A model for estimating soil moisture changes as an aid to irrigation scheduling and crop-water studies: 2. Field test of the model. Soil Use and Management 12: 12-133.

Comber, A. Procter, C. Anthony, S. 2008. The Creation of a National Agricultural Land Use Dataset: Combining Pycnophylactic Interpolation with Dasymetric Mapping Techniques. Transactions in GIS. 12(6): 775–791

Environment Agency (2050) Demand for water in the 2050s – briefing note, Environment Agency, 2008, Bristol, 12pp.

IPCC-TGCIA (1999). Guidelines on the use of Scenario Data for Climate Impact and Adaptation Assessment, Version 1. Prepared by Carter, T.R., Hulme, M., Lal, M. Intergovernmental Panel on Climate Change, Task Group on Scenarios for Climate Impact Assessment.

Jenkins, G.J.,Murphy, J.M.,Sexton,D.S., Lowe,J.A.,Jones,P. and Kilsby,C.G.(2009). UK Climate Projections: Briefing report. Met Office Hadley Centre, Exeter, UK.

King, D. Tiffin, D. Drakes, and K. Smith (2005) Water Use in Agriculture: Establishing a Baseline. Final Report, Project WU0102, 2005, Defra.

Knox, J.W., Rodríguez Díaz, J.A., Nixon, D.J. and Mkhwanazi, M. (2010). A preliminary assessment of climate change impacts on sugarcane in Swaziland. Agricultural Systems 103(2): 63-72.

Knox, J. W., Weatherhead, E. K. and Bradley, R. I.(1997). Mapping the total volumetric irrigation water requirements in England and Wales. Agricultural Water Management 33: 1-18.

Knox, J.W., E.K. Weatherhead, J.A. Rodríguez Díaz, and M.G. Kay, Development of a water-use strategy for horticulture in England and Wales – a case study (2010) Journal of Horticultural Science and Biotechnology 85(2): 89–93.

Knox, J.W., E.K. Weatherhead, J.A. Rodríguez Díaz, and M.G. Kay, Developing a strategy to improve irrigation efficiency in a temperate climate – A case study in England (2009) Outlook on Agriculture, 38(4), 303–309.

Nakicenovic, N., Alcamo, J., Davis, G., de Vries, B., Fenhann, J., Gaffin, S., Gregory, K., Grübler, A., Jung, T.Y., Kram, T., La Rovere, E.L., Michaelis, L., Mori, S., Morita, T., Pepper, W., Pitcher, H., Price, L., Raihi, K., Roehrl, A., Rogner, H., Sankovski, A., Schlesinger, M., Shukla, P., Smith, S., Swart, R., van Rooijen, S., Victor, N. and Dadi, Z. (2000). IPCC Special Report on Emissions Scenarios. Cambridge University Press, Netherlands.

NRC (2001). Nutrient Requirements of Dairy Cattle. 7th Revised Edition. National Academies Press, Washington, DC.

NRC (2000). Nutrient Requirements of Beef Cattle. 7th Revised Edition. National Academies Press, Washington, DC.

44

http://randd.defra.gov.uk/Default.aspx?Menu=Menu&Module=More&Location=None&Completed=2&ProjectID=17363



NRC (1998). Nutrient Requirements of Swine. 10th Revised Edition. National Academies Press, Washington, DC.

NRC (2006). Nutrient Requirements of Small Ruminants. National Academies Press, Washington, DC

NRC (1994). Nutrient Requirements of Poultry. 9th Revised Edition. National Academies Press, Washington, DC.

Rees, B., Cessford, F., Connelly, R., Cowan, J., Bowell, R., Weatherhead, E.K., Knox, J.W., Twite, C.L., and Morris, J. (2003). Optimum use of water for industry and agriculture. Phase III. R&D Technical Report W6-056/TR1. Environment Agency, Bristol.

Rodríguez Díaz, J.A., Weatherhead, E.K., Knox, J.W. and Camacho, E. (2007). Climate change impacts on irrigation water requirements in the Guadalquivir river basin in Spain. Regional Environmental Change 7:149–159.

