An evaluation of C- (developed at SAC) and X- band (data of opportunity) Airborne SAR (ASAR) data was done for flood signatures. The data was collected over parts of a perennially flood affected area of Darbhanga district located in the Bihar State. This activity is carried out as an R&D effort under the Disaster Management Support Programme (DMSP) at SAC. The ASAR data was critically analyzed for identifying flood features and with an aim to develop it finally as a system dedicated for DMS. The need for such a system is important as many a times there are gaps in satellite coverage over tropical regions for flood damage assessment besides it was observed that the varying flight directions provided additional information as regards the flood signatures. Both the airborne data were available in near synchronous time frame and with VV polarization. A comparison of C- and X- band ASAR showed that the inundated area was estimated better in X- band (62.8 %) as compared to C- (50.2 %) and that the combined data set gave further improvement (71.0 %).
Terrestrial vegetation exhibits strong seasonal and inter-annual fluctuations, which are associated with changes in the
environmental variables. The understanding of the vegetation dynamics on a long-term basis requires repetitive
observations over many years at temporal and spatial scale compatible with the regional processes. Space-borne
scatterometer though primarily designed for ocean study, have proved advantageous in studying vegetation dynamics at
regional scale. This could be possible due to availability of high repeat cycle data in addition to availability of processed
scatterometer data by SIR algorithm. In the present study, the temporal behavior of Indian tropical-moist-deciduous
forests has been studied using QuikScat (Ku-band) scatterometer data of year 2000. The impact of vegetation growth and
senescence was investigated by comparing σ0 time-series to NDVI and rainfall observations. It was observed that σ0 varies with different scale than the NDVI. The low correlation coefficients also indicate that, the direct relationship
between σ0 and NDVI is either very week or absent. Therefore, inferring that NDVI, which is a measure of the
vegetation greenness or vigor rather than plant water-content or height, is not important for the explanation of the
backscattering behavior. Further, σ0 was minimum during the deciduous period with prevailing high temperature and no
rainfall condition. The temporal changes in backscattering coefficient were modeled for understanding the vegetation
dynamics. The study suggests the suitability of Ku-band space-borne scatterometer data for understanding the seasonal
dynamics of different forest types.
Conventional imaging radars operating with a single polarization antenna measures only the amplitude of the returned
signal. Therefore, accurate discrimination between similar scattering returns is difficult and, depending on the system's
polarization configuration, many features go undetected. Recent developments in radar technology have led to the
development of imaging radar polarimeters, that is, radars that are capable of imaging the Earth's surface at any and all
possible polarizations through antenna synthesis techniques. The capability enables the complete measurement of a
target's polarization properties, thus permitting a much more detailed understanding of the electromagnetic scattering
process. The work presented in this paper describes the concept of polarimetric signatures, polarimetric scattering matrix
and simple theoretical signatures from various models. The Stokes' matrix for individual class of pixels was generated
using the amplitude and relative phase of the scattered wave. Based on the Stokes' matrices of the full polarimetric Lband
airborne SAR data, polarization signatures were extracted and analyzed for four land cover classes: urban area,
forest, agriculture fallow and water body. These signatures were compared with the theoretical polarization signatures
and the scattering mechanisms were studied. The scattering mechanism of these land cover classes was also analysed
based on the images generated by Pauli as well as Freeman-Durden target decomposition.
This study describes the use of SAR interferometric technique, coherence technique and backscattering coefficient data
for the estimation of canopy height in forest area. The study area is Tundi Reserved Forest falling in Dhanbad district of
Jharkhand state in India. Most predominant species is Shorea robusta (Sal) constituting the top storey of the forest with
maximum height up to 20 m. For the purpose of validation, Ground Truth locations were taken on the basis of variability
in the area. At ground truth location, sample plots were selected for the measurement of stand heights for the forest
stands of different age and height categories. Interferometric SAR technique was used to generate Digital Elevation
Model (DEM) in terms of Top of the Canopy Digital Surface Model (TCDSM). For this, Radarsat SAR interferometric
data pair of 12 February and 7 March, 2004 were selected. The baseline for this pair was 643 m. The TCDSM at ground
truth location was subtracted from the DEM obtained at clear-cut areas. The value of tree height was 8.1 m as against the
measured value of 9m. The two date SLC data set (17 April and 22 May, 2004) of Envisat- ASAR were also used for the
study of coherence pattern and its relation to forest height. Averages of coherence values for each of the GT sites with
stand heights as well as ground height were derived using 5×5 pixel windows. The coherence values with the ground
measured stand heights showed a high correlation. Coherence has been shown to be efficient for establishment of forest
stand height in a forest environment. Thus, this paper demonstrates that SAR derived TCDSM and coherence can be
used as a useful tool along with intensity data for the measurement of tree parameters.
Multi-frequency SAR observation over forested areas has been the subject of research owing to the
frequency dependence on the contribution of radar backscatter from different parts of the vegetation
canopy. The multi-frequency SAR data at P-, L-, C- and X-band was acquired over Rajpipla site (Gujarat).
All the channels were in quad-pole mode except X-SAR, which was in HH and VV-polarization mode.
The study area comprises of the dry and moist deciduous forest. The moist deciduous forest is not
evergreen and shed their leaves during March-April. Teak (Tectona grandis) is the dominant species in the
moist deciduous forest area. Dry deciduous trees are mainly khakhar (Butea monosperma). Multifrequency
SAR data was processed to get geo-referenced images and all the images were co-registered.
Images were converted to the backscattering image using the calibration function. For the purpose of
ground verification, ground data was obtained at different locations. Ground data consisted of
measurements on tree height, diameter at breast height, basal girth, crown diameter etc. for each location;
measurements were done in 10m by 10 m area. The analysis of the data was carried out in relation to
comparison of backscattering coefficient in different frequencies and polarizations. In general, forest type
was better seen in X-band image as compared to other classes. However, X- and C-band could be
comparable in terms of forest classes. Further, backscattering coefficient increases with frequency except in
P-band. However, P-band showed best correlation with biomass as compared to other channels.
Conference Committee Involvement (1)
Microwave Remote Sensing of the Atmosphere and Environment V