Science Service System
You are here : Home : Proposals_Summary

Summary of Proposal HYD0480

TitleAssessment of high spatial resolution SAR data for flood impact on small water bodies: a test of TerraSAR-X data in Bangladesh
Investigator Al-Ali, Mohamed - Durham University, Geography Department
Team Member
Professor Donoghue, Daniel - Durham University, Geography Department
Professor Petley, David - Durham University, Geography Department
PhD Student Al-Ali, Mohamed - Durham University, Geography Department
Professor Huda, Khondaker - Jahangirnagar, Department
SummaryBangladesh is a low-lying flood-prone deltaic plain. The traditional method of creating raised land for safe flood-free homesteads in rural areas is by excavation. This creates, as an important by-product, artificial water bodies that are classified by the local people according to size and function. These are distributed almost equally all over the country. For every twenty people there is approximately a hectare of small water bodies, which range in size from 25 to over 2000 sq. m. These small water bodies form an integral part of the rural economy in Bangladesh because they are used for drinking water, bathing and washing, fisheries and aquaculture, duck raising, irrigation, cattle feeding and washing. Despite the importance of small water bodies to the Bangladesh economy, there is no country-wide, up-to-date inventory of this important resource that is very vulnerable to flooding. Attempts to map water bodies using optical imagery have met with some success (Huda 2004) but cloud cover and poor seasonal coverage severely limit the ability to detect the smallest water bodies because of the lack of image data during the wet season. It is expected that high spatial resolution SAR data could provide the key to identifying the different types of small water body all year round as well separating those from other seasonal large water bodies and flooded areas. This project will evaluate TerraSAR-X data products to establish the most appropriate system characteristics (spatial resolution, polarization, looks), image processing methods (geocorrection, filtering), and ground condition (soil roughness, wetness, vegetation cover, wind speed) for mapping small water bodies in Bangladesh. The experiment will test feature detection methods for a range of differently sized and shaped water bodies, including bodies surrounded or partially obscured by sparse tree cover. A second objective will be to test the ability of the SAR data to detect seasonal changes in the extent of water bodies. This would provide an important new source of information on wetland dynamics that would add considerable value to map products for resource management. Development agencies report that they need (a) information on the number, distribution, size, dimensions and associations of small water bodies and (b) a better understanding of the correlation between the distribution of the small water bodies (which is thought to be a critical variable) and the existence and type of aquaculture at a particular location , so that (c) working with local governments and rural communities, they are better able to undertake appropriate socio-economic research and thence plan aquaculture expansion or other management issues. Small water bodies, in descending size, include canals, beels, communal village pond and ditches, and these can vary in size from 4,000-5,000 m2 to as little as 40 m2. This proposal aims to use TerraSAR imagery to help establish a robust method for mapping spatial distribution of SWB and monitoring the impact of flood events. The project will build upon a large ground inventory already completed in four mouzas of Shahjadpur Thana. The data sources available to the project include a time series of aerial photographs and satellite images, a ground survey of over 300 water bodies, a GIS containing GPS outlines of water bodies and field photographs. It is hoped that suitably processed TerraSAR data will enhance the capability to map the smallest water bodies more accurately, as well as provide information of the water dynamics in the larger areas. It is hoped that this application project will raise awareness of the potential of TerraSAR data by demonstrating a method for producing a rapid assessment of freshwater and wetland resources that can easily be integrated into local and national planning processes that could be applied widely throughout south Asia.

Back to list of proposals

DLR 2004-2016