Silgram, M., Hatley, D., Gooday, R. 2007. IRRIGUIDE: a decision support tool for drainage estimation and irrigation scheduling. Proceedings of the 6th Biennial Conference of the European Federation of IT in Agriculture (EFITA) World Congress on Computing in Agriculture (WCCA) 2007 joint conference “Environmental and rural sustainability”. Glasgow, UK, 2-5 July 2007.

Thomasson, A.J. and Jones, R.J. (1991). An empirical Approach to Crop modelling and the assessment of land productivity. Agricultural Systems 37: 351-367.

University of Warwick and ADAS. 2006. Defra project WU0101: Opportunities for reducing water use in agriculture. http://randd.defra.gov.uk/Default.aspx?Menu=Menu&Module=More&Location=None&Completed=0&ProjectID=14402. Accessed on 30th July 2012.

Weatherhead, E.K (2005) Survey of irrigation of outdoor crops in 2005: England and Wales, 2006, Cranfield University, Cranfield, UK.

45




5 Appendix 5.1 Defining a baseline for demand assessment Irrigation demand varies greatly from year to year depending on the summer weather. For comparability therefore, all irrigation water demand forecasts relate to a ‘design dry year’. This is defined as a year where the unconstrained volumetric irrigation demand has an 80% probability of non-exceedance. This criterion is widely used for planning irrigation water resources such as abstraction licencing and reservoir sizing; the farmer would then have adequate water for 80 years in 100, but would have to under-irrigate some crops in the other 20 (This value of planned headroom is considerably less than for other sectors). Forecasts also need to be made relative to a baseline reference year - in this study 2010, the year in which the most recent Defra Irrigation Survey was conducted. For the future demand forecasting this was then adjusted to a ‘2010 dry year’, ie reflecting what would have happened if a ‘design dry year’ had occurred in 2010.

To assess climate variability across England and Wales a simple water balance model was used to estimate the potential soil moisture deficit (PSMD) using monthly rainfall (P) and reference evapotranspiration (ETo). Potential soil moisture deficit (PSMD) has previously been applied to quantify the irrigation needs at national scales in different countries (e.g. Knox et al., 2010; De Silva et al., 2007; Rodríguez Díaz et al., 2007; Knox et al., 1997). It is also used by the regulatory authority in England and Wales for setting licences (permits) for irrigation water withdrawal (abstraction) (Rees et al., 2003). The advantage of this index over others such as the Wetness Index (ratio of total annual rainfall and total annual evapotranspiration) is that the distribution of rainfall and ETo throughout the year is taken into account. Previous studies have also established a direct correlation between irrigation demand (m3) and PSMD. Figure 13 shows, for example, for selected regions the correlation between EA reported irrigation abstraction and PSMD. Using PSMD as an index of agroclimate to assess irrigation demand is thus a useful, simple and relatively robust variable in demand forecasting.

Figure 13 Linear regression correlation between average agroclimate (PSMDmax) for the region and the EA reported volume of water abstracted (m3) for irrigation, by region.

46


In order to assess the spatial variation of PSMDmax across England and Wales, a gridded 5x5 km monthly dataset available from the Met Office was used to calculate the long term average PSMDmax (1961–1990) for each grid pixel. The data were then classified into agroclimatic zones. The agroclimatic zones with the highest PSMDmax are located in the eastern and south eastern parts of the country, notably in Norfolk, Suffolk, Essex and Kent. These correspond to parts of the country where irrigation is most concentrated (Knox et al., 1997) and where the reliability and availability of water resources for agriculture are under severe pressure (EA, 2001). In contrast, zones with the lowest PSMDmax (< 75 mm) extend across much of Wales, the south west and north-west of England.

