Utilization of GMS Data for Estimation of Ganges River Run-off

上一财政年度提供了GMS-WEFAX数据采集硬件和基本处理软件。进一步开发并实现了对红外云图云面温度分布的统计分析程序。对于与恒河流域相关的大尺度水文研究，云覆盖率可以用FC分布来表示，其中FC是在1 ° C纬度/经度的小正方形框架中指示温度低于215 ° K的像素的比率。将该指数成功地应用于该地区的短期和长期气象分析中，并在对1996年5月袭击孟加拉国的龙卷风的追踪中指出了时间分辨率的不足。然而，沿着龙卷风的东南边缘沿着发现了明显陡峭的FC梯度。这样的发现在大尺度气象学是重要的未来tornado forecasting. Expensive分析FC distributions在1996年季风季节（5月至9月）揭示了定期的日生成和衰减的云在孟加拉国和毗邻的恒河和孟加拉湾的广大地区。用伪矢量表示法较全面地描述了云量的变化。这一技术将有助于季风和相关降水的宏观研究。虽然现有的区域地形、降水和河流径流资料在质量和数量上都不充分，但它们已被汇编。京都大学灾害防灾研究中心，然而，在地面实况数据的积累方面一直在付出相当大的努力，GMS数据的汇编也已成为一项既定的例行工作。这些数据有望在几年内与恒河径流的预测相关联。

Hardware for GMS-WEFAX data acquisition and software for their fundamental processing were provided within the last fiscal year. Programs were further developed and implemented for statistical analyzes of cloud-surface temperature distribution obtained from IR image. For widescale hydrological studies associated with Ganges River basin, cloud coverage is properly represented by FC distribution, where FC is the rate of pixels indicating lower temperature than 215ﾟK in a small square frame of 1ﾟ latitude/longitude. This index was successfully introduced into trial analyzes on both short-term and long-term meteorology in the present region.Insufficiency in time resolution was pointed out in tracing May 1996 tornado which attacked Bangladesh. Notably steep FC gradient, however, was detected along the south-east edge of the tornado. Such a finding in large-scale meteorology is important for future tornado forecasting.Extensive analysis of FC distrbutions in 1996 monsoon season (May through september) revealed regular diurnal generation and decay of clouds over a wide area of Bangladesh and adjoining Ganges River and Bengal Bay. Variation in cloud coverage in comprehensively described by means of pseudo-vector expression. This technique will contribute to the macroscopic study of monsoon and associated precipitation.Existing data on regional topography, precipitation and river run-off were compiled, although they are sufficient neither in quality nor in quantity. The Disaster Prev.Res.Center of Kyoto Univ., however, has been paying appreciable efforts toward accumulation of ground truth data, and compilation of GMS data has also become an established routine work. These data will hopfully be correlated for forecasting of Ganges River run-off within several years.