Zhang Yajie, Wu Hui, Wu Sheng'an, et al. Determination of area precipitation thresholds of rainstorm-flood hazard in the Nandu River Basin. J Appl Meteor Sci, 2014, 25(6): 731-740.
Citation:
Zhang Yajie, Wu Hui, Wu Sheng'an, et al. Determination of area precipitation thresholds of rainstorm-flood hazard in the Nandu River Basin. J Appl Meteor Sci, 2014, 25(6): 731-740.
Zhang Yajie, Wu Hui, Wu Sheng'an, et al. Determination of area precipitation thresholds of rainstorm-flood hazard in the Nandu River Basin. J Appl Meteor Sci, 2014, 25(6): 731-740.
Citation:
Zhang Yajie, Wu Hui, Wu Sheng'an, et al. Determination of area precipitation thresholds of rainstorm-flood hazard in the Nandu River Basin. J Appl Meteor Sci, 2014, 25(6): 731-740.
Using hydrological model to determine area precipitation thresholds of rainstorm-flood hazard is a tendency of hydrological and meteorological forecast. The Nandu River Basin, locating on the north edge of Hainan Island, is the largest basin of tropical regions in China. And the frequent flood of the Nandu River Basin attracts wide public concern. The HBV (Hydrologiska Byråns Vattenbalansavdelning) model is a semi-distributed conceptual hydrological model with multiple versions, used in more than 40 countries and regions around the world. Using HBV-D model which is suitable for large-scale basin, the basin hydrologic characteristics of the Nandu River Basin are simulated and the area precipitation threshold values are determined. These effects may also provide scientific evidence for early warning in the Nandu River Basin.The model is run in terms of observed daily precipitation, air temperature during 1976-1987 and 2009-2010, and the simulated runoff is verified with corresponding hydrological observations of Longtang Hydrologic Station. Taking 1976-1981 as calibration period, several model sensitivity parameters are selected and calibrated by programming. Periods of 1982-1987 and 2009-2010 are selected for model validating, and the Nash-Sutcliffe efficiency index and correlation coefficient are evaluated. Verifications show that the Nash-Sutcliffe efficiency indexes are 0.891, 0.831 and 0.953, and correlation coefficients are 0.944, 0.912 and 0.977, both passing the test of 0.01 level in 3 periods. It indicates that the model can accurately simulate the Nandu River Basin hydrological characteristics. And it's able to determine area precipitation threshold values of rainstorm-flood hazard in the Nandu River Basin.The curve of stream flow and water level is simulated with historical hydrographic data of 1976-1987 when the water level is greater than or equal to 7 m. Curves of area precipitation and different previous water-levels (7 m, 8 m, 9 m, 10 m and 11 m) are determined by hydrological model HBV-D of the Nandu River Basin. Curves can calculate how many meters the water level will rise when storm comes, in the condition of five previous water-levels. Finally, according to water levels of warning, 10-year return period, 30-year return period and 50-year return period as critical criterions for different warning grades, the area precipitation thresholds in different previous water-levels are determined.The accuracy of area precipitation threshold values are verified using observations of four floods. The result indicates that these values are suitable for forewarning, but the missing of warning is still inevitable. To improve timeliness and accuracy, hourly rolling forecast and early warning can be carried out.
Lindström G, Johansson B, Persson M, et al.Development and test of the distributed HBY-96 hydrological model.J Hydrology, 1997, 201:272-288. doi: 10.1016/S0022-1694(97)00041-3
[17]
徐宗学.水文模型.北京:科学出版社, 2009:281-283.
[18]
Krysanova V, Bronstert A, Müller-Wohlfeil D I.Modelling river discharge for large drainage basins:From lumped to distributed approach.Hydrological Sciences Journal, 1999, 44(2):313-331. doi: 10.1080/02626669909492224