Study of Space-time Regression Prediction on Occurrence Quantity of Dendrolimus punctatus Walker
The study was to provide the theoretical basis for controlling the damage of Dendrolimus punctatus Walker. [Method] With pest investigation data and meteorological data of Xianju Country Zhejiang Province in 1 983.9-1 989.9 as the researched object, the dynamic variation of in-sect appearance areas of D. punctatus with the change of meteorological factors, time and spatial in each investigation point was studied. ,and then the space-time regression prediction model on occurrence quantity of D. punctatus was established. The insect appearance areas of 3 generation D. punctatus in 1989 was forecasted and tested with the established model. [Result] Totally 1 linear and 8 nonlinear space-time regression prediction models on occur-rence quantity of D. punctatus were established. The actual villages that damaged by 3 generation D. punctatus and the insect appearance areas were 141, 143, 138 and 4 533.1, 4 181.3, 5 166.0 hm~2 resp. and the villages that damaged by 3 generation D. punctatus predicted by 9 models were 193, and the insect appearance areas predicted by linear model were 3 978.5, 4 159.5, 3 427.5 hm~2 resp. the prediction results of nonlinear model on insect appear-ance areas of 3 generation D. punctatus were 71.686 8, 99.835 9, 119.947 8 hm~2 different with the actual.[Conclusion] The prediction accuracy of the established models on insect appearance areas of D. punctatus were all 66.7% .
国家科技期刊平台
NSTL
万方数据
维普
