| 基于CART算法的夏季干旱预测模型研究及应用 |
| |
| 引用本文： | 王伟,薛丰昌,史达伟,王小洁.基于CART算法的夏季干旱预测模型研究及应用[J].气象科学,2016,36(5):661-666. |
| |
| 作者姓名： | 王伟 薛丰昌 史达伟 王小洁 |
| |
| 作者单位： | 南京信息工程大学 大气科学学院, 南京 210044,南京信息工程大学 地理与遥感学院, 南京 210044,连云港市气象局, 江苏 连云港 222006,南京信息工程大学 大气科学学院, 南京 210044 |
| |
| 基金项目： | 国家自然科学基金资助项目（41375160）；江苏省博士后科研基金项目（1101024B） |
| |
| 摘 要： | 利用标准化降水蒸散指数(SPEI)作为判断干旱是否发生的标准,基于数据挖掘方法中的CART算法探究夏季西太平洋副高北界、夏季西太平洋副高强度指数、准两年振荡(QBO)、东亚夏季风指数、夏季北大西洋涛动(NAO)、夏季太平洋年代际振荡(PDO)、厄尔尼诺指数等多项气候因子与干旱的关系,构造分类决策树,得到干旱预报规则集,从而建立干旱的预报模型。预报模型以各项气候因子为输入变量,是否干旱为目标变量。根据1955—2012年商丘月平均气温和月总降水资料计算出商丘夏季58 a的SPEI指数作为干旱判定指标;以同期的多项气候因子数据作为输入变量,随机选取46 a的数据得到7条分类规则集,分类准确率为86.96%。使用剩余12 a的数据验证,准确率高达91.67%。结果有力地证明了基于CART算法建立干旱预报模型的可行性、科学性、有效性以及与干旱研究理论的一致性,为干旱模型的研究及季节性预测提供了科学有效的新思路。
|
| 关 键 词： | 标准化降水蒸散指数 干旱 CART算法 |
| 收稿时间： | 5/7/2015 12:00:00 AM |
| 修稿时间： | 2015/8/13 0:00:00 |
Research on summer drought prediction model based on CART algorithm |
| |
| WANG Wei,XUE Fengchang,SHI Dawei and WANG Xiaojie.Research on summer drought prediction model based on CART algorithm[J].Scientia Meteorologica Sinica,2016,36(5):661-666. |
| |
| Authors: | WANG Wei XUE Fengchang SHI Dawei and WANG Xiaojie |
| |
| Institution: | School of Atmospheric Sciences, Nanjing University of Information Science & Technology, Nanjing 210044, China,School of Geography and Remote Sensing, Nanjing University of information Science & Technology, Nanjing 210044, China,Lianyungang Meteorological Bureau, Jiangsu Lianyungang 222006, China and School of Atmospheric Sciences, Nanjing University of Information Science & Technology, Nanjing 210044, China |
| |
| Abstract: | By using the Standardized Precipitation Evapotranspiration Index (SPEI) as the judgment standard of drought occurrence, based on the CART algorithm of data mining method, the relation between drought and multiple climate factors such as the north boundary of the summer western Pacific subtropical high, summer western Pacific subtropical high intensity index,Quasi-Biennial Oscillation (QBO), the summer East Asian monsoon index and summer North Atlantic Oscillation (NAO), summer Pacific Decadal Oscillation (PDO), El Niño index and so on, was analyzed. Moreover, the classification decision tree was built and the prediction rules set of drought, the prediction model of drought was developed. The climatic factors included as input variables, and the result whether drought appears or not is considered as the output variable. Taking summer in Shangqiu for example, SPEI index of 58 years were calculated by the 1955-2012 monthly precipitation and monthry mean temperature data of Shangqiu as the drought index; multiple climatic factors in the same period were used as input variables. The random data of 46 years was used to get 7 classification rules, whose accuracy is 86.96%. Then, if all the data of 58 years was included, the results show the accuracy reached 91.67%. The results present that the drought prediction model established by CART algorithm provides a new idea for drought research and seasonal prediction. |
| |
| Keywords: | SPEI Drought CART algorithm |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《气象科学》浏览原始摘要信息 |
| 点击此处可从《气象科学》下载免费的PDF全文 |
|