However, the latest gridded monthly data produced by the Met Office correspond to 2006 and therefore it was necessary to produce a PSMDmax map for 2010 using an alternative and indirect method. The Met Office produce a summary report for each year where regional monthly climatic parameters of that year are compared against the long term historical averages. Hence, each grid (5x 5km) of the long term average PSMD (1961-1990) was multiplied by the appropriate climatic changing factor for each region to produce a 5km grid of 2010 PSMD (Figure 14). The main limitation of this method is that it assumes the weather pattern of 2010 is similar to the long term historical one and the changes in the weather parameters are uniform across each region. The baseline map (Figure 2a) shows the long term average PSMD between 1961 and 1990 and highlights the agroclimate zones where irrigation needs are currently greatest including parts of Suffolk, Kent, areas in West Midlands, Nottinghamshire, and the south coast.

Figure 14 Spatial variability in agroclimate across England and Wales, using PSMDmax as an aridity indicator; based on the long term average (1961-1990) (a), and data generated for 2010 (b).

(a) Long-term average agroclimate (b) Agroclimate in 2010

5.2 Crop modelling The crop scenarios, irrigation scenarios, soil types and locations used for the IRRIGUIDE modelling were based on those described in Knox et al (1997). Information on crop type, planting date, emergence date, harvest date, crop cover %, and max rooting depth were input into IRRIGUIDE to

47


define the crop scenario. These agronomic parameters were sufficient to provide realistic crop growth outputs for each of the crop types tested. There were a total of seven crop types (Table 1) each with an associated irrigation schedule which differed depending on soil type (Table 2). The crop/irrigation scenarios were modelled for three differing soil types of low, medium and high available water capacity (AWC): loamy medium sand (LMS), medium sandy loam (MSL) and a loamy peat (LP). Each crop/irrigation/soil scenario was modelled at 11 agroclimatically different locations with Irriguide requiring an OS Grid Reference and altitude for each site. All scenarios were modelled from 1986 – 2009 inclusive providing 24 years of model output data.

A total of 5016 annual irrigation demand estimates were generated using the cropping and site location information described above. Within these model runs there was no attempt to artificially reset the SMD to zero at the start of each season. This decision was taken because all crops were irrigated and should therefore not have given rise to significant SMDs at the end of each season thus allowing SMDs to return to zero naturally following winter rainfall. Also, this approach is more representative of reality for crops such as permanent grass, as low winter rainfall is then recognised in the following season’s irrigation demand.

Table 18 Crop characteristics used for Irriguide modelling (after Knox et al., 1997).

Crop Characteristics Early potatoes

Maincrop potatoes

Sugar beet

Wheat Grass Carrots

Date of planting 01-Mar 01-Apr 01-Apr 01-Apr na 01-May Date of emergence 10-Mar 10-Apr 15-Apr 10-Apr na 09-May Date of 20% cover 01-Apr 10-May 15-May 25-May na 01-Jun Date of full cover 20-May 10-Jul 01-Aug 15-Jul na 15-Aug Date of maturity 20-May 10-Aug 01-Aug 15-Jul na 20-Sep Date of harvest 01-Jun 15-Sep 09-Oct 10-Aug na 20-Sep Date of max root depth 20-May 10-Jul 01-Aug 15-Jul na 15-Aug Planting depth (m) 0.05 0.15 0.05 0.03 na 0.01 Max root depth (m) 0.55 0.70 1.20 1.20 1.00 0.80 Max crop cover (%) 100 100 100 100 100 100 Mulch cover at planting (%) 0.00 0.00 0.00 0.00 0.00 0.00 Crop coefficient full cover (Kc) 1.10 1.10 1.00 1.00 1.00 1.00

A review of the data generated (excluding permanent grass) confirms that in all seasons and at all sites on all soil types SMDs returned to zero for early potatoes and main crop potatoes. In the remaining scenarios field capacity was reached at all LMS sites, 99% of MSL sites and 92% of LP sites. The incidence of sites not returning to field capacity was greatest for sugar beet on LP soils. It is not uncommon, on such high AWC soils, for significant SMDs to develop under deep rooted crops such as sugar beet or wheat, so the failure to return to field capacity in some seasons reflects reality. In the context of this study, it is unlikely that a carry-over SMD from the previous season substantially affected irrigation demand estimates produced by the model as the data indicate that, in all cases, the root zone SMD returned to zero.

