The use of multi-temporal NOAA/AVHRR data to monitor surface moisture status in the Huaihe River Basin, China

Abstract

Main part of the Huaihe River Basin, China, is an extensive flat plain between Yellow River and Yangzi River and there exists a steep climatological gradient from wet southeast to dry northwest region. The divide between paddy field zone and cropland runs through the basin. These characteristics facilitate the verification of satellite remote sensing to extract hydrological information. Runoff, precipitation and pan evaporation data in the several sub-basins are used to calculate water balance in each sub-basin. Annual runoff( Q), pan evaporation( Epan) and water loss( P- Q) are plotted against the annual average precipitation( P). Average annual evaporation( Ea) as expressed by annual water loss( P- Q) shows relatively small variations among years except northwestern part of the Huaihe Basin, and it ranges from 600 mm/year in northwestern part to 800 mm/year in southeastern part of the basin. Steady Ea suggests that there is enough water to evaporate under the given climatological conditions. Annual precipitation decreases from southeast to northwest. In the northwestern sub-basin(DW), Ea decreases with decreasing P when P falls less than steady Ea. It means that evaporation is suppressed by water deficit. This suppression should be detectable by the satellite remote sensing. Multi-temporal NOAA/AVHRR GAC datasets are used to estimate surface moisture status by combining surface temperature(Ts) and vegetation index(VI). Modified soil adjusted vegetation index(MSAVI) is used as VI for this study to minimize the soil background effect. Eighty scenes of GAC from DOY(Day of the Year) 198 to DOY 277 in 1990 are used to extract the slope values in Ts/VI relationship. The time series of the slope values correspond well to that of antecedent precipitation index(API). The spatial distribution patterns of the slope values are also consistent with those of the API. Although API is not a proper indicator of surface wetness, the results indicate the usefulness of the Ts/VIs relationship to estimate surface moisture status over extensive area.