Agriculture

These images are the false color composites showing the land surface observed by the Precipitation Radar (PR) aboard the TRMM satellite for February (above) and August (below) 1998. PR observes backscatters from rain and Earth’s surface. It scans 17 degrees across the track. Since the backscatter from 3 to 8 incident angles is highly affected by the absorption (scattering) by vegetation canopy over the land, weaker backscatter corresponds to dense vegetation. Therefore, dark green is used for low backscattering areas to express more dense vegetation. Since more reflectance is observed from a wetter surface at incident angles of 9 to 13 degrees, blue corresponds to stronger backscatter. At 14 to 18 degrees, weaker signals are from smoother surfaces such as deserts with mirror-like surfaces. Dark red is used for weaker backscattering.
Thus in these false color composites,
　+Green areas correspond to the areas with dense vegetation,
　+Blue corresponds to wet surfaces,
　+Red represents arid and desert area, and
　+Yellow represents dry grasslands (opposite to blue).
In dry areas, the color may change from green, yellow or orange to red, according to the vegetation amount. If the area is wet enough, blue may be mixed and change to purple (less vegetation) or cyan (more vegetation). These color changes are shown schematically as a triangle diagram in the figures.
Comparing the false color composites for February and August, the distribution of tropical rain forests in the Amazon river basin in South America, in the Asia Monsoon region, and in the Congo river basin in Africa is clearly seen. The increase and decrease of wet (blue) and vegetated (green ) regions according to the rainy season are apparent as well.
Red, which corresponds to deserts and areas, is seen in the Sahara desert in Northern Africa, and in the Rub’al Khali desert in the Arabian Peninsula, The Great Victorian desert in central Australia indicates missing PR observation data due to a frequency conflict. Areas thought to be water surface from PR observation, such as oceans, big lakes and wide rivers, are shown in black.
　Courtesy of Oki Lab. (IIS, University of Tokyo) in cooperation with JAXA

Backscatter data from the Precipitation Radar (PR) aboard the Tropical Rain Measuring Mission (TRMM) satellite includes not only information on soil wetness but also on the vegetation amount and land surface roughness. Using the information of NDVI from visible and infrared sensors and the theory of microwave scattering, the effects from vegetation and roughness were first subtracted, then soil wetness was estimated.
The upper panel is the estimated soil wetness for February 1998, and the lower panel is for August 1998.
Comparing the estimates for February and August, the Amazon River Basin in South America is wet in February and the rainy season and dries up in August, corresponding to the dry season. The Orinoco River Basin, which is adjacent to north of the Amazon, behaves in the opposite manner. Wetting in the Asian Monsoon region is dominant as well. In the false color composite maps are classified as arid regions patterns and recognized in the desert area. However, the Sahara desert and the Rub’ al Khali desert are classified as arid regions in the quantitatively estimated soil wetness maps. Soil wetness can not be estimated by this algorithm some mountainous areas where the scattering theory of dependency of incident angle cannot be applied due to the effects of steep slopes and in some tropical rainforest area where the forest is extremely dense. Such regions are shown in black missing data.
The black region in the Northwestern part of Australia indicates missing PR observation data due to a frequency conflict.
　Courtesy of Oki Lab. (IIS, University of Tokyo) in cooperation with JAXA