(1) School of Environmental Science, The University of Shiga Prefecture
Quantitative evaluation of precipitation in the Tibetan Plateau is quite important to understand the land-surface interaction processes and to estimate atmospheric latent heat. Radar or satellite estimates are fundamentally expected in such remote areas. To improve the algorithms and validation of estimates, we need to observe the characteristics of sub-grid scale precipitation distribution, which is closely related to the meso-scale structures of clouds and convections.
GMS/IR satellite data shows remarkable diurnal changes of Tbb in the monsoon season over the plateau, which reaches below 220K in the evening. Such convections are especially dominant in the southeastern mountainous plateau, and provide diurnal changes of precipitation distribution affected by the large scale topography, such as precipitation of the daytime in the mountain and nighttime in the valley. GAME/Tibet project conducted intensive precipitation observation at Naqu basin, south of Tanggula mountains in the central plateau. In such plain area, diurnal change of precipitation amount was not obvious, and precipitation intensity was not so strong as expected from the low Tbb. Rather, every dayfs frequent and weak precipitation provided more than 400mm amount of monsoon precipitation (Ueno et al., 2001). Diurnal change of hourly precipitation intensity was also found. Radar observation showed passing isolated convective echoes with a small scale provided weak precipitation intensity in the daytime, and stagnant widespread stratiform echoes provided rather strong intensity in the nighttime.
Experimental formula was constructed by adjusting the cloud amount or number of convections calculated from GMS/IR to the surface measurements of precipitation, and one-degree daily scale precipitation distribution was estimated over the plateau. But, such formula miss-estimated the precipitation from the clouds without the true precipitation in the pre-monsoon season, and caused large biases in the areas far from the original calibration point. Precipitation was also estimated by using scattering index (SI) of SSM/I passive microwave (PMW) data. The signal of SI was very small comparing to the lower elevation areas, and the feature agrees with the observational results previously mentioned. Unusual increase of SI, which was not caused by the precipitation or snow cover, was found in the non-monsoon (Ueno et al., 2000). This noise was associated with diurnal change of SI, such as an increase in the nighttime when the surface skin temperature was colder than the deeper soil temperature. New scattering index was constructed by subtracting the lower frequency difference (19v-37v) which showed better agreement with radar estimated precipitation. In the future, new algorithms are expected by combining IR and PMW data, with the consideration of differences between daytime and nighttime cloud structure.
Ueno K., H. Fujii, F. Yamada, and L. Liu., 2001: Weak and frequent monsoon precipitation over the Tibetan Plateau. Accepted to the Journal of the Meteorological Society of Japan. J. Meteor. Soc. Japan, 79, 419-434.
Kenfichi UENO, N Grody, R Ferraro, and A Gruber, 2000: Improvement of the SSM/I derived precipitation in the Tibetan Plateau during the monsoon season. Proceedings of the 2000 western pacific geophysics meeting, Tokyo, Japan.
