DEFORMATION MONITORING IN AND AROUND THE NATIONAL CAPITAL REGION OF INDIA USING DINSAR TECHNIQUE

Urbanization increases the pressure on land and ground water. Due to pressure on land and over extraction of the ground water for modernization and development, deformation of land increase. The traditional deformation monitoring techniques are costly and time taking. The Differential Interferometry Technique (DInSAR) has capability to monitor regional scale deformation. Recently Vasundra region of the NCR has experienced a road subsidence near to the multiple stories building. In the study we have attempted to identify the Vasundra and nearby region by the DInSAR technology to identify the deformation in Delhi and NCR focused on the Vasundra area of Ghaziabad, Uttar Pradesh, India. This is one of the developed and planned area in the Ghaziabad District. The Sentinel-1 C data has been used to monitor the deformation in that area. The Vasundra area has been deformed due to heavy rainfall and water logging in the surrounding area by which the road has been subsided up to 30 feet. Using the sentinel data we were able to identify the subsidence in Vasundra and nearby areas. Since the Sentinel scene was very big so a clear picture of the deformation indication in the NCR region are found which need more in-depth study further. Conclusively it can be said that DInSAR technique with Microwave data is a competent tool to monitor the minute deformation.


INTRODUCTION
This research focuses on space borne SAR imagery for monitoring surface deformation in urban area.The technology is fast improving with enhanced synthetic aperture radar (SAR) sensors on different space born (Jiang et al. 2011) platforms, providing high-resolution imagery at any atmospheric conditions, day-night capturing capability and covers large area (Tomiyasu 1978) (Geymen 2014) (Ng et al. 2011).SAR interferometry concept has been introduced in the last 1980 (Gabriel, Goldstein, and Zebker 1989).Interferometry SAR (InSAR) (Aydoner, Maktav, and Alparslan 2004) techniques are widely used to measure the topographic profile (Bürgman, Rosen, and Fielding 2000) and surface deformations (Ishwar and Kumar 2017) (Fulton 2000) (Burgmann, Rosen, and Fielding 2000;Bürgmann, Rosen, and Fielding 2000;Iglesias et al. 2015;Klees and Massonnet 1999;Otir and Komac 2007;Peltier et al. 2010;Qiu et al. 2016;Samieie-Esfahany et al. 2009) and subsidence monitoring (Strozzi et al. 2001).Differential SAR-Interferometry (D-InSAR) is one of the competent technique use to identify the land surface horizontal motion and land subsidence (Ishwar and Kumar 2017) (Strozzi et al. 2001) (Bhattarai et al. 2017;Caló et al. 2017;Champati, Head, and Studies 2001;Chang, Ge, and Rizos 2005) over a large area at low cost than the traditional techniques (Ng et al. 2011) (Ferretti, Prati, and Rocca 2001).The precise leveling, total station and global navigation satellite system (GNSS) are common ground-based traditional measurement techniques which are used to monitor the subsidence at a pin * Corresponding author point location in an area.Though these methods are timeconsuming, require more man power and costly, they can measure height information at millimeter to centimeter levels of accuracy on local area level (Cao et al. 2007).Typically deformation occurs due to resource extraction (States 1991) (Sahu and Lokhande 2015) and less buoyancy to support the external pressure, it occurs at low rates ranging from fraction to as much as 3 inch per year.(Fulton 2000) Subsidence is a certain effect of underground mining (Singh 1992).In the mining subsidence affected surface area is generally larger than the extraction area.The subsidence not only impact on the human also impact their environment (Sahu and Lokhande 2015) due to release of toxic gases (Jiang et al. 2011).

STUDY AREA
The total processed area shown in figure 1 is covered under 28⁰59′22′′ N, 76⁰50′30′′ E, 27⁰59′31′′N and 77⁰39′54′′E.The area under investigation covered under 28⁰14′37′′ N, 76⁰59′27′′ E, 28⁰53′14′′ N and 77⁰31′37′′ E. The study area corresponds to the eastern part of Delhi and National capital region (NCR) of India.The study area includes major developing cities of Delhi and NCR since 1980s such as Noida, Faridabad Guru Gram and Ghaziabad (Figure 3).The major focus of this study has been on urban areas of NCR region of India which has been urbanized years back.Vasundra, Ghaziabad, India has been chosen for validation and ground data collection for deformation Vasundra is a township area since 1990s and well planned and developed.
It was shaped by Uttar Pradesh Housing and Development Authority in the late 1990s.The area map has been showing in figure 1.The area which has been chosen for surface deformation under investigation has been estimated approximate 3,000 sq.km form 6,378 sq.km total processed area (figure 3).In the area, Yamuna and Hindan River flow in the north east part of the area which is under investigation.The Vasundra area situated between both the river and near to Hindan River.

