Frost Damage Assessment Of Winter Wheat Based On Multiple Meteorological Factors

Spring frost of wheat in winter is an agricultural meteorological disasters means that the leaves,spikelets,and other organs of wheat are damaged by extreme low temperatures in spring,resulting in a decline in wheat yield or quality in the year.The object of study is the typical frost-prone region-Henan Province,in which frost events recruited by statistical yearbook during 1964-2014 are the basis for validations.This article select minimum grass temperature close to the wheat plant body temperature as the frost evaluation criteria,analyzing the effects of weather conditions on spring frost damage of winter wheat by combining correlation analysis and K-mean cluster analysis and setting up stepwise regression equation.The main results and conclusions of this paper are as follows:?1?The spatial and temporal distributions of actual frost and meteorological frost in Henan Province from 1964 to 2014 are quite different.The areas with high incidence of actual frost damage in50 years are concentrated in western Henan and northern Henan;the areas with high incidence of meteorological frost are in central Henan and western Henan.The frequency of meteorological frost in the province was significantly higher than the actual frost.And meteorological frosts have a lower degree of agreement with actual frosts.The accuracy of traditional meteorological indicators for single station frost damage assessment was only 14.2%.?2?Analyzing through actual frost show that the actual slight frost damage is sensitive to previous heat factors with larger diurnal temperature range,sunshine hours and lower relative humidity,wind speed belonging radiation-type frost.And the severe actual frost damage is more sensitive to the heat factors to the previous moisture factors.There are two kinds of weather conditions during freezing,one have relatively high relative humidity,lower daily temperature,low sunshine hours and average wind speed belonging mixed frost;the other is same with actual slight frost.In general,the effect of previous meteorological conditions on the actual frost was relatively weak,and the actual frost occurred mainly in the weather condition during freezing.?3?This paper set up spring frost model by multiple weather factors from 6 automatic observing stations in Shangqiu respectively during sunny days and rainy days.Two models pass and F test and the regression result is significant.x₁Daily range?x₂ Precipitation?x₃ Sunshine hours?x₄ Average wind speed?x₅ Average relative humidity andx₆ Minimum temperature.The model is as follows:Clear day model:y?28?-.75 5-.0102x₁?10?0.433x₄?10?0.055x₅?10?0.96x₆Rain and cloudy day model:y_rain?28?-.80 6?10?0.98x₄?10?0.55x₅?10?0.94x₆The model was used to predict the spring frost in 17 stations of Henan province from 2007 to 2011.The results show that the percentage of right prediction in the actual frost is 87.1%and the precision of the model is 36%by distinguish between sunny and rainy days.The evaluation results of the traditional meteorological indicators in the corresponding years show that the percentage of right prediction in the actual frost is 51.6%and the precision is only 26.7%.Obviously,the right number of events predicted from multiple weather factors model distinguishing between sunny and rainy days are higher than tradition meteorological indicators.However,due to the relatively high number of empty prediction,the accuracy rate of model results is only higher than that of traditional meteorological indicators 9.3%.