HR: 0800h
AN: H41D-0424 [Abstracts]
TI: Effect of Sub-Pixel Variability of Land-Cover on Soil Moisture Retrieval using SAR Data
AU: * Lakhankar, T
EM: tarendra@ce.ccny.cuny.edu
AF: NOAA-CREST, The City University of New York, New York, NY 10031
United States
AU: Ghedira, H
EM: ghedira@ce.ccny.cuny.edu
AF: NOAA-CREST, The City University of New York, New York, NY 10031
United States
AU: Khanbilvardi, R
EM: khanbilvardi@ccny.cuny.edu
AF: NOAA-CREST, The City University of New York, New York, NY 10031
United States
AB:
The soil moisture response to microwave systems from ground surface is mostly influenced by parameters such as: land cover,
and vegetation density, soil texture; which make the retrieval process more complex. In this research, we have used a
back-propagation neural network to retrieve the surface soil moisture from Synthetic Aperture Radar (SAR) data acquired by
RADARSAT-1 satellite. The soil moisture data measured by Electronically Scanned Thinned Array Radiometer during the SGP97
campaign were used as truth data in the training and the validation processes. The sub-pixel variability of land-cover and
vegetation on soil moisture retrieval has been investigated through a sensitivity analysis of spatial variability of soil
moisture.
All the 800 m resolution pixels of land-cover have been labeled as either homogeneous or heterogeneous based on the
percentage of most commonly occurring land cover class of sub-pixels of 25m resolution. The results showed that
homogeneous pixels are more likely to have better accuracy than heterogeneous pixels in soil moisture retrieval. Further,
the illustration of sub-pixel variability of land-cover and vegetation (Normalized Difference Vegetation Index and Vegetation
optical depth) was investigated through spatial structure of soil moisture classes.
Confusion matrices and Kappa Coefficients calculated from independent datasets have been used to evaluate the accuracy of the
retrieved soil moisture. A better correlation between soil moisture and SAR backscattering was found in areas with high
soil moisture content. The modeling results have shown that the retrieval of soil moisture in highly vegetated areas was
less accurate than bare soil areas. Further, the same results have shown that the additions of vegetation optical depth and
Normalized Difference Vegetation Index as additional Neural Network input with the SAR data had a significant effect on the
overall soil moisture classification accuracy.
DE: 1855 Remote sensing (1640)
DE: 1866 Soil moisture
SC: Hydrology [H]
MN: Fall Meeting 2005