黄土丘陵区土壤侵蚀因子敏感性分析
|
摘要
为探究黄土丘陵区土地利用类型、地形及降雨等因素对土壤侵蚀的影响程度,以彭阳县1995、2005、2015年土地利用、降雨量、DEM(5 m)等为数据源,选取高建堡、虎沟等11个小流域作为试验区,采用修正通用土壤流失方程(RUSLE)的算法,计算各试验区降雨侵蚀力、坡度坡长等5个土壤侵蚀影响因子和土壤侵蚀模数;基于BP神经网络模型(BPNN),构建以各试验区降雨侵蚀力、坡度坡长等因子为输入变量,侵蚀模数为输出变量的关系模型,预测、验证模型的有效性。结果如下:1) BP神经网络模型能够有效预测土壤侵蚀影响因子的显著性; 2)研究区小流域尺度上,地形因子对土壤侵蚀的显著性最强,土壤可蚀性因子的显著性最弱; 3)在时间尺度上,小流域土壤侵蚀影响因子显著性略有差异,降雨侵蚀力因子的显著性通过降雨量体现出来;水土保持措施因子和植被覆盖与管理因子的显著性与退耕还林(草)工程等生态建设项目的实施有关,在2005年体现出对土壤侵蚀的抑制性; 4) 2015年的显著性预测结果适用于以生态自然恢复的区域,土壤侵蚀影响因子的显著性表现为:SL> P> R> C> K。研究结果表明,基于BP神经网络模型预测土壤侵蚀影响因子显著性的方法适用于黄土丘陵区,可为后续小流域综合治理提供科学依据。
[Background]The terrain of loess hilly gully region is complex and varied,and soil erosion is serious. Soil erosion is a nonlinear system,influenced by various uncertain factors,such as soil,vegetation,terrain and others,the result of erosion is very complicated. [Methods]In order to explore the effects of land use types,topography,rainfall and other soil erosion factors in the loess hilly gully region,take Pengyang county as an example,based on the data source in 1995,2005,2015,such as land use,annual rainfall,DEM( 5 m),etc. In the ecological zoning of Pengyang county,11 small watershed areas such as Gaojianpu and Hugou were selected as experimental zones. Firstly,the revised universal soil loss equation( RUSLE) was adopted to calculate the soil erosion modulus and five influence factors,such as rainfall erosion force,slope and slope length in each small watershed. Then the BP neural network method was applied to construct the relational model. The five influence soil erosion factors in various small watershed as input variables,soil erosion modulus as the output variable,and finally the validity of the model was predicted and verified. [Results] 1) BP neural network model effectively predicted the significance of soil erosion influencing factors. 2) In the small watershed scale of the study area,topographic factors had the strongest significance to soil erosion,while soil erodibility factors had the weakest significance to soil erosion. 3) On the time scale,there was a slight difference between soil erosion influnence factors in small watershed. The rainfall erosion factor was reflected by rainfall,the significance of soil and water conservation measures factor and the vegetation cover and management factor were related to the implementation of ecological construction projects such as the project of returning cultivated land to forest( grass). In 2005,soil and water conservation measures factor and vegetation cover and management factors showed the inhibition of soil erosion. 4) The significance prediction results of 2015 were applicable to the study area with ecological natural restoration,and the significance of soil erosion influencing factors was as follows: SL > P > R > C > K.[Conclusions] The results show the method based on BP neural network model to predict the significance of soil erosion impact factor is applicable to loess hilly region,and it can provide scientific basis for the follow-up comprehensive governance of small watershed.
[Background]The terrain of loess hilly gully region is complex and varied,and soil erosion is serious. Soil erosion is a nonlinear system,influenced by various uncertain factors,such as soil,vegetation,terrain and others,the result of erosion is very complicated. [Methods]In order to explore the effects of land use types,topography,rainfall and other soil erosion factors in the loess hilly gully region,take Pengyang county as an example,based on the data source in 1995,2005,2015,such as land use,annual rainfall,DEM( 5 m),etc. In the ecological zoning of Pengyang county,11 small watershed areas such as Gaojianpu and Hugou were selected as experimental zones. Firstly,the revised universal soil loss equation( RUSLE) was adopted to calculate the soil erosion modulus and five influence factors,such as rainfall erosion force,slope and slope length in each small watershed. Then the BP neural network method was applied to construct the relational model. The five influence soil erosion factors in various small watershed as input variables,soil erosion modulus as the output variable,and finally the validity of the model was predicted and verified. [Results] 1) BP neural network model effectively predicted the significance of soil erosion influencing factors. 2) In the small watershed scale of the study area,topographic factors had the strongest significance to soil erosion,while soil erodibility factors had the weakest significance to soil erosion. 3) On the time scale,there was a slight difference between soil erosion influnence factors in small watershed. The rainfall erosion factor was reflected by rainfall,the significance of soil and water conservation measures factor and the vegetation cover and management factor were related to the implementation of ecological construction projects such as the project of returning cultivated land to forest( grass). In 2005,soil and water conservation measures factor and vegetation cover and management factors showed the inhibition of soil erosion. 4) The significance prediction results of 2015 were applicable to the study area with ecological natural restoration,and the significance of soil erosion influencing factors was as follows: SL > P > R > C > K.[Conclusions] The results show the method based on BP neural network model to predict the significance of soil erosion impact factor is applicable to loess hilly region,and it can provide scientific basis for the follow-up comprehensive governance of small watershed.
