Measurement and Simulation of Topographic Effects on Passive Microwave Remote Sensing over Mountain Areas: A Case Study from the Tibetan Plateau.
Xinxin Li,Lixin Zhang,Lutz Weihermueller,Lingmei Jiang,Harry Vereecken
DOI: https://doi.org/10.1109/tgrs.2013.2251887
IF: 8.2
2013-01-01
IEEE Transactions on Geoscience and Remote Sensing
Abstract:Knowledge about the surface soil water content is essential because it controls the surface water dynamics and land-atmosphere interaction. In high mountain areas in particular, soil surface water content controls infiltration and flood events. Although satellite-derived surface soil moisture data from passive microwave sensors are readily available for most regions globally, mountainous areas are often excluded from these data (or at least flagged as biased) due to the strong topographic influence on the retrieved signal. Even though a substantial volume of literature is available dealing with topographic effects on spaceborne brightness temperature, no systematic analysis has been reported. Therefore, we present a comprehensive analysis of topographic effects on brightness temperature at C-band using a two-step approach. First, a well-controlled field experiment is carried out using a mobile truck-mounted C-band radiometer to analyze the impact of geometric and adjacent effects on the radiometer signal. Additionally, a comprehensive radiative transfer model is developed accounting for both effects and tested on the ground-based data. Second, recorded Advanced Microwave Scanning Radiometer-Earth Observing System (AMSR-E) data over the Tibetan Plateau were used to analyze the error due to the impact of topography using the developed model. The results of the field experiment clearly show that the geometric effect of a single hill has a much larger impact on brightness temperature compared to the adjacent effect of multiple hills, whereby, due to the geometric effect, the bias is up to +20 K for horizontal and -13 K for vertical polarization. For the adjacent effect, the bias is less than 3 K for both polarizations. Additionally, the developed radio transfer model was able to reproduce both effects with high accuracy. For the AMSR-E data, the model shows that the brightness temperature recorded is biased in the same way as the ground-based measurements and that uncertainties induced by the wide existence of atypical mountain regions in the Tibetan Plateau will have a great impact on the retrieving error (maximum 30%). The largest impact on the retrieval error, on the other hand, is calculated for the soil moisture with a maximum relative error of 44%. The negligible impact can be attributed to false parameterization of the soil texture, soil surface temperature, and sky temperature. Finally, the overall absolute error in the estimated water content is quantified on average with 4%, whereby single pixels indicate a maximum absolute error of up to 16%. In conclusion, we show that recorded spaceborne brightness temperatures are highly biased by topographic effects in mountainous regions using a comprehensive radiative transfer model. Additionally, we suggest using this model to invert the effective surface emissivity of mountain areas for standard processing of higher level data products such as surface soil water content.