A RATIONALE FOR USING LOCAL AND REGIONAL POINT RAINFALL DATA FOR DESIGN AND ANALYSIS OF URBAN STORM DRAINAGE SYSTEMS

The Danish measuring network for high-resolution rainfall data consist s of approximately 70 tipping bucket rain gauges of which 41 have been operated for more than 10 years. The gauges are separated by one to 3 00 km and cover an area of 43,000 km2. Significant geographical variat ions of extreme rainfall characteristics have been observed. Part of t hese variations can be explained by correlation with the mean annual r ainfall and the existence of a metropolitan effect in the Greater Cope nhagen area. The remaining variation may be attributed to sampling err ors and small-scale spatial variations close to the gauges. Engineerin g methodologies all require rain data of some kind, ranging from desig n storms based on idf-curves for use in simple calculations to high-re solution time series for use in detailed simulation studies. A compreh ensive regional analysis was carried out to account for the geographic al variation and to improve estimation for large return periods exceed ing the actual length of the measured time series. Ideally, rainfall d ata used as input to urban drainage calculations should always be base d on regional rain information. Regional design storms can be made rea dily available based on theory developed in this study. However, a sat isfactory framework for generating synthetic rain series from regional rain information is not yet available. Thus, there will still be a ne ed for using historical rain series in the near future. To improve the basis for choosing representative historical rain series all the Dani sh gauges have been classified according to their deviations from the regional distribution. (C) 1998 IAWQ. Published by Elsevier Science Lt d. All rights reserved.