| Focus on the oilseed rape waterlogging in Hunan,this paper inverts the waterlogging process samples from historical disaster information and builds the impending hazard sample set and affected sample set.As the disaster-causing factors in different status have significant difference,T-Test is used to quantify the significance of precipitation factor difference in different attenuation coefficients and create Effective Accumulation Preparation(PE).According to PE and continuous rain days(D),the critical line is derived from Fisher principle to build oilseed rape vernal waterlogging catastrophe distinction index which can distinguish affected status day by day.The waterlogging catastrophe index is verified by inverting waterlogging process samples and calculating the value of independent samples.Based on catastrophe index,this paper performed a regression analysis to examine the relationship between yield and level of waterlogging and find the thresholds of light catastrophe and severe catastrophe.Then the time-space analysis of frequency to the area affected has been done.After that two waterlogging affected index has been built based on the relationship of yield and level of waterlogging.The results show that:(1)When the attenuation coefficient is 0.899,the difference between the PE of the impending hazard sample set and affected sample set is the most significant.The oilseed rape vernal waterlogging catastrophe distinction index I can be used to monitored the waterlogged catastrophe process on a daily basis,and show the catastrophe time,the intensity of disaster and the aftermath of disasters.(2)The waterlogging catastrophe thresholds of not affected,light catastrophe and severe catastrophe are I=1 and I=1.44.The thresholds of not affected and light catastropher is the lower limit ofaffected,which represents the risk of waterlogging disaster.The thresholds of light catastropher and light catastropher is the lower limit of damaged,which is related to frequency to the area affected,terrain and disaster prevention and mitigation capabilities.(3)The waterlogging-affected frequency of oilseed rapeat different levels in Hunan allemerges an overall trend of high in southeast and low in northwest from 1961 to 2008.The frequency to the area affected increased first,then decreased and existed mutation in flowering period.In mature period,the frequency increased.(4)The effect weights of affected days in flowering,fruiting and mature period in waterlogging affected index M1 is 0.16,0.66 and 0.18.The effect weights of damage days in flowering,fruiting and mature period in waterlogging affected index M2 is 0.11,0.75 and 0.14.M2 is more precise.The decreasing rate of oilseed rape yield in couty has a similar patial distribution with M1 and M2.M1 and M2 are applicative in Hunan. |
