Spring thaw classification based on AMSR-E brightness temperature in the central Tibetan Plateau

摘要

Changes in soil freeze/thaw dynamics in the Tibetan Plateau (TP) had significant influences on regional hydrology and ecosystem processes. We developed an algorithm to detect spring thaw onset in the central TP using microwave brightness temperature (T-b) data. We assumed that the T(b)at lower frequencies is more sensitive to changes in soil freeze/thaw status, while the T(b)at higher frequencies is subjected more to scattering effects of snow cover and vegetation. Therefore, the standard deviation of brightness temperature at 6-GHz horizontal polarization was used to detect soil F/T status, and a scattering index based on higher frequencies was used to constrain the scattering effects of snow cover. The algorithm was calibrated and validated with surface ground temperature and daily minimum air temperature. Results showed that our algorithm had a mean bias of 3.7 similar to 17.6 days and a RMSE of 7.5 similar to 19.7 days, and this method behaved better than the previous one based on all-frequency standard deviation of microwave T-b, which had a mean bias of 8.3 V 41.2 days and a RMSE of 13.4 similar to 27.6 days. Further validation is needed over more extensive area with diverse surface conditions.