K. S. Rao

A study on the applicability of repeat‐pass SAR interferometry for generating DEMs over several Indian test sites

Repeat‐pass synthetic aperture radar interferometry (InSAR) appears to be an excellent tool for generating digital elevation models (DEMs) of high spatial and vertical resolution. However, in recent years, it has been noticed that the atmosphere causes errors in the elevation accuracy. The objective of this paper is to try to identify climatic regions in India where InSAR can be used for generating DEMs. A detailed study was carried out to quantify atmospheric effects on the DEMs derived from repeat‐pass InSAR. Three test sites were selected for the analysis: one over Mumbai, which is a part of the western ghats region with hilly terrain, lakes and forests; another is the Kolar area with gently undulating terrain with agricultural and forested lands; and the last was Bhuj, an arid plains region which is close to the Rajasthan Thar desert, India. Six interferograms of the Bhuj area derived from two tandem pairs were analysed and it was found that there were varying atmospheric effects of the order of 0.2λ to 1λ and at specific locations more than λ. In the case of the Mumbai and Kolar study areas with two interferograms each, the atmospheric effects were almost negligible. In these cases, it is shown that it was also possible to estimate the height of buildings to an accuracy of 1–2 m. Therefore, it is concluded that one should be cautious when using InSAR techniques for generating DEMs of desert/arid regions where the climate may be very dynamic and only small amounts of water vapour can cause strong atmospheric artefacts.

A model for the retrieval and monitoring of soil moisture over desert area of Kuwait

A simple statistical inversion model is developed for the retrieval of soil moisture, based on two frequencies. The model is applicable to desert semi‐arid areas such as Kuwait, where the surface is covered with scant vegetation and moderate roughness. In this study, Nimbus‐7 SMMR brightness temperature (T b) data at two bands (T b(H) of 6.6 GHz and T b(V) of 37 GHz) acquired during the six‐year period (1979–1985) are used for the retrieval and monitoring of soil moisture. The model parameters are fine‐tuned with field data. The accuracy of estimation of soil moisture is of the order of 3%. The soil moisture varies from 1% (in dry summer) to 13% (during the rainy period), with an average of 4%.

Assessment of geo-coding and height accuracy of the dem derived from preliminary data sets of x-band srtm

The first data set of X-band Digital Elevation Model (DEM), derived from Shuttle Radar Topography Mission (SRTM), of Gujarat Earthquake affected area was processed to assess the geo-coding and height accuracy. A detailed survey of the study area was carried out using Differential-GPS systems to get accurate control points. Though, it is the first data set processed without Attitude and Orbit Determination Avionics, the geo-coding and height accuracy is good. It is observed that there is an average shift of-1.5 pixels in Latitude and 2.5 pixels in Longitude with 1.5 pixels Standard Deviation (STD). A comparison of SRTM and Differential-GPS heights indicates that there is a bias of 7.4 m with STD of 3.4 m. This is in agreement with the height errors reported by the Deutschen Zentrums fur Luft- und Raumfahrt (DLR) in Height Error Map (HEM) data set.

Study of dems derived from ERS-1/2 SAR and SRTM data

The DEM of the Bhuj earthquake affected area of 50 x 50 km was generated using the ERS-1/2 SAR tandem data (May 15—16,1996). Region growing algorithm coupled with path following approach was used for phase unwrapping. Phase to height conversion was done using D-GPS control points. Geocoding was done using GAMMA software. A sample data of DEM of Shuttle Radar Topography Mission (SRTM) of the Bhuj area is made available by DLR Germany. The intensity image, DEM and Error map are well registered. The spatial resolution of this DEM is about 25 m with height accuracy of a few meters. The DEM derived through ERS SAR data is prone to atmospheric affects as the required two images are acquired in different timings where as SRTM acquired the two images simultaneously. An RMS height error of 12.06 m is observed with reference to SRTM though some of the individual locations differ by as much as 35 m.

Observation of atmospheric effects on repeat-pass interferometric synthetic aperture radar over the Kuwait desert

Twenty-eight advanced synthetic aperture radar (ASAR) scenes from the Environmental Satellite (ENVISAT) are analysed to select suitable pairs for generating a digital elevation model (DEM) and displacement maps. For this purpose, the repeat-pass interferometric synthetic aperture radar (InSAR) technique is implemented using GAMMA interferometric modules. The perpendicular component of baseline (B┴) is taken as the criteria for selecting the pairs: 0 < B┴ <100 m for displacement maps and 200 < B┴ < 400 m for the DEM. Though there are many pairs satisfying the above criteria, only four case studies are presented here to illustrate the effects of atmosphere on the DEM and displacement maps over the Kuwait desert climate. In each case study, two examples are selected: one where the atmosphere is a serious problem and another example the atmosphere has no significant problem. The DEM of the Shuttle Radar Topographic Mission (SRTM) is taken as a reference for root mean square (RMS) error estimation in the DEM. The RMS error varies from as low as 2 m to as high as 40 m. Some DEMs showed fringe-like structures resembling precipitable water vapour (PWV) fields. Similarly, the measured displacement values were found to vary randomly from place to place and time to time. The displacement maps showed vertical structures similar to PWV. The DEM was corrected for PWV. The results are encouraging. From this study, it is clear that, even for desert areas, there is a need to look into the effects of PWV on the DEM and displacement maps before the results are used.