引文
[1]GHOBADI M,TAHERABADI S,GHOBADI M E,et al.Antioxidant capacity,photosynthetic characteristics and water relations of sunflower(Helianthus annuus L.)cultivars in response to drought stress[J].Industrial Crops and Products,2013,50(5):29.
[2]HESSEL R.Consequences of hyperconcenttrated flow for process-based soil erosion modeling on the Chinese Loess Plateau[J].Earth Surface Processes and Landforms,2006,31(9):1100.
[3]XU J X.Hyperconcentrated flows in the slope-channel systems in gullied hilly areas on the loess plateau,china[J].Geografiska Annaler:Series A,Physical Geography,2004,86(4):349.
[4]XU J X.Erosion caused by hyperconcentrated flow on the Loess Plateau[J].Catena,1999,36(1-2):1.
[5]郑粉莉,刘峰,杨勤科,等.土壤侵蚀预报模型研究进展[J].水土保持通报,2001,21(6):16.ZHENG Fenli,LIU Feng,YANG Qinke,et al.Research progress in soil erosion prediction model[J].Soil and Water Conservation Bulletin,2001,21(6):16.
[6]SMITH D D.Interpretation of soil conservation data for field use[J].Agricultural Engineering,1941,22:173.
[7]WISCHMEIER W H,SMITH D D.Predicting rainfall erosion losses from cropland east of the rocky mountains[M].USDA Agricultural Handbook,No.292,1965.
[8]WISCHMEIER W H,SMITH D D.Predicting rainfall erosion losses:A guide to conservation planning[M].US-DA Agricultural Handbook,No.537,1978:10(11),28.
[9]方开泰.实用多元统计分析(上)[M].上海:华东师范大学出版社,1989:145.FANG Kaitai.Practical multivariate statistical analysis(upper)[M].Shanghai:East China Normal University Press,1989:145.
[10]李扬,王伯昕,陈冬昕,等.基于BP神经网络预测复合盐侵蚀后混凝土的相对动弹性模量[J].混凝土,2018(7):21.LI Yang,WANG Bxin,CHEN Dongxin,et al.Prediction of relative dynamic elastic modulus of concrete after composite based on BP neural network theory[J].Concrete,2018(7):21.
[11]王尧,蔡运龙,潘懋,等.贵州省乌江流域土壤侵蚀模拟:基于GIS、RUSLE和ANN技术的研究[J].中国地质,2014,41(5):1735.WANG Yao,CAI Yunlong,PAN Mao,et al.Soil erosion simulation in Wujiang river basin,Guizhou province:Based on GIS,RUSLE and ANN technology[J].China geology,2014,41(5):1735.
[12]郭成久,颉丽,朱淼淼,等.基于BP神经网络的土壤侵蚀预测模型研究[J].沈阳农业大学学报,2013,44(4):495.GUO Chengjiu,JIE Li,ZHU Miaomiao,et al.Research on prediction model of soil erosion based on BPneural network[J].Journal of Shenyang Agricultural U-niversity,2013,44(4):495.
[13]赵明伟,汤国安,李发源,等.基于BP神经网络的陕北黄土高原侵蚀产沙影响因子显著性研究[J].水土保持通报,2012,32(1):5.ZHAO Mingwei,TANG Guoan,LI Fayuan,et al.Study on the significance of erosion of sediment yield in loess plateau in northern Shaanxi province based on BP neural network[J].Soil and Water Conservation Bulletin,2012,32(1):5.
[14]何文秀.彭阳县生态恢复措施景观格局尺度效应研究[D].银川:宁夏大学,2016:17.HE Wenxiu.Study on the scale effect of ecological restoration measures in Pengyang county[D].Yinchuan:Ningxia University,2016:17.
[15]李志伟.提高BP神经网络学习速率的算法研究[J].考试周刊,2016(21):103.LI Zhiwei.An algorithm for improving the learning rate of BP neural network[J].Test Journal,2016(21):103.
[16]谢云,刘宝元,章文波.侵蚀性降雨标准研究[J].水土保持学报,2000,14(4):6.XIE Yun,LIU Baoyuan,ZHANG Wenbo.Study on the standard of erosive rainfall[J].Journal of Soil and Water Conservation,2000,14(4):6.
