基于农业灾情数据的贵州省粮食单产灾损估算模型

Establishment of the Estimation Model for Disaster Loss of Grain Yield per Unit Area Based on Agricultural Disaster Data in Guizhou

【目的】探明贵州省粮食气候减产量与灾情程度的关系,为贵州省防范农业自然灾害风险和科学预测粮食单产提供决策依据。【方法】利用2010—2020年贵州省粮食灾情统计数据,计算干旱、洪涝、风雹、低温的受灾率、成灾率及绝收率,并用灰色关联法评价各灾种及灾害程度对全年粮食及水稻、玉米、马铃薯气候单产的影响,运用多元回归方法分析粮食气候减产量与灾情的关系,构建基于灾情的粮食单产灾损估算模型。【结果】从发生频率看,干旱为贵州省2010—2013年主要的农业气象灾害,2014年起逐步由旱转涝;发生程度中等、范围较广的干旱及程度较重、局地性强的低温灾害是造成贵州省粮食作物减产的主要灾害。构建了粮食气候减产量(ΔY_(w))与总受灾率(Kts)、总成灾率(Ktc)、总绝收率(Ktj)的灾损评估模型:ΔY_(w)=189.2Kts+4490.6Ktc+50.0Ktj+85.2(R^(2)=0.760,P<0.01),通过模型估算得到的模拟单产与实际单产的相关性极显著(P<0.01),且误差在合理范围内。【结论】贵州省粮食总受灾率、总成灾率和总绝收率与气候减产量呈极显著正相关关系,即灾害程度越高,气候减产量越大,粮食因灾损失也越大;基于灾损的粮食单产估算模型可估算当年粮食单产。

【Objective】The relationship between climate yield reduction of grain crops and disaster degree is discussed to provide the decision basis for precaution of agricultural natural disaster risk and scientific forecast of grain yield per unit area in Guizhou.【Method】The disaster rate,real disaster rate and non-harvest rate of grain crops under the conditions of drought,flood,hailstorm and low temperature are calculated based on disaster statistical data of grain crops in Guizhou from 2010 to 2020.The effects of different disaster types and degrees on yearly yield of grain crops and climate yield per unit area of rice,maize and potato is evaluated by the grey correlation method,and the estimation model for disaster loss of grain yield is established according to the relationships between climate yield reduction and disaster degree analyzed by the multiple regression method.【Result】From the aspect of occurrence frequency,drought is the major agricultural meteorological disaster during 2010-2013 but flood is the major agricultural meteorological disaster since 2014 gradually.The drought with moderate occurrence degree and wide range,severe drought and local low temperature are the major disasters to cause yield reduction of grain crops in Guizhou.The established estimation model for disaster loss of grain crops isΔY_(w)(climate yield reduction)=189.2 Kts(total disaster rate)+4490.6 Ktc(total real disaster rate)+50.0 Ktj(total non-harvest rate)+85.2(R^(2)=0.760,P<0.01).The correlation between actual and simulated grain yield reaches at very significant level(P<0.01)and its relative error is within a reasonable range.【Conclusion】Total disaster rate,total real disaster rate and total non-harvest rate of grain yield are positively related to climate yield reduction very significantly,which means that grain yield loss increases with increase of climate yield reduction and disaster degree.The established estimation model for disaster loss of grain crops can be used to estimate grain yield of the year.