Better Understanding of Land Surface Emissivity Map from passive microwave data

Land emissivity is a crucial boundary condition when passive microwave data is used in Numerical Weather Prediction (NWP) models or when is used to retrieve the soil moisture. According several investigations, a high accuracy of less than one percent in the retrieved emissivity is required to minimize the impact of cloud liquid water in retrieving air temperature, water vapor and surface temperature. The objective of this study is to develop global land emissivity products using AMSR-E passive microwave data and to investigate the effect of penetration depth in different areas especially in deserts. This issue will be more highlighted when the amplitude of diurnal cycle of skin temperature is higher. Additional ancillary data such as water vapor amount and cloud mask will be used to assess the atmospheric contribution and to identify clear sky conditions. For this study, several global satellite data as well as regional airborne and ground based measurements were collected and processed. All AMSR-E frequencies (6.9, 10.7, 18.7, 36.5 and 89.0 GHz) have been used to retrieve the emissivity. Skin temperature and cloud mask product (DX) produced by the International Satellite Cloud Climatology Project (ISCCP) was used in this study. Moreover, water vapor information produced by ISCCP (TOVS data) was used to calculate the upwelling and the downwelling temperatures as well as atmospheric transmission. Several land emissivity maps have been developed at different AMSR-E frequencies. The analysis of the obtained maps has shown an acceptable agreement with the global pattern of land use/land cover conditions. At this stage of the project the developed emissivity maps will be validated qualitatively by comparing the emissivity maps at the AMSR-E frequencies to the previous study using SSM/I data. Comparing the brightness temperatures between SSM/I and AMSR- has shownsome calibrating issues, since there is slightly difference between their incident angles. The reason for this problem still is under investigation. The differences between daytime and nighttime emissivity as well as differences in retrieved emissivities from different frequencies have been studied. Additionally, the sensitivities of retrieval to using different available datasets such as MODIS-LST for skin temperature and NCEP for atmospheric correction parameters will be studied. Potential of extrapolating the L band emissivity to the AMSR-E frequencies is being investigated in this study.