ISEIS Report on UN-ESCAP & SSEF supported Participants in 2018-2019

Report

On

Ms. Kisolel Lina POSANAU

Climate Scientist - PNG National Weather Service, Papua New Guinea

Mr. Heshan GUNASEKERA

Computer Programmer - Arthur C. Clarke Institute for Modern Technologies, Sri Lanka

For MSc in Geo-survey and Public Management (2018-2019)

Travel Grant by: UNESCAP

Scholarship funded by: SSEF

Institute of Space and Earth Information Science

Chinese University of Hong Kong, Shatin

Hong Kong

April 2019

Master of Science in Geo-Survey and Public Management at ISEIS

The programme commenced on September 3, 2018 at the Institute of Space and Earth

Information Science in the Chinese University of Hong Kong. The full-time MSc programme

covers one year with three semesters. The objectives of this programme are:

1. Students get an appreciation of geo-information on national conditions 2. Be knowledgeable in geo-survey and geo-analysis methods 3. Be better prepared for public policy studies and precise public management

The academic staffs at ISEIS are very skilled professionals and knowledgeable in their field or

area of study and all the classes were taught in English.

were superb with very high level of skills and knowledge in their field or area of study and all

the classes were taught in English.

REPORT: Heshan GUNASEKERA

A. The course modules

Total courses that I took in both semester one and two were 8 courses (6 major and 2 electives).

The final course project in which we write our thesis is the major course project to be taken in

summer term. This will have 4 credit points. The oral defence is set later at the end of June.

B. Field trips

In Semester one I participated in a field trip at Tai O fishing village for the course

Environmental Remote Sensing Technology, we classified a satellite image of the study area

prior to the field trip and observer the true locations and compared to our classification results.

In semester two, I participated in a combined field trip for the Global Navigation Satellite

System (GNSS) and Sensor Network for Environmental Monitoring course and

Geoinformation Technologies for Risk and Crises Management course to Tung Ping Chau

island. We tracked the whole field trip with GPS trackers provided to us by Dr. Li Gang

(Lecturer of GNSS course).

C. Projects and ongoing research

For the last two semesters, I did 6 group projects and two individual projects. From our research

and the project, I intend to continue with funding when I return to my country. Here are some

outline information of the projects and research work that was carried out.

� Individual Projects 1. Investigating Impact of deforestation on mountain areas for landslides in Sri

Lanka.

Course - Introduction to Geo-survey and Public Management

Instructor - Dr. WONG C. Janice

Landsat Images with 30m resolution and R, G, B and NIR Bands was selected as the

data for spatial analysis. Central mountain area of the country above 300m height was selected

as the study area. Satellite images of 1992 and 2001 was Classified and the overall accuracy of

land use classifications were 98.20% and 97.41% respectively. It showed that the “bare land”

area has been increasing and the forest cover has been decreasing when the results were

compared.

Satellites images of 1988, 2006 2009 and 2018 was used to calculate NDVI and then was

compared to the number of landslides events in those years except 2018. This comparison

clearly shows a direct proportional relationship between the deforestation of mountain areas

and the increasing of landslides incidents.

So, the results of this project are clearly emphasizing that it is very important to do reforestation

in the mountain area of the country to minimize the number of annual landslides.

2. Comparison of Centralities between cities having Higher Traffic Indices & Lower Traffic Indices.

Course - Urban Networks

Instructor - Dr. WONG C. Janice

The main goal of this network analysis project is to identify whether there is a

relationship between common network centrality measures and the traffic situations in the road

networks of major cities around the world.

Urban Road Network Data was retrieved as shapefiles from website called FIGSHARE. this

dataset contains the Urban road networks of 80 of the most populated urban cities in the world.

Then 8 cities were selected in 2 categories considering the available road network data and the

current Traffic Index values retrieved from a website called NUMBEO. 2 categories are cities

with good traffic management and cities with Bad traffic management.

Then the network centralities of the road networks of the selected cities were calculated in

ArcGIS with the help of “Urban Network Analysis Toolbox”. 3 centralities were calculated

namely Betweenness Centrality, Straightness Centrality and Closeness Centrality

Finally, the average values of these calculated centralities were compared between all the

selected cities along with the population densities of those cities.

� Group Projects 3. Investigating Invasion pattern of Sonneratia in Mai Po Nature Reserve.

Course - Special Topics in GeoInformation Sciences

Instructor – Prof. Hui Lin & Dr. Ann Mak

Group Members - Luoma WAN & Heshan Gunasekera

This project is carried out to find out where is the Sonneratia in Mai Po, what is the distribution

pattern in Mai Po area as well as to examine what factors have supported the distribution

pattern. Since Sonneratia was first found in 2000 inside the Mai Po Reserve, we used high

resolution image of Mai Po mangrove forest area from Google Earth in February 2000 and

February 2003.

