Study of small time-scale rainfall amount distribution: a case study on Xiaheyan Station, the Yellow River Basin

摘要

Rainfall is the most important resource of regional water resources; also it is a key part of the global water cycle. Rainfall data is the crucial input of all hydrological models. Usually the rainfall data that are credible and easily obtained are daily rainfall data. How to downscale daily rainfall amounts to small time-scale amounts is an important point in distributed hydrological models. The two widely-used methods, random distribution pattern method and triangular distribution pattern method, based on no physical background or statistical basis, perform not as well as we expected. A reasonable method should be based on appropriate rainfall amount distribution patterns. This study investigated the real distribution of 2-h duration rainfall amounts using the long-term rainfall records from 1951 to 1997 of the Xiaheyan hydrological station of the Yellow River basin. The rainfall records were grouped into five categories according to the daily rainfall amount: 0-2 mm, 2-10 mm, 10-20 mm, 20-50 mm and 50-100 mm. For each category, statistical analysis showed that rainfall amount distribution patterns were close to the Pearson-III distribution. But different categories had different Pearson-III parameters. Thus new methods of downscaling daily rainfall amounts could be developed, This would significantly improve the accuracy of distributed hydrological models. Furthermore, this study provided a new tool to evaluate global climate change influence on regional rainfall and water resources .