[17]章文波,谢云,刘宝元,等.利用日雨量计算降雨侵蚀力的方法研究[J].地理科学,2002,22(6):706.ZHANG Wenbo,XIE Yun,LIU Baoyuan,et al.Study on the method of calculating rainfall erosion force by daily rainfall[J].Geographical Science,2002,22(6):706.
[18]江忠善,郑粉莉.坡面水蚀预报模型研究[J].水土保持学报,2004,11(1):66.JIANG Zhongshan,ZHENG Fenli.Study on the prediction model of slope water erosion[J].Journal of Soil and Water Conservation,2004,11(1):66.
[19]符素华,刘宝元,周贵云,等.坡长坡度因子计算工具[J].中国水土保持科学,2015,13(5):105.FU Suhua,LIU Baoyuan,ZHOU Guiyun,et al.Calculation tool for slope length and slope factor[J].Chinese Soil and Water Conservation Science,2015,13(5):105.
[20]申楠.黄土坡面细沟水流分离能力对水力学特征的响应过程研究[D].陕西杨凌:西北农林科技大学,2015:10.SHEN Nan.Study on the response process of the water flow separation capacity of the surface of the loess slope of the loess slope[D].Yangling,Shaanxi:Northwest Agricultural and Forestry University,2015:10.
[21]WILLIAMS J R,RENARD K G,DYKE P T.EPIC:a new method for assessing erosion's effect on soil productivity[J].Journal of Soil&Water Conservation,1983,38(5):382.
[22]张旭群,陈耀强,陈浩昆,等.基于GIS和RUSLE的粤东黄冈河流域土壤侵蚀评估[J].中国水土保持,2013(2):35.ZHANG Xuqun,CHEN Yaoqiang,CHEN Haokun,et al.Soil erosion assessment in the Huanggang river basin based on GIS and RUSLE[J].China Soil and Water Conservation,2013(2):35.
[23]WISCHMEIER W H,SMITH D D.Predicting rainfall erosion losses:A guide to conservation planning.[J].U-nited States Dept.of Agriculture Agriculture Handbook,1978:537.
[24]覃杰香,王兆礼.基于GIS和RUSLE的从化市土壤侵蚀量预测研究[J].人民珠江,2011,32(2):39.TAN Jiexiang,WANG Zhaoli.Prediction of soil erosion in the Conghua city based on GIS and RUSLE[J].People's Pearl River,2011,32(2):39.
[25]谢红霞.流域土壤侵蚀时空变化及水土保持环境效应评价研究[D].西安:陕西师范大学,2008:48.XIE Hongxia.Evaluation of soil erosion and environmental effects of soil and water conservation in Yanhe river basin[D].Xi'an:Shaanxi Normal University,2008:48.
[2]HESSEL R.Consequences of hyperconcenttrated flow for process-based soil erosion modeling on the Chinese Loess Plateau[J].Earth Surface Processes and Landforms,2006,31(9):1100.
[3]XU J X.Hyperconcentrated flows in the slope-channel systems in gullied hilly areas on the loess plateau,china[J].Geografiska Annaler:Series A,Physical Geography,2004,86(4):349.
[4]XU J X.Erosion caused by hyperconcentrated flow on the Loess Plateau[J].Catena,1999,36(1-2):1.
[5]郑粉莉,刘峰,杨勤科,等.土壤侵蚀预报模型研究进展[J].水土保持通报,2001,21(6):16.ZHENG Fenli,LIU Feng,YANG Qinke,et al.Research progress in soil erosion prediction model[J].Soil and Water Conservation Bulletin,2001,21(6):16.
[6]SMITH D D.Interpretation of soil conservation data for field use[J].Agricultural Engineering,1941,22:173.
[7]WISCHMEIER W H,SMITH D D.Predicting rainfall erosion losses from cropland east of the rocky mountains[M].USDA Agricultural Handbook,No.292,1965.
[8]WISCHMEIER W H,SMITH D D.Predicting rainfall erosion losses:A guide to conservation planning[M].US-DA Agricultural Handbook,No.537,1978:10(11),28.
[9]方开泰.实用多元统计分析(上)[M].上海:华东师范大学出版社,1989:145.FANG Kaitai.Practical multivariate statistical analysis(upper)[M].Shanghai:East China Normal University Press,1989:145.
[10]李扬,王伯昕,陈冬昕,等.基于BP神经网络预测复合盐侵蚀后混凝土的相对动弹性模量[J].混凝土,2018(7):21.LI Yang,WANG Bxin,CHEN Dongxin,et al.Prediction of relative dynamic elastic modulus of concrete after composite based on BP neural network theory[J].Concrete,2018(7):21.