After performing various GIS operations and analysis we found that waterway is an important

fact for Sonneratia distribution, but not the major driven factor. Tide is the major factor for

Sonneratia distribution along with the support from the mudflat.

4. Investigating Spatial and temporal land cover changes in Xiong'an New Area in 2016-2018.

Course - Environmental Remote Sensing Technology

Instructor – Prof. Hongsheng Zhang

Group Members - YAN Yizhen, JIANG Xiaolu & Heshan Gunasekera

This project aims at analyzing the short-term land cover changes in order to explore the

development nature of Xiong’an New Area. The remote sensing images from Sentinel-2

satellite were used for a supervised classification through the Support Vector Machine (SVM)

algorithm in the three years of 2016, 2017 and 2018. The accuracy of the land cover

classification ranged from 86.0% to 94.5%. The results revealed that the urban land areas were

consistently increased, while the rate of expansion was decreasing in the three years. In

addition, the spatial distribution of urban land cover changed into the centralization from

decentralization multipoint. These results will be helpful for understanding the development of

the Xiong’an New Area.

5. Spatial analysis of house prices and public facilities in Shenzhen. Course - Spatial Analysis for Public Policy

Instructor – Dr. Li Rongrong

Group Members - Yan Yizhen, Jiang Xiaolu, Kisolel Lina Posanau & Heshan

Gunasekera

This research aims at exploring the relationship between the residential house prices and public

facilities through combining the geographic information system (GIS) spatial analysis and the

geographical weighted regression (GWR) modelling in Shenzhen city, China. The hotspot

analysis and Kriging interpolation were applied to define the spatial distribution of the

residential house prices in Shenzhen city. The study defined the factors as two groups, the

distances from the houses to the nearest public facilities and the number of public facilities in

a certain neighbourhood of each house, respectively. The traditional ordinary least squares

(OLS) models were firstly constructed to identify the key factors that significantly influenced

the residential house prices. Finally, the GWR model results illustrated the residential house

prices in Shenzhen city were more influenced by the factor of the distances from the houses to

the nearest public facilities. Among, the variables that were considered as factors contributing

to house prices was subway, mall and secondary school.

6. Identification of high-risk tree collapse area during typhoon in Hong Kong. Course - Geoinformation Technologies for Risk and Crises Management

Instructor – Prof. Hongsheng Zhang

Group Members - Chan Ching Yan Alice, Wang Tianyi & Heshan Gunasekera

By identifying the high-risk tree collapse area, we can make prevention in advance to reduce

the adverse effect.

With the help of advanced remote sensing technology, we can easily produce risk map like

this which help to do better risk management. We believe that the technology will become

more advance in the future and the result could be produced with high accuracy.

In this project we produced a Tree Collapsing probability Disaster Risk Map of Hong Kong

with the help of Sentinel-2 Satellite images and several GIS tools.

REPORT: Kisolel Lina POSANAU

A. The course modules

Total courses that were taken in both semesters were 9 courses (6 major and 3 electives). The

final course project in which we write and present our thesis is the major course project to be

taken in summer term from May 13 – June 25.

B. Field trips and visits

ESGS5006 course Principals of Hydrology: San Tin Flood Pumping Station, Yuen Long By-

pass Floodway, Compensation Wetland, Fabric Dam and Pumping station, Confluence of Kam

Tin River and Shan Pui River, Water Supplies Department. We visited the sites with our

Professor Chen and combined with Environmental Science students of CUHK. To study how

Hong Kong manages its water systems and flood monitoring systems.

ESGS5019 course Global Navigation Satellite System (GNSS) and Sensor Network for

Environmental Monitoring and ESGS5017 course Geoinformation Technologies for Risk and

Crises Management: Visit to Tung Ping Chau island. We tracked our path with GPS trackers

provided to us by Dr. Li Gang (Lecturer of GNSS). Tung Ping Chau is Hong Kong’s UNESCO

Global Geopark.

C. Seminars

11 Jan Point Cloud Processing at GI Lab by Prof. Ruisheng Wang, Department of Geomatics

Engineering, University of Calgary, Canada. He presented an overview of point cloud

processing in the Geospatial Intelligence Laboratory at their university. Their recent work using

aerial LiDAR data.

