冰雹对美国玉米带作物产量影响模拟

    Simulation of hail effects on crop yield losses for corn-belt states in USA

    • 摘要: 为了尝试计算机模拟方法能否成为制定作物保险保费的辅助性工具,该研究运用计算机模拟模型预测冰雹对作物产量的影响。通过对现有EPIC(综合气候因素的环境政策)模型增加冰雹天气模块,即冰雹事件发生概率的数学模型,模拟冰雹对作物产量的影响。除此之外,该研究还模拟了干旱和霜冻等天气因素对美国Iowa,Illinoi和Indiana等3个玉米带州作物产量的影响。首先介绍数据来源及处理方法,讨论建立冰雹模拟模型过程以及对模型进行有效性检验;然后运用统计分析方法对模型模拟结果与实际观察结果进行比较,检验模型模拟结果的准确性。结果表明EPIC模型可以达到95%甚至更高的产量预测的准确性;同时,冰雹灾害模拟结果也可以达到一个较为合理的准确性(R2>0.7)。这些结果表明本研究所建立的增加冰雹模块的EPIC模型可以作为一个较为可靠的冰雹引起的作物产量损失的预测方法.该模型可以用来模拟冰雹事件发生的概率以及其对各种作物产量造成的损失。

       

      Abstract: In this study, a computer simulation model was used for predictive analysis of hail effect on crop yield losses. A pre-existed environmental policy integrated climate ( EPIC ) model was modified by introducing the hail weather module using an embedded stochastic probability function. This study focuses on estimating effects of the three important weather factors (hail, dry and cold) which make the most important contribution to the crop yield losses in U.S. corn-belt states of Iowa, Illinois and Indiana. Data sources, model development, calibration, and validation were described in detail, the model performance was tested, and statistical comparisons of simulated losses of crop yields against observed hail-induced crop yield losses were made. The results showed that the crop yield predictions reach 95% or higher accuracy and hail damage simulation also achieve a reasonable level of reliability ( R2 was above 0.7). These suggest that using the hail-integrated EPIC model can properly provide a reliable method for hail-related crop yield loss estimation. The model can be utilized to simulate hailstorm events and their damages to various field crops.

       

    /

    返回文章
    返回