Scale-dependent differences of areal reduction factors: from minutes to hours

Abstract

Design rainfall is required for numerous applications in hydrology. If based on rain gauge time series, an adoption in space without further processing leads to an overestimation of spatial rainfall extreme values. Areal reduction factors (ARF) reduce point extreme values in space to achieve more realistic areal rainfall extreme values. The necessity of reduction increases with higher temporal resolution. However, the low density of most rain gauge networks hinders the estimation of representative ARF. In this study ARF are derived from 5 min rainfall time series of the high-density WegenerNet in Austria with 143 rain gauges distributed over 300 km² (~ 0.5 gauges per km²).ARF dependency on area (up to 81 km²), rainfall duration (5 min to 6 hours), return period (1 year to 10 years), seasonality (four seasons) and altitude (260 m to 400 m) are studied. The results provide new insights into the research field, especially for the short durations. In addition to providing explicit ARF values, the main conclusions are:ARF decrease with increasing areal extent considered for all durations. ARF decrease with increasing temporal resolution for all return periods. While ARF for hourly values (and coarser) decrease with increasing return period, the opposite is found for shorter durations. ARF vary strongly between seasons, with lowest values found for spring. Altitude-dependency of ARF increases with areal extent considered, whereby ARF values increase with altitude. The resulting ARF data set is unique with its applicability for high-resolution extreme values as needed for urban hydrology. The results are assumed to be transferable to other regions with similar hydro-climatologic characteristics.