To map the irrigation needs spatially, a correlation between the irrigation needs calculated with IRRIGUIDE and the nationally mapped climatic dataset is needed. Therefore, the maximum potential soil moisture deficit at each of the 11 selected weather stations was calculated and correlated against the corresponding irrigation need of a design dry year and at different soil type. An example of the correlation is presented in Figure 23 and the final regression equations per crop and soil type are listed in Table 22. Using this look up table, it is possible to estimate the theoretical irrigation need of each crop based on the pedo-climatic characteristics of the area where they are grown.

48


Table 19 Irrigation schedule for each crop by soil type determined by Available Water Capacity (AWC); mm water applied at mm soil moisture deficit (after Knox et al., 1997).

Crop type Irrigation period Low AWC Medium AWC High AWC

Early potatoes 1st May - 30th Jun 15@25 25@40 25@50 Main crop potatoes 1st May - 31st May 12@15 12@15 12@15 1st Jun - 31st Aug 25@30 30@55 30@70 Carrots 1st May - 31st Aug 20@25 40@50 Unirrigated Permanent grass 1st May - 31st Aug 25@40 25@75 25@100 Spring wheat 1st May - 30th Jun 25@50 25@80 25@100 Sugar beet 1st Jun - 30th Jun 20@25 25@35 25@75 1st July - 31st July 25@35 25@50 25@150 1st Aug - 31st Aug 25@50 25@75 25@200 1st Sept - 30th Sept 25@65 25@125 25@250

Table 20 Soil types and characteristics

Characteristic

Available water capacity (AWC) Low Medium High

Soil type Loamy sand Medium sandy loam Loamy peat

Soil depth (m) 0.5 0.5 0.5

Total AWC (mm m-1) 130 170 350 Easily available AWC (mm m-1) 90 110 260 Subsoil total AWC (mm m-1) 90 150 350

Subsoil easily available AWC (mm m-1) 60 110 260

Table 21 Linear regression equations derived for each crop category on three different soil types.

Crop Loamy sandy Medium SandyLoam Loamy peat

a b r2 a b r2 a b r2 Earlies potato 0.21 5.61 0.88 0.17 3.01 0.88 0.19 -22.83 0.96 Maincrop potato 0.46 104.21 0.90 0.39 84.07 0.90 0.40 45.63 0.90 Grass 0.91 90.61 0.95 0.96 47.15 0.94 1.07 4.04 0.94 Sugarbeet 0.66 72.36 0.93 0.66 21.57 0.91 0.56 -85.21 0.89 Wheat 0.22 -12.02 0.87 0.11 -17.93 0.62 0.07 -11.99 0.38 Carrots 0.18 6.34 0.82 0.10 4.09 0.53 - - - Strawberries 0.23 78.33 0.58 0.24 57.65 0.75 - - -

49


Figure 15 Correlation between Irriguide modelled design dry year irrigation needs (IN) and agroclimate (PSMDmax) for maincrop potatoes grown on three contrasting soil types (LMS, MSL, LP).

50


Figure 16 Total ‘design’ dry year volumetric irrigation water demand (m3) by crop category in England and Wales.

(a) Total potatoes (b) Sugar beet

51


(c) Field vegetables (d) Cereals

52


(e) Soft fruit (f) Grassland

53


Figure 17 Total annual average volumetric water demand (m3) for livestock, by sub-sector in England and Wales.

(a) Beef cattle (b) Dairy cattle

54


(c) Poultry (d) Pigs

55


(e) Sheep

56

Defra FFG1129 (Phase I)

57

Documents

Assessment of the impacts of climate change and …randd.defra.gov.uk/Document.aspx?Document=11705_DefraFFG...2013/12/05 · Defra’s Sustainable, Secure and Healthy Food Supply