甘肃小陇山林区位于我国四大自然植被区系(华北、华中、喜马拉雅、蒙新)交汇处的温带向亚热带过渡带,林业植物种类繁多、茂密,但区内地形、地貌复杂,气候差异大,林业生物灾害频繁,环境气候条件是其发生发展的主要制约因素。为此,进行区内林业生物灾害的气象预测预报技术研究,对林业有害生物灾害的预防、防治和林业产业发展具有显著的指导意义。利用逐步回归统计方法统计分析了影响甘肃小陇山林区甘肃鼢鼠危害面积的主要气象因子,建立了林区春、秋两季甘肃鼢鼠危害面积分区集成与全区综合统计预测预报数学模型。甘肃鼢鼠适宜温暖、较为干燥的气候环境,怕光、怕雨、怕高温高湿天气,林区暖冬天气气候和春、秋季温暖少雨天气、初夏较好的日照和适宜的温度条件配置是造成甘肃小陇山林区甘肃鼢鼠发展蔓延的主要气候原因。通过预测效果历史回代拟合检验,春、秋季16 a平均预测准确率均在97%以上。通过对2008年、2009年和2010年甘肃鼢鼠春、秋季危害面积进行试报检验,3 a平均试报误差分别为1.8、2.0和1.9、1.6,平均试报准确率均达98%以上,预测和试报效果均较理想,2种预报方法均可满足当前业务服务需求。 The Xiaolongshan forest in Gansu is located in the transition zone from temperate zone to subtropical zone at the confluence of the four major natural vegetation areas (North China, Central China, Himalayas and Mengxin). There are many kinds of forestry plants in the area, but the terrain and landforms are complex and the climate difference is large The frequent occurrence of forestry biological disasters and the climatic and climatic conditions are the main constraints on their occurrence and development. Therefore, carrying out the meteorological forecasting and forecasting technology of forestry biological disasters in the region has significant guiding significance for the prevention, prevention and control of forest pest disasters and the development of forestry industry. The main meteorological factors influencing the damage area of ​​Zokor in Gansu Xiaolongshan forest area were statistically analyzed by using stepwise regression method. The integrated mathematical forecast model of zokor area in Gansu Province in spring and autumn was established. Gansu Zokor suitable for warm, dry climate, fear of light, fear of rain, afraid of hot and humid weather, warm winter forest climate and spring, autumn warm and dry weather, good summer sunshine and the appropriate temperature conditions are caused by configuration The Main Climatic Reasons for the Development and Spread of Zokor in Gansu Xiaolongshan Forest Region. By predicting the effect history back to the fitting test, the average prediction accuracy of spring and autumn 16 a was above 97%. Through the test report on the damage area of ​​zokor spring and autumn in 2008, 2009 and 2010 in Gansu Province, the average test reporting errors for 3 years were 1.8, 2.0 and 1.9, 1.6 respectively, and the average accuracy of test reports reached more than 98% Forecast and test results are better, two kinds of forecasting methods can meet the current business needs.