Analyzing the Inundation Pattern of the Poyang Lake Floodplain by Passive Microwave Data

Abstract

Abstract The soil wetness condition is a useful indicator of inundation hazard in floodplains, such as the Poyang Lake floodplain. Special Sensor Microwave Imager (SSM/I) passive microwave data were used to monitor water-saturated soil and open water areas of the Poyang Lake floodplain from 2001 to 2008, capturing the inundation patterns of this area in space and time. The polarization difference brightness temperature (PDBT) at 37 GHz is sensitive to the water extension even under dense vegetation. The zero-order radiative transfer model was simplified to retrieve the vertical–horizontal (V–H)-polarized effective emissivity difference from the PDBT at 37 GHz. Vegetation fractional area and vegetation transmission function were derived from NDVI to represent the vegetation attenuation. This effective emissivity difference has a quasi-linear relationship with the fractional area of water-saturated soil and standing water, no matter the frequency. Using the multifrequency-polarization surface emission (Qp) model and the Dobson model of the soil–water mixture, the two segments of this relationship were combined into a quasi-linear model. Comparing the retrieved water-saturated soil and standing water area of Poyang Lake with the lake area obtained from the MODIS and synthetic aperture radar (SAR) image at higher spatial resolution, the calculations show a good fit with the MODIS and SAR data, with R2 = 0.7664 and relative RMSE = 17.74%. The cross-correlation analysis shows that the Poyang Lake extension fluctuates with a 5-day time lag with the upstream land area of water-saturated soil and standing water. Since the closure of the Three Gorges Dam, this relationship is more evident.