22 Feb Integrating Geospatial Technologies, Environmental Science and Health Research by

Prof. Mei Po Kwan – Director of Space-Time Analysis and Research Lab (StarLab),

Department of Geography and Geographic Information Science, University of Illinois at

Urbana-Champaign. Her presentation was on recent conceptual and methodological

developments that help advance interdisciplinary research on interaction between human health

and environment. `

15 Mar Making Sense of the Climate-war Nexus in History by Prof. Harry F. Lee, Department

of Geography and Resource Management, The Chinese University of Hong Kong. Presentation

on the current debate of the climate-war nexus to reveal its controversies and also a simple

conceptual framework proposed to solve those controversies.

22 Mar Remote sensing application and geospatial analysis to evaluate community health risk

associated with weather-related air pollution events by Prof. Hung Chak Ho, Department of

Urban Planning and Design, The University of Hong Kong. Study used Satellite-derived

aerosol optical depth (AOD) product retrieved from the Moderate Resolution Imaging

Spectroradiometer (MODIS) and was applied to estimate spatial variability of air quality during

the weather-related air pollution events.

D. Research Meetings

This semester, I attended our regular Remote Sensing meeting and also attended a course on

Research Methods. We had the opportunity to hear the PhD students present their proposal and

what research they are conducting at ISEIS. In the group meetings, we are able to do a lot of

critical thinking and open discussions on papers that were reviewed as well as how to write

good proposals for funding a research or project.

E. Projects and ongoing research

My final project for summer term will be based on analysis of EL NINO series in Papua New

Guinea using satellite images against statistical data. There could be a possibility to introduce

a new drought index to help meteorological and agricultural drought.

For the last two semesters, I did a total of six group projects and two individual projects. Two

of the projects were focused on areas in Papua New Guinea, the others were on disaster risk

reduction, environmental issues and some policy making decisions. Here is a brief outline on

some of the projects that were carried out for assessment.

1. The spatial analysis of solar power locations applying Geographic Information System and Decision-Making methods: Papua New Guinea

Course - Introduction to Geo-survey and Public Management

Instructor - Dr. WONG C. Janice

The proposal question was how would solar energy contribute to solving the 78% of Papua

New Guinea (PNG) that has no access to electricity? Almost 80% of the total population of

PNG live in rural areas. An analysis in assessing possible site location for solar power stations

in Papua New Guinea by using GIS, meteorological data and also decision-making methods.

This GIS map was generated using the following attributes which are: provincial boundaries,

major road network, water areas and major rivers or lakes.

2. Temporal Analysis of Land use and coastal reclamation in Hong Kong Island

Course - Special Topics in GeoInformation Sciences

Instructor – Prof. LIN Hui & Dr. Ann MAK

Group Members – WANG Yuan & POSANAU Kisolel

Aim & Objectives

� Showing the spatial & temporal change in reclamation of Hong Kong Island. � To identify that most of the reclaimed land was used for development of infrastructure

and compare with population analysis.

3. Typhoon Impact Assessment on Super Typhoon Haiyan in Barangay, Tacloban

city Philippines

Course - Environmental Remote Sensing Technology

Instructor – Prof. ZHANG Hongsheng

Group Members – LAU Chun Yin, PANG Kit & POSANAU Kisolel

Two images of Tacloban city obtained from Google Map on 17th October 2013, which is the

situation before typhoon Haiyan made a landfall on the Philippines and on 18th February 2014,

3 months after typhoon Haiyan made a landfall and struck Tacloban.

The images were processed by ENVI 4.7 in order to investigate the difference in the same area

before and after the catastrophe and later classify different features in the images in order to

find out the number of pixels that corresponds to the feature. We can eventually find out the

difference in the 2 images and find out the devastating effect done by the typhoon.

4. Flood Monitoring in Markham River, Morobe Province in Papua New Guinea.

Course - Geoinformation Technologies for Risk and Crises Management

Instructor – Prof. ZHANG Hongsheng

Group Members – LAU Chun Yin, WONG Yee Tung & POSANAU Kisolel

There are 3 objectives for this project. Firstly, flood and trend of Markham River was analysed.

Secondly, current policy was investigated on flood management and prevention. Thirdly,

insights and suggestions on disaster management in Markham River were discussed. After

taking a brief look at the current situation of monitoring and prevention of disasters in Markham

River, some suggestions were provided in order to aid a better disaster management and

prevention for the government and NGOs to co-operate and local people to act before disaster

happens.

� Analyses of satellites images (LANDSAT 7) on the river size and river width � Images for river area measurements were taken in December � River width was measured by observation and measurement on Google Earth � River size was measured by classification of satellite images in ENVI � Showing different trends with precipitation patterns

� From findings and analyzing the results, there are changes in the river during wet and dry season from 2009 - 2018.

� These changes affect the people who live along the river and also those who travel through this river to their villages (Watut).

� Markham river is the largest river in Morobe Province & prone to floods � This study stressed the need of an effective and reliable flood monitoring system.