Extreme rainfall events: Vulnerability analysis for disaster management and observation system design

摘要

Extreme rainfall events today pose a serious threat to many populated and urbanized areas worldwide. An accurate estimate of frequency and distribution of these events can significantly aid in policy planning and observation system design. We report here a high-resolution (10 km) analysis of heavy rainfall episodes (defined as 24-h rainfall exceeding 250 mm) over the Indian region. The dataset, recently developed by NOAA, USA, provides daily composite rainfalls for the period 2001 - 06 at locations approximately 10 kin apart. We first assess the reliability of the dataset by comparing it with daily gridded (1 degrees x 1 degrees) rainfall data from IMD and three-hourly gridded (0.25 degrees x 0.25 degrees) data from TRMM for the overlap years (2001 - 04). A category-wise analysis of the high-resolution data reveals a number of hotspots of vulnerability; in particular, the semiarid region in northwest India emerges as a high-vulnerability area in terms of extreme rainfall events. The high-resolution analysis also clearly reveals the corridor of the monsoon trough, lined by a flower-pot distribution of extreme rainfall events along the flanks. This can be a valuable input for precision design of field experiments on the continental trough or on localized extreme events like thunderstorms. Other important implications for areas like vulnerability assessment, planning and mesoscale forecasting are discussed.