| dc.description.abstract | Taiwan, being located at the subtropical zone in western Pacific Ocean, recently suffers severe droughts in two consecutive years. The weather in Taiwan is the typical island climate, with its characteristics such as high intensity rainfall and disproportionate distribution. For this reason, a serous drought condition can be recovered with one heavy rainfall event. This study investigates not only the effects of antecedent rainfall on the rainfall-runoff relationships, but also the possibility of recovery from a given drought, to perform risk analysis.
The precipitation record from 1957 to 1994 is used for analysis of the effects of antecedent precipitation, and Shih-Men watershed is chosen as the case study area. Firstly, three sets of 10-days periods were used to classify antecedent rainfall into three populations, namely “wet”, “normal”, and “dry”. These groups were then interrelated with three groups of “current” rainfall conditions, subdivided upon a similar bases. This resulted nine unique subpopulations consisting of combinations of antecedent and current periods. A “production ratio” (runoff/rainfall) was defined for each current period. Furthermore, runoff is subtracted from base flow by applying linear recession assumption. A simple statistical analysis was performed to evaluate the differences of production ratios between subpopulations.
Furthermore, a conceptual rainfall-runoff model was suggested by separating the runoff into two portions, namely, base flow portion and current rainfall induced runoff portion. Four models, based on different parameters, were investigated for their performances, with the rainfall-runoff record from 1995 to 1999 is used for verifications. The results show that compared with the classification of antecedent period, the classification of current periods plays a much more important role on the rainfall-runoff relationship.
Finally, two severe draught event in 1996 and 1997 were chosen to demonstrate the applicability of risk analysis with use of the proposed model. From the results, probabilities of recovery from these two droughts were estimated as 0-0.3, and 0-0.7. | en_US |