资源环境科技发展态势分析平台

Heavy rainfall, floods and other hydroclimatic extremes may be related to specific states of organization of the atmospheric circulation. The identification of these states and their linkage to local extremes could facilitate a physically meaningful quantification of local extremes in future climates and could allow forecasting extremes conditioned on the large-scale atmospheric state. A novel methodology is presented that combines non-linear, non-parametric methods to link heavy precipitation events (HPEs) to atmospheric circulation states. Using daily rainfall data for the period 1951-2015 from 37 gauges in the Lazio region in Italy, HPEs are defined. For the same period, two atmospheric variables, namely, the 850hPa geopotential height field and the integrated vapour transport (IVT), are derived from reanalysis data. The geopotential configurations driving heavy precipitation in the region are identified by combing self-organized maps and event synchronization. First, a finite number of representative geopotential configurations is identified. Rainfall gauges are pooled into clusters, which show synchronized occurrence of heavy precipitation. Furthermore, geopotential configurations are identified, which tend to drive HPEs. For these geopotential states, the probability of HPE occurrence as a function of IVT is calculated through a local logistic regression model. Finally, it is explored whether the identified patterns are related to the occurrence of atmospheric rivers, which govern the atmospheric humidity transport from the tropics and subtropics to Europe. The relation found demonstrates the reliability of the proposed methodology.