The influence of soil condition in remote sensing data have not been clearly explained. Therefore, it is important to study the relationship between ground survey data and remote sensing data. Ground survey on soil condition such as topography, soil moisture, soil color and soil particle size were carried out in Mandal-GOBI, Mongolia on every August 1998 to 2002. The test area was covered by thin grass, the coverage ranges from 12% to 28%.
Topographic survey was used kinematic GPS. Soil moisture was measured in 3cm depth by four-pin prove sensor. The measurement on ground surface was very difficult. Measured soil moisture ratio indicated 0.05 to 0.20 and topography was no related to the soil moisture. Soil color was measured by HSI value in digital camera images. The result showed soil color was almost same condition. Therefore, soil moisture and soil color was disregard in this study. Soil particle size was also measured by the digital camera images. Visual interpretation of the images was carried out to classify five categories of soil particle size.
In this study, TERRA ASTER (Advanced Spaceborne Thermal Emission and Reflection Radiometer) image was compared with soil particle size. Firstly, ASTER image was geometrically rectified with topographic data by image matching. And topographic influence of the ASTER image was removed by logarithmic residuals method. The digital number in band 7 and band 8 had tendency to drop with large soil particle size. The digital number in Band 9 had tendency to rise with large soil particle size. Those tendencies were noticeable results in vegetated area. Therefore, vegetation condition should be discussed to understand the results.
抄録全体を表示