The Estimation of probable maximum precipitation for Sri Lanka

Abstract

On 24 and 25lh of December 1957 heavy rain fell in the North Central Province; 206.5mm and 219.7mm of rainfall were recorded at Anuradhapura meteorological station and were nearly twice that recorded during the previous high floods. Some of the major tanks such as Kalawewa, Nachchaduwa and Huruluwewa in Anuradhapura breached, resulting in catastrophic damage to infrastructure. The storm of 392.5 mm on 8,h June 1996 over Ratnapura resulted in a disastrous flood which caused enormous damage and rendered many people homeless. The design of a hydraulic structure often requires an estimate of the design flood which has considerable uncertainty about its magnitude. However it should be determined in relation to the project feature for which it is required. When 100% virtual security against failure is required, then the design criteria should be based on the Probable Maximum Flood, which in turn requires the estimation of Probable Maximum Precipitation (PMP). The research compares the results of 24-hour PMP obtained by two methods. One is based on storm maximization or a hydro-meteorological method and the other uses the statistical method developed by Hirschfield. In this study, the last 105 years (1896-2000) of daily rainfall data from the meteorological stations at Anuradhapura and Ratnapura have been analyzed. The 24-hour point PMP for Anuradhapura showed that the results were in the range of 600-625mm based on the statistical method while storm maximization yields a PMP of 425mm with the adjustments of moisture and wind maximization. On the other hand the statistical result for Ratnapura was in the order of 990 to 1120mm and the PMP calculated by storm maximization was 1025mm. If storm maximization was based on true rainfall, PMP values were increased up to 623mm and 1058mm for Anuradhapura and Ratnapura respectively. Finally suggestions have been made to improve the effectiveness of the different approaches, in estimating PMP.