基于风险预测的山地典型农业生长期干旱指标筛选——以贵州遵义为例

Selection of a drought index based on risk forecast in the typical mountainous agricultural growth periods: a case study of Zunyi,Guizhou Province

为探明适宜贵州的干旱指标,以便更准确的预报干旱,本文以遵义为例,统计分析了1961-2012年这52年的逐日降水气温数据。通过分析0、5、10、15和20℃5个典型农业生长期特征,分别构建了降水距平百分率(Pa)和干燥度(K)两种类型的干旱指标序列并进行旱涝状态的长期预报,根据旱涝风险预报的准确度来筛选典型农业生长期的干旱指标。分析得到了不同农业生长期内旱涝状态界限和两种类型干旱指标对应下的旱涝状态预测准确率,结果表明对于遵义地区来讲,除0℃生长期外,其余4个典型农业生长期内Pa指标作为干旱指标进行旱涝灾害预测时更准确。

In order to select a drought index for correctly forecasting drought in the typical mountainous agricultural growth periods,the data of daily precipitation and temperature from 1961- 2012 were analyzed for Zunyi,Guizhou Province.Through the analysis of the characteristics in five typical agricultural growth periods which are represented by 0℃,5℃,10℃,15℃ and 20℃,the sequence of two drought indexes,i.e. precipitation anomaly percentage（ Pa） and dry degree（ K） were established and used to predict the risk of drought. According to the forecast accuracy of the risk of drought,more suitable drought index was selected. Our analysis produced the boundaries of drought and the forecast accuracy in comparing the two indexes in predicting risk of drought in different agricultural growth periods. The results indicated that,apart from 0℃ growth period,the index of precipitation anomaly percentage（ Pa） is more accurate and suitable to forecast the risk of drought for the other four growth periods in Zunyi.