Extreme hydrological events in karst areas of Slovenia, the case of the Unica River basin

Abstract

The study presents natural hazards in Slovenia's karst, focusing on flooding in karst poljes. A specific study was done on the flood dynamics of two typical and connected karst poljes (Cerknica and Planina) of the Classical Karst region. In the case of particularly extreme hydrological conditions in the autumn of 2008, detailed analyses of the recharge-discharge regime and the interrelationship of flooding on the two poljes were done. Daily precipitation, discharge, and water level values from, several monitoring sites were analyzed and cross-correlated, and additional hydrological analyses were done using a digital elevation model in order to acquire water level increase and decrease intensity, flood water volumes, and the extent of flooding and to understand the conditions controlling karst flooding. The results reveal that the hydrological functioning of the studied karst poljes is influenced by the hydrogeological and temporary hydrological conditions in the catchment area. The response of the binary karst system (i.e., the influence of autogenic and allogenic recharge) is especially distinct. The study shows that during extremely intense recharge, the reactions of karst aquifer systems to precipitation, are as rapid as the response of surface waters (the water level of Cerknica Lake increased with an intensity of 38-63 cm/day or 55 m3/s respectively) while retention capacities are negligible. In contrast to flash floods, floods in karst areas may last from several weeks to several, months. For the observed period a three-dimensional simulation of the flooding was made. At the maximum recorded water level, the volume of water on the Cerknica polje was 51 million m 3, and 26 million m3 on the Planina polje. The maximum extent of flooding on the Cerknica polje was 23 km2 and on the Planina polje 9.5 km2, On the basis of the study, information was provided regarding future hazard mitigation However, the study demonstrated that a sufficiently dense monitoring network is necessary to predict the occurrence and duration of floods with greater certainty, © 2010 Lavoisier SAS. All rights reserved.