Exploring the use of MODIS forest transmissivity for correcting passive microwave observation of snow-covered terrain/landscape
MetadataShow full item record
Snow is one of the most important parts in Earth’s hydrologic cycle especially at high latitude. Observation of snow accumulation by passive microwave measurements is an effective way for estimating snow mass at the regional to hemispheric scales because microwaves have the capability of interacting with the snow and the amount of interaction is controlled by the bulk properties of a snowpack such as snow water equivalent (SWE) or snow depth (SD). Compared with optical approaches, microwave observations can be made under nearly all weather and lighting conditions. Forest coverage is one of the challenges in the estimation of snow accumulation by passive microwave observations. Canopy can decrease the accuracy of SWE retrieval by attenuating microwave emission from ground and by producing additional emission. Because forest is one of the major land cover types, retrieval SWE in forested domains is the key challenge in the estimation of snow properties with passive microwave. Transmissivity of radiation is an important variable that describes how a tree canopy attenuates microwave emission from ground. If this variable is known, inverse microwave emission retrieval schemes can provide reasonable estimates for SWE in forest area. Although transmissivity can be measured in the field or retrieved by model which based on field data, field data is not always available especially at regional to global scales. Therefore, following the work of Metsamaki et al. (2005) a transmissivity model driven reflectance data from the Moderate Resolution Imaging Spectroradiometer (MODIS) been applied for retrieval transmissivity in this study. Because SCAmod need reflectance from snow covered condition to drive, it only can be applied in high the latitude area. MOD44B data were used to extend this transmissivity data to lower latitude area because MOD44B data and transmissivity data are highly correlated. The vegetation's influence on PM brightness temperature were explored by compare the PM brightness temperature at open area with forest covered area. In general, the brightness temperature contributed by the vegetation increases with the increase of forest vegetation density. In the higher frequency bands, vegetation tends to contribute more brightness temperature than lower frequency bands. This finding can be used to solve SWE or SD underestimate in the forest region.