Analysis of short-term precipitation extremes for design purposes in Hungary

The rainfall intensity for various return periods are commonly used for hydrological design. In this study, we focus on rare, short-term precipitation extremes and related return values which are relevant durations in the planning and operating demands of drainage and sewerage systems in Hungary.

Hungarian Meteorological Service together with the National Water Directorate (Municipal Water Management Department) with the professional assistance of the Hungarian Chamber of Engineers (Water Management and Water Construction Section) developed a user service for design purposes.  The user can download the return levels of the short-term rainfall intensity for the closest meteorological station to the location of the planned object specified with geographical coordinates.

Automatic stations replaced the ombrometer in many places in Hungary, particularly from the late 1990s. The change of the measurement practice do not allow simply merging the data recorded form the registering paper in the past and the recent 10 minutes measurements.   The most intense 5, 10, 20, 30, 60, 180 min sub-totals per rainfall events were recorded from the ombrometer registering paper before atomization, typically until 1993. By contrast, the 10 min precipitation sum from the AWSs are stored in the meteorological database of the Hungarian Meteorological Service from automatization. In order to join together the older and the AWS measurements it was necessary to develop a method to make this possible. Therefore, we downscaled the 10 min data in time. The sampling of the AWSs is one minute, although the 1-minute data are available only for some stations in the digital database. We applied a linear regression model to downscale the 10-minute data for 1 min. After this, we can derive the most intense sub-totals per events from the AWS data as if they have been measured with the ombrometers. Therefore, several station data series can be made longer, thus the confidence intervals of the return level estimates are narrowing, and the quality of the service is improving.