[1]吴小君,方秀琴,任立良,等.基于随机森林的山洪灾害风险评估——以江西省为例[J].水土保持研究,2018,25(03):142-149.
　WU Xiaojun,FANG Xiuqin,REN Liliang,et al.Risk Assessment of Mountain Torrents Disaster Based on Random Forest—A Case Study in Jiangxi Province[J].Research of Soil and Water Conservation,2018,25(03):142-149.
点击复制

基于随机森林的山洪灾害风险评估——以江西省为例

《水土保持研究》[ISSN:1005-3409/CN:61-1272/P] 卷: 25 期数: 2018年03期 页码: 142-149 栏目: 出版日期: 2018-04-10
Title:
Risk Assessment of Mountain Torrents Disaster Based on Random Forest—A Case Study in Jiangxi Province
作者:
吴小君1, 方秀琴1, 任立良2, 吴陶樱1, 苗月鲜1
1. 河海大学 地球科学与工程学院, 南京 211100;
2. 河海大学 水文水资源与水利工程科学国家重点实验室, 南京 210098)
Author(s):
WU Xiaojun1, FANG Xiuqin1, REN Liliang2, WU Taoying1, MIAO Yuexian1
1. College of Earth Science and Engineering, Hohai University, Nanjing 211100, China;
2. State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Hohai University, Nanjing 210098, China
关键词:
山洪灾害; 风险评估; 随机森林; GIS
Keywords:
mountain torrent disaster; risk assessment; random forest; GIS technology
分类号:
X43;S422
摘要:
依据流域灾害系统理论，以江西省山洪灾害的特点为基础，从触发因子、下垫面孕灾环境和承灾体3个角度确定影响山洪灾害的9个评价指标。将历史发生山洪灾害点的正样本集和5组随机采样点的负样本集组成不同的总样本集，并进行训练样本和测试样本的随机分组。通过训练样本构建了基于随机森林的模型，并通过测试样本对模型进行检验，平均精度为86.26%，表明构建的随机森林模型具有较好的拟合效果。对山洪灾害风险图在行政分区上进行了风险高值区面积绝对值和面积占比的定量统计分析，结果发现景德镇和上饶市山洪灾害风险度较高和很高的区域不仅面积占比大，所占面积绝对值也很大，属于山洪灾害高发区域。
Abstract:
According to the theory of mountain torrent disaster system, based on the characteristics of mountain torrent disaster in Jiangxi Province, nine factors were selected to determine the impact of mountain torrent disaster from three aspects including trigger factors, the underlying surface of the disaster environment and the relief body. We took the historical data of the mountain torrent disaster as the positive sample set, five groups of random samples as a negative sample set to form the total sample set, and randomized the training samples and test samples, respectively. Through training samples, the estimation models were set up based on the random forest. Accuracy assessment showed an average accuracy of 86.26%, which indicated the satisfying performance of random forest model. Based on the mapping of mountain torrent disaster risk assessment, statistical analyses were carried out according to different administrative areas. The results showed that Jingdezhen and Shangrao were the most vulnerable to mountain flood disaster as they had both relative and absolute larger values in the higher or greater area of the risk degree of the mountain flood disaster.

参考文献/References:

[1] 管珉,陈兴旺.江西省山洪灾害风险区划初步研究[J].暴雨灾害,2007,26(4):339-343.
[2] 赵士鹏.中国山洪灾害系统的整体特征及其危险度区划的初步研究[J].自然灾害学报,1996(3):93-99.
[3] 马建华,胡维忠.我国山洪灾害防灾形势及防治对策[J].人民长江,2005,36(6):3-5.
[4] 孟东勇,秦亚丽.山洪灾害防治非工程措施项目建设探讨[J].江西建材,2014(6):82-82.
[5] 李林涛,徐宗学,庞博,等.中国洪灾风险区划研究[J].水利学报,2012,43(1):22-30.
[6] Woodruff J D, Irish J L, Camargo S J. Coastal flooding by tropical cyclones and sea-level rise[J]. Nature, 2013,504(7478):44-52.
[7] Hallegatte S, Green C, Nicholls R J, et al. Future flood losses in major coastal cities[J]. Nature Climate Change, 2013,3(9):802-806.
[8] 刘小平,黎夏,叶嘉安,等.利用蚁群智能挖掘地理元胞自动机的转换规则[J].中国科学:地球科学,2007,37(6):824-834.
[9] Martens D, Backer M D, Haesen R, et al. Classification with ant colony optimization[J]. Ieee Transactions on Evolutionary Computation, 2007,11(5):651-665.
[10] 赵铜铁钢,杨大文,蔡喜明,等.基于随机森林模型的长江上游枯水期径流预报研究[J].水力发电学报,2012,31(3):18-24.
[11] Dong L J, Xi-Bing L I, Peng K. Prediction of rockburst classification using Random Forest[J]. Transactions of Nonferrous Metals Society of China, 2013,23(2):472-477.
[12] Tesfamariam S, Liu Z. Earthquake induced damage classification for reinforced concrete buildings[J]. Structural Safety, 2010,32(2):154-164.
[13] Chen X, Ishwaran H. Random forests for genomic data analysis[J]. Genomics, 2012,99(6):323-329.
[14] MihaiIlescu D M, Gui V, Toma C I, et al. Computer aided diagnosis method for steatosis rating in ultrasound images using random forests[J]. Medical Ultrasonography, 2013,15(15):184-190.
[15] 赖成光,陈晓宏,赵仕威,等.基于随机森林的洪灾风险评价模型及其应用[J].水利学报,2015,46(1):58-66.
[16] Wang Z, Lai C, Chen X, et al. Flood hazard risk assessment model based on random forest[J]. Journal of Hydrology, 2015,527:1130-1141.
[17] Feng Q, Liu J, Gong J. Urban flood mapping based on unmanned aerial vehicle remote sensing and random forest classifier:A Case of Yuyao, China[J]. Water, 2015,7(4):1437-1455.
[18] 刘筱琴,李昆.江西省暴雨洪水地理变化规律研究[R].华东七省(市)水利学会第二十次学术研讨会,2007.
[19] 李世勤,邱启勇,王述强.江西山洪灾害防治实践及思考[J].中国水利,2012(3):51-54.
[20] 李明辉,熊剑英.江西省山洪灾害防治规划概述[J].江西水利科技,2005,31(2):73-77.
[21] 林俊.江西省山洪预警系统分析与设计[D].昆明:云南大学,2012.
[22] 杨妩,柏林, YANG W,等.浅谈江西省山洪灾害防治监测、通信和预警系统的规划[J].江西水利科技,2006,32(2):83-85.
[23] 程晓陶.城市型水灾害及其综合治水方略[J].灾害学,2010(S):10-15.
[24] 王全才.随机森林特征选择[D].辽宁大连:大连理工大学,2011.
[25] 吴喜之.应用回归及分类:基于R[M].北京:中国人民大学出版社,2016.
[26] 尼娜·朱梅尔,约翰·芒特, NinaZumel,等.数据科学:理论、方法与R语言实践[M].北京:机械工业出版社,2016.
[27] 李欣海.用R实现随机森林的分类与回归[R].北京:第五届R语言会议,2012.
[28] 赵北庚.基于R语言randomForest包的随机森林建模研究[J].计算机光盘软件与应用,2015(2):152-153.
[29] 方秀琴,王凯,任立良,等.基于GIS的江西省山洪灾害风险评价与分区[J].灾害学,2017,32(1):111-116.

