Runoff Forecast In Flood Season For Huanren Basin Based On Climate Factors And SEAS5 Precipitation Forecast

With the rapid economic and social development,people's demand for water resources continues to increase,and the contradiction between the supply and demand of water resources has become increasingly prominent,and the management of water resources in my country is therefore facing severe challenges.Runoff is an important manifestation of water resources.Improving its predictive ability can provide important help for water resources management.Flood season is the main concentrated period of runoff.It is more realistic to carry out flood season runoff forecasting to improve the water resources management capacity of river basins.The main research contents are as follows:(1)A factor screening method based on the CFS index was proposed,and a flood season runoff forecast model based on climate factors such as atmospheric circulation and sea temperature was constructed,and the influence of precipitation and the initial state of the basin on the runoff was quantitatively analyzed.First,combine the principle of runoff generation,use r-CFS and MIC-CFS to select the initial state factors and climate factors of the basin;then,build the flood season and monthly runoff forecast models based on the screened factors;finally,compare the results with the previous ones.The runoff forecast results of the wrapper method were compared and analyzed to verify the validity of the CFS index and the applicability of rCFS and MIC-CFS.In addition,based on the measured precipitation,random forests are used to construct a "perfect" runoff forecast model to quantitatively analyze the impact of precipitation and the initial state of the basin on runoff.The results show that the previous climate factors can reflect the future climate information to a certain extent,and can be effectively applied to the flood season runoff forecast;compared with the forward search package method,the factor screening method based on the CFS index can improve under the premise of ensuring the effect of factor screening Factor screening efficiency;compared to rCFS,the MIC-CFS index can better measure the relationship between the forecast factor set and the runoff during the flood season;for the runoff during the flood season and July,precipitation has a leading role,and the degree of influence is 0.92,respectively.0.86,while in other periods,precipitation and the initial state of the basin have important influences,and the degree of influence of precipitation is 0.55-0.63.(2)The applicability of SEAS5 precipitation forecast products in Huanren Basin was evaluated,and a precipitation forecast correction model based on Bayesian probabilistic joint probability model was established.The SEAS5 original precipitation forecast and the corrected precipitation forecast are evaluated from four aspects: system deviation,reliability,sharpness,and forecasting skills.The results show that the original forecasts have large system deviations,and the forecast reliability is insufficient,and affected by the deviations and reliability,most of the forecasting skills scores are negative;and the correction model can effectively eliminate the system deviations and improve the forecast reliability.At the same time,the sharpness of some forecasts will be reduced,but most of the forecasting skills will be upgraded to positive skills,and the overall performance will be improved.But overall,the SEAS5 precipitation forecast has potential application value in the Huanren watershed(3)The flood season runoff forecast based on the SEAS5 precipitation forecast was carried out,and the runoff forecast correction model based on the Bayesian joint probability model was established.Evaluate the deterministic results(ensemble mean)of the original and corrected runoff forecasts and the probabilistic forecast results,and finally compare the results with the runoff forecast results based on climatic factors to evaluate their availability.The results show that in the corrected deterministic runoff forecast results,the pass rate in August was reduced from 93% to 89%,and the rest of the period was improved.Compare the results with the results based on climate factors,except for June In all periods of the year,the runoff forecast based on the SEAS5 precipitation forecast has been improved;and in the corrected probabilistic runoff forecast results,the reliability of each month has been improved,and the low-tech forecasts have been greatly improved.Overall,the SEAS5 precipitation forecast has good availability in the Huanren watershed,and can provide support for water resources management...