[11]王尧,蔡运龙,潘懋,等.贵州省乌江流域土壤侵蚀模拟:基于GIS、RUSLE和ANN技术的研究[J].中国地质,2014,41(5):1735.WANG Yao,CAI Yunlong,PAN Mao,et al.Soil erosion simulation in Wujiang river basin,Guizhou province:Based on GIS,RUSLE and ANN technology[J].China geology,2014,41(5):1735.
[12]郭成久,颉丽,朱淼淼,等.基于BP神经网络的土壤侵蚀预测模型研究[J].沈阳农业大学学报,2013,44(4):495.GUO Chengjiu,JIE Li,ZHU Miaomiao,et al.Research on prediction model of soil erosion based on BPneural network[J].Journal of Shenyang Agricultural U-niversity,2013,44(4):495.
[13]赵明伟,汤国安,李发源,等.基于BP神经网络的陕北黄土高原侵蚀产沙影响因子显著性研究[J].水土保持通报,2012,32(1):5.ZHAO Mingwei,TANG Guoan,LI Fayuan,et al.Study on the significance of erosion of sediment yield in loess plateau in northern Shaanxi province based on BP neural network[J].Soil and Water Conservation Bulletin,2012,32(1):5.
[14]何文秀.彭阳县生态恢复措施景观格局尺度效应研究[D].银川:宁夏大学,2016:17.HE Wenxiu.Study on the scale effect of ecological restoration measures in Pengyang county[D].Yinchuan:Ningxia University,2016:17.
[15]李志伟.提高BP神经网络学习速率的算法研究[J].考试周刊,2016(21):103.LI Zhiwei.An algorithm for improving the learning rate of BP neural network[J].Test Journal,2016(21):103.
[16]谢云,刘宝元,章文波.侵蚀性降雨标准研究[J].水土保持学报,2000,14(4):6.XIE Yun,LIU Baoyuan,ZHANG Wenbo.Study on the standard of erosive rainfall[J].Journal of Soil and Water Conservation,2000,14(4):6.
[17]章文波,谢云,刘宝元,等.利用日雨量计算降雨侵蚀力的方法研究[J].地理科学,2002,22(6):706.ZHANG Wenbo,XIE Yun,LIU Baoyuan,et al.Study on the method of calculating rainfall erosion force by daily rainfall[J].Geographical Science,2002,22(6):706.
[18]江忠善,郑粉莉.坡面水蚀预报模型研究[J].水土保持学报,2004,11(1):66.JIANG Zhongshan,ZHENG Fenli.Study on the prediction model of slope water erosion[J].Journal of Soil and Water Conservation,2004,11(1):66.
[19]符素华,刘宝元,周贵云,等.坡长坡度因子计算工具[J].中国水土保持科学,2015,13(5):105.FU Suhua,LIU Baoyuan,ZHOU Guiyun,et al.Calculation tool for slope length and slope factor[J].Chinese Soil and Water Conservation Science,2015,13(5):105.
[20]申楠.黄土坡面细沟水流分离能力对水力学特征的响应过程研究[D].陕西杨凌:西北农林科技大学,2015:10.SHEN Nan.Study on the response process of the water flow separation capacity of the surface of the loess slope of the loess slope[D].Yangling,Shaanxi:Northwest Agricultural and Forestry University,2015:10.
[21]WILLIAMS J R,RENARD K G,DYKE P T.EPIC:a new method for assessing erosion's effect on soil productivity[J].Journal of Soil&Water Conservation,1983,38(5):382.
[22]张旭群,陈耀强,陈浩昆,等.基于GIS和RUSLE的粤东黄冈河流域土壤侵蚀评估[J].中国水土保持,2013(2):35.ZHANG Xuqun,CHEN Yaoqiang,CHEN Haokun,et al.Soil erosion assessment in the Huanggang river basin based on GIS and RUSLE[J].China Soil and Water Conservation,2013(2):35.
[23]WISCHMEIER W H,SMITH D D.Predicting rainfall erosion losses:A guide to conservation planning.[J].U-nited States Dept.of Agriculture Agriculture Handbook,1978:537.
[24]覃杰香,王兆礼.基于GIS和RUSLE的从化市土壤侵蚀量预测研究[J].人民珠江,2011,32(2):39.TAN Jiexiang,WANG Zhaoli.Prediction of soil erosion in the Conghua city based on GIS and RUSLE[J].People's Pearl River,2011,32(2):39.
[25]谢红霞.流域土壤侵蚀时空变化及水土保持环境效应评价研究[D].西安:陕西师范大学,2008:48.XIE Hongxia.Evaluation of soil erosion and environmental effects of soil and water conservation in Yanhe river basin[D].Xi'an:Shaanxi Normal University,2008:48.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |