不同尺度农业预测差异及其处理

DIFFERENCES BETWEEN AGRICULTURE FORECASTING BY DIFFERENT MEASUREMENTS & THEIR TREATMENTS

不同尺度农业预测差异是指在进行区域及其各个亚区域农业预测时,区域尺度预测结果和各个亚区域尺度预测结果之和不相等的现象。该文对此差异的产生原因进行分析后指出,历史资料的波动性是普遍存在的,历史资料拟合值的差异是导致不同尺度农业预测差异的原因。在此基础上,系统提出了以大尺度预测结果为主调整小尺度预测结果的原则,按照逐步寻优的思路,通过选择或组合应用排序修正、相关修正、增率修正、结构修正、布局修正、政策修正和趋势修正等处理方法解决不同尺度预测差异问题。

Differences between agriculture forecasting by different measurements mean the phenomenon that when conducting agriculture forecast on various regions ＆ sub- regions, the forecast results by regional measurement are different from the sum of the forecast results by sub- regional measurement. After analyzing the causes of producing these differences, this paper points out that the fluctuation of historical materials exists universally, because historical materials are of matching value. The diversity is the cause for differences in agriculture forecast by different measurements. On this basis, this paper systematically puts forward the principle of taking the forecast results by large measurement as the major and adjusting the forecast results by small measurement. According to the thought of gradually selecting the superior one, revising by order, relative revise, rate increasing revise, structure revise, allocation revise, policy revise, tendency revise and treatment methods are adopted through selection or combination to solve the differential problems of different measurements.