Study On Monthly Scale Forecasting Method Of Winter Wheat Water Requirement In China

China is the world’s largest wheat producer and wheat production accounts for 20%of the grain output,winter wheat production can account for more than 90%of the total wheat production,occupying an important position.Based on the large regional and inter-annual variation of water resources in China,uneven distribution of water resources and low utilization efficiency of agricultural water resources,the accurate prediction of winter wheat water demand has become an important basis for making reasonable irrigation decisions and realizing the sustainability of agricultural production and water resources.At present,irrigation decision making is mainly based on historical statistical data,which cannot meet the practical needs of timeliness and refinement of dynamic water resources management.The prediction of crop water requirement（ET_c）is mainly based on short-and medium-term prediction of 1-7d.The results of this study are mostly applied to the irrigation practice at the site scale.However,there is a problem of mismatch of forecast time in the water scheduling guidance of large-scale irrigation areas.This study is based on the accuracy evaluation of 1-30d weather forecast data from 2391meteorological stations in China.Penman-Monteith Forecast（PMF）and temperature-ured were selected for this study Six reference crop evapotranspiration（ET₀）prediction models,Penman-Monteith（PMT）,Hargreaves（HG）,Mc Clound（MC）,Makkink（MK）and Priestley-Taylor（PT）,were studied.The suitable ET₀ prediction models for different forecast periods in China are studied.The crop coefficient method was used to forecast the ET_c of winter wheat,and the water surplus and deficit of winter wheat during the growing period was analyzed by combining with the precipitation forecast data of the same period,so as to provide reference for water scheduling in the growing areas of winter wheat in China.The results of the study show that.1.Meteorological factor forecastFrom the forecast time,the forecast accuracy of temperature,wind speed and sunshine duration decreased with the increase of the forecast period,and the forecast accuracy of each meteorological factor decreased little with the increase of the forecast period in 16～30 days.On the whole,the accuracy of minimum temperature forecast is higher than that of maximum temperature forecast,while the accuracy of wind speed and sunshine duration forecast is not high.From the perspective of spatial distribution,the prediction accuracy of meteorological factors in high latitude and high altitude region is lower than that in low latitude and low altitude region.The accuracy of temperature forecast is the worst in arid and semi-arid areas in northern China and Qinghai-Tibet Plateau,and the coincidence of temperature forecast in high latitude is better than that in low latitude.The error of wind speed forecast is the largest in arid and semi-arid areas in north China and northeast China plain.The accuracy of sunshine duration in Northeast plain and Qinghai-Tibet Plateau is the worst in short and medium term forecast,while the accuracy of sunshine duration in eastern region is lower than that in central and western region in long term forecast.2.Reference crop evapotranspiration forecastIn terms of forecast time,the HG model performs best in the short-term ET₀ forecast,and the PMF model is the second best;in the medium-term forecast,the HG and PMF models have the best forecasts,while the PMT and MC models have the worst forecasts;in the long-term forecast,the MC and PT models have the best forecasts,and the PMF model has the second best forecasts.With the increase of the forecast time,the forecast accuracy of each forecast model shows a decreasing trend,among which the decreasing trend of the long-term ET₀forecast accuracy is not obvious.In terms of spatial distribution,the HG model has the best global applicability in short-term and medium-term ET₀ forecasting,and the PMF model is the next best;in long-term forecasting,the most suitable ET₀ forecasting models from north to south in China are PT,PMF,and MC,but there is no obvious distribution feature in the selection of ET₀ long-term forecasting models in the Qinghai-Tibet Plateau region and the northern arid and semi-arid region.The preferred forecast models have the highest forecast accuracy in the Qinghai-Tibet Plateau region and the northern arid and semi-arid region,and the worst forecast accuracy in the middle and lower reaches of Yangtze River,and the overall ET₀ forecast value in China is 11%smaller.3.Winter wheat crop water demand forecastThe forecast water requirement of winter wheat in China was higher in the north and lower in the south.The forecast water requirement of crops in Qinghai-Tibet Plateau region was the highest,and the forecast water requirement of crops in South China was the lowest.The water requirement of winter wheat crops increased from October to April of the next year,and the maximum water requirement of winter wheat crops in April was 100mm.The water requirement of winter wheat crops from July to September is small.The water surplus and deficit of winter wheat showed the distribution characteristics of surplus in southern regions and deficit in northern regions.From October to June of the following year,the water deficit of winter wheat was in deficit state,especially in April and May,which were 65mm and 78mm respectively.The spatial distribution of winter wheat precipitation satisfaction rate was high in the south and low in the north,and the precipitation satisfaction rate in April and May was the highest,which were 57%and 63%,respectively..