DATA USE AND METHODOLOGY
The two sceans of the C-band, data product of single look complex (SLC) were obtained through the European Space Agenacy (ESA) satellite Sentinel-1 (wavelength of 5.6 and temporal resolution is 12 days and spatial resolution 5x20m) (Bourbigot et al. n.d.)

RESULTS AND DISCUSSION
The results described aim to monitor the surface deformation in the urban and developed region of NCR of India by focusing on Vasundra, Ghaziabad (pin point), India using Cband data of Sentinel-1 displayed in figure 3.
The products derived with the help of the data pair IW SLC Sentinel-1 images, spanning the period from 14 th July 2018 to 26 th July 2018 for surface deformation map.In figure 4 the total area covered derived coherence image with both the polarization with a time interval of 12 days.The VV polarized coherence show good results in the city areas then VH polarized coherence image.In VV coherence image bridges on the river are also easily distinguish which shows good coherence due to volume scattering.In resulted images of coherence showing that coherence varies from low to high with range from 0.6 to 0.79.The values near to 0.6 are showing the low coherence due to presence of water body, vegetation, etc. and a good coherence has been observed near to 0.79 (high coherence due to less temporal resolution) which have good back scatting features such as buildings, bridges, etc.In the residential and industrial areas of Faridabad, Noida, Ghaziabad, Guru Gram and Delhi observed good coherence due to presence of building which have the prominent source of volume and double bounce scattering.In very low coherence, fringes have not been formed very clear due to heavy rain in the area.Where less or low coherence it has been observed due to water lodging in the area.
The consequent SAR interferometry images shown in figure 5 describe the fringes has been formed by the both VV and VH polarization.The location of deformation can be clearly identified but the high phase gradient over a large spatial extent made the phase fringes distinguishable.The fringes are closely to each other can be clearly identified more deformation (in square) in figure 6.The image having good coherence in urban area due to presence of building, increase double bounce scattering which has high reflectivity with a time interval of 12 days.The results show deformation in Delhi and NCR region but Delhi shows less deformation as compare to the NCR.The enclosed area in rectangle in figure   The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-5, 2018 , from14 July 2018 to 26 July 2018.The SAR TOPS (terrain observation with scanSAR ) mode processing has been done on interferometric wide (IW) sub-swath1.The data detailed specification are given below in table.1.The methodology adopted in present study is shown in figure 2. The methodology consists in different steps: a) selection of suitable data pair which is a key step for any SAR processing; b) application of orbit file for precise position and velocity of the satellite at the time of acquiring; c) Coregistration of the images with the orbit and a Shuttle Radar Topography Mission(SRTM) digital elevation model (DEM) of 3 arc secound of same subswath with same polarisation (VV or VH) of the product.d) resampling the slave image into master frame and, equalization, and calibration, e) Interferogram formation of two image by complex congugate of (orbital data processing, interferogram calculation, interferogram flattening, called phase flattening which is corrupted by noise from temporal decorelation, geometric decorelation volumetric scattteing and processing errors, etc.)(Ferretti, Prati, and Rocca n.d.)   of SLC data.The output images such as intensity, phase and cohrence has been examined for interpretation of the deformation in Delhi and NCR region.Terrain correction has been processed for compensate the distortion so that the geometeric representaion of image should be reperesented as close as posible to real world.The identification and interpretation has been done on google earth by analysing and comparing the resulted data with the real world imagenary.

Figure 2 Figure 1
Figure 2 Methodology flow chart

Figure 4
Figure 4 Coherence image generated from two images of NCR area of India, with a time interval of 12 days by (a) VV polarization and (b) VH polarization

Figure 3
Figure 3 coherence image of NCR overlapped on google earth image to show major cities in rectangle and pin point area is Vasundra.

Figure 8 A
Figure 8 A road subsidence of Vasundra Ghaziabad photographs