Analysis of ASAR data for geo-location accuracy and desert signatures

The utility of ASAR data will be greately enhanced if the Radiometric and Geometric quality of the data satisfies the requirements of applications. 7 secenes of ASAR data acquired during April 2004 - June 2005 have been processed to assess the Geo-location accuracy and also temporal / spatial variability of backscattered signatures. Ground Crontrol Points (GPS) approach was used for Geometric accuracy assessment. 80 GCPs spread all through the image were measured using Trimble 5700. It is estimated that the geometric accuracy of the data is within 55 m. This is in agreement with the reported accuracies in the literature. About 20 well defined targets were considered for the backscattering signature study. The size of each target varies from 1000 to 5000 SLC pixels for good statistical stability of the signature. It has been estimated that the Standard Deviation (STD) of the sitgnature is 1.3% and the temporal / spatial variability of the signatures ( considering all varieties of targets such as built-up area, grass lands, airports, agricultureal plots, desert soils, etc) are within its STD. During this study period the soil moisture varied from 10% to 6% as recorded from AMSR-E on board Acqua satellite.The ASAR data has been filtered for speckle noise.

An assessment of brightness temperature data quality of MSMR of IRS‐P4 satellite

Frequency dependence of brightness temperature (T B) in the range 6–21 GHz is studied as a function of surface roughness, vegetation cover and soil moisture using theoretical models. Frequency dependence of T B is plotted from Multi‐channel Scanning Microwave Radiometer (MSMR) T B data of Indian Remote Sensing Satellite (P‐series) IRS‐P4 for different target types such as ocean, desert, forest, agricultural and snow regions. MSMR responds to the variations of the target parameters; however, it is noticed that the T B data at 18 GHz deviate from the theoretically predicted trend. The effect of atmosphere (water vapour) is taken into account in the theoretical trend. However, the frequency dependence of T B plots from Scanning Multi‐channel Microwave Radiometer (SMMR) of Nimbus‐7 closely resemble the theoretical trend. This indicates that the relative calibration of 18 GHz T B of IRS‐P4 has some problem and needs to be looked into. Approximate offset values for each land type are computed and presented here. A general comparison of T B of MSMR and SMMR indicates that T B of MSMR is higher by 4 K, which needs to be further investigated.

Time Series Analysis of the Digital Elevation Model of Kuwait Derived from Synthetic Aperture Radar Interferometry

The digital elevation model derived from SAR Interferometry is prone to atmospheric, penetration into soil medium, system noise and decorrelation errors. Eight ASAR images are selected for this study which have unique data set forming 7 InSAR pairs with single master image. It is expected that all the DEMs should have the same elevation values spatially with in the noise limits. However, they differ very much with one another beyond the noise levels indicating the effects of atmosphere and other disturbances. The 7 DEMs are compared with the DEM of SRTM for the estimation of errors. The spatial and temporal distribution of errors in DEM are analyzed by considering several case studies.

Time series analysis of SAR interferometry derived DEM of Kuwait desert to reduce the affects of atmosphere

Twenty five ASAR scenes are analyzed to find the suitable pairs for generating DEM of Kuwait desert area. About 40 pairs are suitable for generating DEMs. A unique combination of seven DEM are possible with a single master image whose perpendicular baseline component varies between 233 to 393m. GAMMA inerferometric package coupled with ERDAS image processing package are used in the analysis. The seven DEMs are compared with the 90 m DEM derived from SRTM. It has been found that the RMS error varies from 1.9 m to 15.3 m. The highest RMS error refers to day-difference of 525 days with average coherence value of 0.71 (lowest of all). Therefore, correlation will be the cause of higher errors. However, 35 days day-difference pair gave RMS error of 5.5 m with highest coherence value (0.93) which is supposed to produce the lowest RMS error. For the study of spatial distribution of errors, Interferometric DEMs are subtracted from DEM of SRTM. From this analysis, it is observed that the atmospheric affects are aligned and varies systematically. It can give an error in elevation as high as 15 m. Therefore, one should be very careful in using SAR interfoermetry technology for DEM generation even for desert regions.