Abstract:
The objective of streamflow synthesis is to study the usage of HEC-4 (Sacramento watershed model), a physical resemblance model. Streamflow data of four principal gaging stations, namely B5, B6, B7 and B8 in Petchaburi river basin during the period of B.E. 2495-2534 were used in this study. The results could be summarized as flollows : The stremflow synthesized by HEC-4 model of individual station was inconsistent between adjacent stations. While streamflow synthesized on the river basin exhibited better consistency. Therefore, it is recommended that streamflow should be synthesized on the river basin basis. Including linear trend in streamflow synthesis, in readjustment, precaution should be taken since negative value could be encountered. In this river basin, the relationship of linear trend was found to be not of significant level. Streamflow synthesis on the river basin basis aggregated annually, the mean value as well as the variance of the observed data with or without trend adjusted were maintained. If the maximum and minimum values of the synthesized data were seemed to be abnormal, it is recommended that outliers test should be carried out before data adjustment. The calibration and verification of streamflow simulated by SCMT model showed that the parameters were not constant. Streamflow synthesis of B5 station where only one rainfall station situated near B5 gaging station showed low correlation between the observed and synthesized data comparing to other stations, since the rainfall data was not a good representative of the area. The streamflow synthesized on river basin basis, impact of B5 station was significant since B5 subcatchment was approximately 50% of the river basin area considered. Therefore the index station used was not a good representative of the basin rainfall. It was noted that other subcatchments streamflow synthesized exhibited good correlation due to good representative rainfall data of the area. It is recommended that if a longer period of streamflow data than observed data is needed for analysis. A physical resemblance model, for example SCMT, should be first used to verify data compatibility, thereafter, a statistical model, for example HEC-4, should be used to synthesis streamflow data on the river basin system basis.