TIME SERIES ANALYSIS OF SUBSIDENCE IN DHAKA CITY, BANGLADESH USING INSAR

Abstract

Despite existing literature suggesting that Dhaka City, Bangladesh is undergoing subsidence, few researches have been carried out to actually measure the subsidence rate. Previously conducted studies either do not provide sufficiently accurate subsidence results, or the study period is not long enough. In this research, we have tried to address that gap by performing time series subsidence analysis of Dhaka City utilizing Interferometric Synthetic Aperture Radar (InSAR) technique for a study period of 20 years. Synthetic Aperture Radar (SAR) C-band images from ERS, ENVISAT and Sentinel-1A were used to obtain the results. We had to use C-band SAR data from multiple sensors considering data availability issue of the period of investigation (i.e. 1992 to 1999(using ERS); 2003 to 2010(using ENVISAT); 2014 to 2017(using Sentinel 1A)). Most parts of the city is found to be subsiding. Mirpur and Uttara have subsided by over 221mm and 232mm respectively over the 20 years. Ramna and Cantonment subsided ar ound 205mm compared to their level in 1992, whereas both Gulshan and Tejgaon have subsided by about 200mm. Demra and Lalbagh show similar subsidence to the Ramna area, whereas Dhanmondi and Mohammadpur show subsidence rates similar to Tejgaon. We have also assessed the parameter sensitivity to perform this time series subsidence analysis. The parameter selection of coregistration, filtering and unwrapping was found to greatly influence the results. The result is being calibrated with the available GPS observation.

Keywords

• Subsidence
• InSAR
• Time Series Analysis