基于IDRISI降雨量的时间序列分析与预测

Time Series Analysis and Forecast of Rainfall Based on IDRISI

笔者以全球1982年至2000年5月、6月、9月的降水量数据为研究对象,分别从空间尺度和时间尺度对此期间降雨量进行分析,并运用全球趋势模型对未来降雨量进行预测。结果表明,1982年至2000年,全球这3个月的降雨量普遍增加,但是增量不大;5月和6月的降雨量的差值在长时间序列上基本保持稳定,9月份的降雨量在时间序列上明显增加。全球6月份降雨量的预测公式为y=5sin[π/2(T-1982)].以此预测的降雨量数据与实际数据差异不大,且变化趋势相似。

The global precipitation data of May,June and September from 1982 to 2000 were taken as the research object.The precipitation during this period was analyzed from the spatial and temporal scales,and the future rainfall was predicted using the global trend model.Results showed that the global precipitation of these 3 months from 1982 to 2000 had generally increased,but the increment was not large.The difference between May and June rainfall was basically stable in the long time series,and the rainfall in September was significantly increased in the time series.The prediction formula of global rainfall in June was y=5sin[π/2(T-1982)].The difference between the predicted rainfall data and the actual data was not significant,and the change trend was similar.