基于广义线性模型的水稻种植风险评估

Risk Assessment of Rice Planting Based on Generalized Linear Modelbh

农作物风险评估是农业风险管理的前提与基础.为探索科学有效的水稻种植风险评估模型,将水稻种植风险评估过程分为承灾体脆弱性评估和致灾因子危险性评估两个过程,利用广义线性模型解决自然灾害会对水稻产量造成多大程度的损失问题,在产量波动服从布尔分布的假设下,建立起致灾因子(用连续降雨天数、连续低温天数等指标表示)与产量波动之间的线性关系,完成承灾体脆弱性评估;利用天气发生器模型模拟各地区致灾因子发生情况解决自然灾害概率分布的问题,完成致灾因子危险性评估;最终给出风险评估的量化表达.通过黑龙江农垦气象及水稻产量数据的实证分析验证了模型的有效性,并讨论了该模型在保险费率厘定方面的应用.

Crop risk assessment is the premise and basis of agricultural risk management.To explore a scientific and effective rice planting risk assessment model, this paper divides the risk assessment process into two processes: vulnerability and hazard risk, and discusses the frequency of natural disasters and the impact on rice. In this paper, the generalized linear model is used to solve the problem of the loss of crop yield caused by natural disasters.Under the assumption that the yield fluctuates from the Burr distribution, the causative factors(continuous rainfall days, Continuous low temperature days and other indicators that)and the volatility of the linear relationship between production to complete the assessment of vulnerability. Using the weather generator model to simulate the occurrence of disaster risk factors in various regions to solve the problem of natural disaster probability distribution, to complete the hazard risk assessment. Finally, a quantitative expression of the risk assessment is given. This paper verifies the validity of the model through empirical analysis of meteorological and rice yield data in Heilongjiang Land Reclamation, and discusses the application of the model in determining the premium rate.