相似文献/References:

[1]宾建华,窦新英.乌鲁木齐市小流域山洪灾害临界雨量分区及防治区划研究[J].水土保持研究,2005,12(05):252.
　BIN Jian-hua,DOU Xin-ying.Study on the Critical Rainfall Distribution and Prevention and Control Regionalization of Mountain Flood Disasters in Urumqi[J].Research of Soil and Water Conservation,2005,12(03):252.
[2]郭跃.自然灾害的风险特征及风险管理模型的探讨[J].水土保持研究,2006,13(04):15.
　GUO Yue.The Conceptual Model for Natural Hazards Risk Management[J].Research of Soil and Water Conservation,2006,13(03):15.
[3]王凯博,高保玲,李洪滨,等.博州山洪灾害防治对策探讨[J].水土保持研究,2007,14(01):289.
　WANG Kai-bo,GAO Bao-ling,LI Hong-bin,et al.Disscussion on the Prevention and Control Measures of Mountain Torrents in Bozhou District[J].Research of Soil and Water Conservation,2007,14(03):289.
[4]朱静,唐川.城市山洪灾害风险管理体系探讨[J].水土保持研究,2007,14(06):407.
　ZHU Jing,TANG Chuan.Discussion on Risk Management System for Urban Flash Flood[J].Research of Soil and Water Conservation,2007,14(03):407.
[5]陈香.福建省暴雨洪涝灾害风险评估与管理[J].水土保持研究,2007,14(04):180.
　CHEN Xiang.Assessment and Management of Rainstorm-flood Hazard Risk in Fujian Province[J].Research of Soil and Water Conservation,2007,14(03):180.
[6]张茂省,薛富平,王晓勇.陕西省山洪灾害防治区划[J].水土保持研究,2005,12(02):163.
　ZHANG Mao-sheng,XUE Fu-ping,WANG Xiao-yong.Features on Torrential Floods and Its Precaution in Shaanxi Province[J].Research of Soil and Water Conservation,2005,12(03):163.
[7]赵宗权,周亮广.江淮分水岭地区旱灾风险评估[J].水土保持研究,2017,24(01):370.
　ZHAO Zongquan,ZHOU Liangguang.Assessment on Drought Disaster Risk in Jianghuai Watershed Region[J].Research of Soil and Water Conservation,2017,24(03):370.
[8]王春山,巴仁基,周洪福,等.四川省石棉县县城泥石流灾害风险评估[J].水土保持研究,2014,21(04):219.
　WANG Chun-shan,BA Ren-ji,ZHOU Hong-fu,et al.Risk Evaluation of Debris Flows in Urban Area of Shimian County of Sichuan Province[J].Research of Soil and Water Conservation,2014,21(03):219.
[9]林孝松,崔梦瑞,牟凤云,等.基于多情景的巫山县山洪灾害危险评价[J].水土保持研究,2018,25(06):370.
　LIN Xiaosong,CUI Mengrui,MOU Fengyun,et al.Assessment of Hazard of Flash Flood Disaster in Wushan County Based on Multi-Scenarios[J].Research of Soil and Water Conservation,2018,25(03):370.
[10]陈楷俊,陈艺仪,陈菁.近30年粤东地区台风灾害风险评估与分析[J].水土保持研究,2019,26(03):362.
　CHEN Kaijun,CHEN Yiyi,CHEN Jing.Risk Assessment and Analysis of Typhoon Disaster in East Guangdong in the Period 1989-2017[J].Research of Soil and Water Conservation,2019,26(03):362.

备注/Memo

收稿日期:2017-07-19;改回日期:2017-08-18。
基金项目:“十三五”国家重点研发计划项目“陆地水循环演变及其在全球变化中的作用研究”（2016YFA0601500）；中央高校基本科研业务费专项资金资助（2016B05414）
作者简介:吴小君(1991-),女,安徽阜阳人,研究生,从事遥感方面研究。E-mail:2360868477@qq.com
通讯作者:方秀琴(1978-),女,安徽池州人,博士,从事地表参数的遥感反演及遥感、地理信息系统和水文模型的集成等方面的研究。E-mail:kinkinfang@hhu.edu.cn

更新日期/Last Update: 1900-01-01