Spatiotemporal stochastic event-rainfall process in a small mountainous catchment

Abstract

The spatial distribution and temporal process of rainfall are important with respect to rainfall-induced hazards, which represent major problems in small mountainous catchments. This study considered the rainfall process as a stochastic process, and investigated its spatial variation and temporal evolution using data obtained from a dense monitoring network in a small catchment. The event-scale rainfall–elevation and rainfall–area relations were proposed, and the temporal process was investigated at 1-min resolution and characterized by an index that included the onset time, duration, and center time of the rainfall peak period at each gauge. Analysis of the errors associated with using the relations revealed that the errors were random and followed a normal distribution. Consequently, the rainfall process could be expressed asy = f(x) + ε, wheref(x)represents the specific deterministic relations used to estimate the rainfall parameters, andεrepresents the estimation errors. A stochastic simulation method that incorporates the deterministic empirical relations of rainfall amount and the stochastic process of rainfall errors was proposed to simulate the rainfall process at any gauge. The simulation procedure was described in detail and validated using real event-rainfall data, also the influence of gauge selection to the hydrological forecasting (e.g. water flood simulation and debris flow forecasting) was evaluated. This study marks an attempt to establish a framework for generation of the temporal process of rainfall at a high resolution (1-min) which is significant to the hydrological hazards forecasting, and could be extended to other unknown places, thereby realizing the spatial distribution of rainfall within a catchment.