摘要
介绍了Casati于2004年提出的强度尺度检验技术,并将该方法应用到上海区域中尺度模式的预报检验中。结果表明,强度尺度检验技术可以提供一种从不同降水率阈值和空间尺度对模式定量降水预报进行评估的方案。该方法首先对预报场和观测场进行降水率阈值获得二进制误差场,并通过二维离散Haar小波对二进制误差场进行各个空间尺度的分解,从而评估模式在各个降水率阈值和空间尺度上的均方差和预报技巧。通过计算各空间尺度二进制误差场的均方差百分比和能量平方偏差,可分析不同的降水特征尺度误差对总体误差的贡献情况以及预报降水与观测的偏差情况。
An and the method is provides a way of intensity-scale method developed by Casati in 2004 is introduced in detail in this paper applied to the Shanghai regional mesoscale model. The results show that this technique evaluating precipitation forecast at different precipitation rate intensities and differentspatial scales. The binary error images between forecast and analysis is obtained by thresholding at differ- ent precipitation rate intensities, the mean square deviation and skill score is assessed on different spatial scales using a two-dimensional discrete Haar wavelet decomposition of binary images. By computing the mean square deviation percentage and energy squared bias of different scale binary error images, the error contribution of different spatial scale and bias between forecast and observation can be analyzed.
出处
《气象与环境科学》
2012年第1期1-7,共7页
Meteorological and Environmental Sciences
基金
2010年公益性行业(气象)科研专项"强对流天气短时临近预报非常规检验新技术研发"(GYHY20100600)资助
关键词
小波
尺度分解
均方差
检验
wavelet
scale analysis
mean square deviation
verification
