Estimation And Prediction Model Of Crop Reference Evapotranspiration In Taizi River Basin

China is a main country of agriculture, water used in agriculture is about 70% out of that in all national economy, and crop requirement is a main portion of water used in agriculture. Crop requirement is important datum in water resources development and utilization; Necessory basal data in river basins plans. It's also important in project planning, design and management of irrigation and drainage. Refined calculation and forecast of crop requirement is significant in increasing agriculture water usage and improving irrigation management. Reference evapotranspiration is a key parameter to estimate and predict crop requirement, the study about reference evapotranspiration provides a common foundation of crop requirement calculation.Taizi river basin crosses the middle part of Liaoning province, industrial cities centralize in this basin, and the river basin is an important industrial base and marketable grain growing district. Climate has changed largely recent years, and reference evapotranspiration changed as climate variation, the trend of reference evapotranspiration will effect on water resource evaluation and allotment. Searching reference evapotranspiration space-time variation regularity can provide a reference datum on basin planning, proper irrigation scale and irrigation program. It is significance on releasing water resource shortage, promoting proper systemic circulation about water resource-ecology-economy-society and realizing basin agriculture sustainable development and utilization.Using monthly meteorologic data of eight weather stations in Taizi River Basin during 1960 and 2005, the following contents are studied in this thesis:Monthly reference evapotranspiration was calculated by the penman-montieth formula, the characteristics and the trend of ET₀ and meteorological factors changed with year and month were analyzed, and the main factors affected ET₀ were discussed with statistic method also. The results show that the ET₀ in the Taizi river basin is decreased slowly over the past 46 years, and the changing frequency becomes quick since 1980s. The ET₀ in May and June is the highest and that of in January is the lowest in the year. The sequence of the main factors affected ET₀ is sunshine time, wind velocity, temperature and relative humidity. The main factors affected ET₀ in summer is sunshine time, and in the other three seasons is wind velocity. ET₀ shows a tendency that it is higher gradually from west to east.This study evaluates the Hargreaves equation for estimation of monthly evapotranspiration in Taizi river basin during 1960-2005. Comparing the estimate of Hargreaves method with P-M method, the former overestimates the latter during March to October, other months underestimating. The differences are highest in summer, lowest in winter. The main reason about difference of the methods is relative humidity and wind velocity, but relative humidity has more impact on Taizi river basin. The Hargreaves equation produces considerable bias in different region, and the bias is reducing from Upper River to Lower River. A linear relationship between Hargreaves estimation and P-M estimation is proposed for the regional adjustment of the Hargreaves coefficient.Three time series forecasting models are established with monthly ET₀ datas from 1960 to 2004: stochastic time series model, deterministic time series model, wavelet denoise stochastic model. Time series model is established based on historical process of itself, and autocorrelation is the premise to fit and forecast data. The idea about model establishing of stochastic time series model is opposite to deterministic time series model, the former takes out of "signal" and the the latter distills "signal". Both of the models have good fitting ability to datas, but bad in the peak value month. Wavelet denoise stochastic model has a denoise process with original data, eliminating the useless disturbing signal. Wavelet denoise stochastic model improves the disadvantage that the two fomer models can't fit peak value well, but it has good fitting result, and is an ideal predicted model. The three models all have stability in model form, and have no significant difference between fitting value and true value.After analyzing present status and problem of crop reference evapotranspiration, the changing character, estimation and forecast methods of crop reference evapotranspiration have been studied and searched.in Taizi river basin. This thesis includes three part, that ET₀ ragularity and influencing factors based on Penman-Monteith equation, operational evaluation and error analysis of Hargreaves equation, ET₀ time series establishing.