[关键词]
[摘要]
利用1981 年- 2010 年西北地区夏季降水观测数据和MEERA( Mo dem2Er a Retr ospect ive Analysis for Re2 search and Applicatio ns) 再分析资料, 耦合DEM( Digital Elevat ion M odel) 数据, 采用基于区域分月的多元回归算法, 建立模型进行降水估算。结果表明, 6 月、7 月、8 月和夏季平均绝对误差分别为8. 0 mm、8. 0 mm、7. 9 mm、20. 8 mm, 平均相对误差分别为19. 0%、14. 4%、22. 2%、15. 0%, 模型估算结果较MERRA 降水资料误差明显减小。模 型估算结果不仅保持着原始MERRA 降水产品的分布特征, 而且可以精细反映局部降水分布特征。
[Key word]
[Abstract]
In this paper , a monthly multiple reg ressio n alg o rithm for the certain r eg io n w as used to establish a comput at ional model for the precipitat ion est imatio n based on the summer precipit ation observat ion data, MERRA (M odem2Er a Retro spectiv e Analysis fo r Research and Applicat ions) reanaly sis, and the DEM ( Dig ital Elevatio n Mo del) data fr om 198122010 in the No rth2 w estern Regions. The results show ed that the av erage absolute err or s o f estimat ed precipitation ar e 8. 0 mm, 8. 0 mm, 7. 9 mm, and 20. 8 mm in June, July, Aug ust, and the w hole summer , respectively, and the aver age relative err or s ar e 19. 0%, 14. 4%, 221 2% and 15. 0%, r espectively. The estimated precipitat ion has few er err or s compared w ith t hat from M ERRA analy sis. Mo re2 o ver, the mo del estimates o f pr ecipitation have the same dist ribution cha racterist ics as those o f the o rig inal MERRA precipitat i2 o n pro duct, and can also reflect the local distr ibution.
[中图分类号]
[基金项目]
国家自然科学基金项目( 40801145) ; 江苏高校优势学科建设工程项目; 南京师范大学虚拟地理环境教育部重点实验室开放基金