Risk evaluation of agricultural disaster impacts on food production in southern China by probability density method

Meteorological disaster occurred frequently in China and inflicted great losses to agriculture. The strengthening of disaster risk assessment is necessary, which also has important practical significance for reducing the influence of and losses from meteorological disasters. Due to the advantages, risk probability method is adopted. Using the serial provincial crop yield data from 1949 to 2012 and the probability density function algorithm, the meteorological disaster risk of southern China is analyzed. First, the trend yields of various crops are extracted, then the probability density function curves and distribution functions of the relative meteorological yields are constructed and calculated. The probabilities of different yield decrease sections are also estimated. Finally, the risk levels associated with agriculture and various crops suffering from an agricultural disaster are also assessed. The results show that the variations in trend yields can be divided into three stages and it increased greatly from 1962 to 2004. The food, autumn food and summer food yield has increased each year at a mean rate of 48.3, 50.55 and 36.6 kg/ha a. The increasing rates for the trend yields in Yunnan and Guizhou Provinces are slower than the rates for the other provinces. Additionally, the increasing rate in south China is higher than that of southwest China. Additionally, the yields of all food grains and economic crops in southern China have increased each year. Additionally, the relative meteorological yield of the primary crops possessed distinct stochastic, fluctuating features. The relative meteorological yield was reduced year by year for every province. During the early period, the relative meteorological yield clearly fluctuated, but in recent years, this change has been small. The fluctuation extents for different crops or regions exhibited obvious differences. For example, in some regions or during some years, the fluctuant extents of the relative meteorological yields are large, which suggests that the risk level is high during this year and in this area. The rice yield fluctuation at a mean value of 3.7 % is the smallest and is relatively stable. In general, after suffering an agricultural disaster, the yield increase or decrease section for most crops are primarily concentrated in an interval from -10 to 10 %, but for some crops or individual areas, the probabilities of large losses are relatively high, which suggests that the risk level in this area is high, and the ability to prevent agricultural disasters there should be improved. Among these provinces, the agricultural risk rate of Guizhou Province reaches 20 %, and this province possesses the highest risk associated with grain production, but it is also the most unstable. The agricultural risks for Sichuan and Yunnan Provinces are lower, and they show a skewed distribution and an increasing production trend. All of these results suggest that the agricultural costs in Guizhou Province are higher than the costs in other provinces. The results also provide a scientific basis for agricultural production and for government decision-making in relation to disaster prevention and mitigation.