基于机器学习算法的喀斯特峰丛洼地石漠化程度评估

doi:10.16843/j.sswc.2023.05.006

Abstract
Figure/Table
References
Related Citation (15)

Download: PDF (4778 KB) HTML (1 KB)
Export: BibTeX | EndNote (RIS)

Abstract [Background] Rocky desertification (RD) has become one of the most serious ecological and environmental problems in karst areas. The ecological environmental security problems such as soil erosion caused by rocky desertification have seriously affected people's living environment and sustainable development. Accurately evaluating rocky desertification is the key to implementing soil and water conservation projects, and ecological projects in karst areas. The study is aimed to compare the performance of seven different optimization algorithms machine-learning models so as to assess the degree of rocky desertification and then invert the spatial and temporal distribution of rocky desertification in a typical peak cluster depression basin in Southwest Guangxi by selecting the optimal model. [Methods] Based on Boruta, the Maximal Information Coefficient (MIC) and Extreme Learning Machine (ELM) feature selection method, Cuckoo Search (CS) algorithm, Whale Optimization Algorithm (WOA), Firefly Algorithm (FA), Artificial Bee Colony (ABC) optimization algorithm, Difference Evolution (DE) algorithm, Particle Swarm Optimization (PSO) algorithm, and Gravitational Search Algorithm (GSA) were used to adjust the super parameters of Support Vector Machine (SVM) model for assessing the extent of rocky desertification, using Moderate-resolution Imaging Spectroradiometer (MODIS) remote sensing data, topographic data, and meteorological data from 2001 to 2020 as well as the field survey data in 2020. [Results] 1) The bare rock rate and fractional vegetation coverage played an important role in assessing the degree of rocky desertification, followed by slope. 2) As shown by comparative analyses of three feature selection methods, the feature set constructed by Boruta had the best dimension reduction effect and the highest accuracy. 3) Seven intelligent optimization algorithms could effectively assist in the super parameter optimization of SVM. In addition, the accuracy of the optimization models, in descending order, were PSO-SVM, FA-SVM, GSA-SVM, CS-SVM, ABC-SVM, WOA-SVM, and DE-SVM. The corresponding overall accuracy values were 96.2%, 95.4%, 95.4%, 93.1%, 93.1%, 93.1% and 93.1%, and the Kappa coefficients were 0.95, 0.93, 0.93, 0.90, 0.90, 0.90 and 0.90, respectively. 4) In terms of spatial variation, in the early study period (2001-2010), severe rocky desertification and extremely severe rocky desertification accounted for large areas. However, from 2011 to 2020, the degree of rocky desertification was contained, and the extremely severe rocky desertification was scattered in small areas. [Conclusions] Generally speaking, the rocky desertification in the study area demonstrates a positive trend of improvement. The optimization algorithm and support vector machine have a good prospect for assessing rocky desertification in peak cluster depression basins. In short, this result may provide data support for the control of rocky desertification and the implementation of soil and water conservation plans in the karst area of the peak cluster depression in Southwest Guangxi.

Key words： rocky desertification support vector machine optimization algorithm feature selection machine learning peak cluster depression

Received: 14 March 2022

PACS:

P931.5

	Service

	E-mail this article
	Add to my bookshelf
	Add to citation manager
	E-mail Alert
	RSS
	Articles by authors
	ZHANG Yali
	TIAN Yichao
	WANG Donghua

Cite this article:

ZHANG Yali,TIAN Yichao,WANG Donghua. Evaluation of rocky desertification degree in karst peak cluster depression based on machine learning[J]. SSWC, 2023, 21(5): 51-61.

URL:

http://journal12.magtechjournal.com/Jweb_stbc/EN/10.16843/j.sswc.2023.05.006 OR http://journal12.magtechjournal.com/Jweb_stbc/EN/Y2023/V21/I5/51

[1]	汪泓, 王跃跃, 蔡宏, 等. 基于多源数据的多尺度石漠化时空变化研究[J]. 地下空间与工程学报, 2021, 17(S2):1044. WANG Hong, WANG Yueyue, CAI Hong, et al. Research on temporal and spatial changes of multi-scale rocky desertification based on multi-source data[J]. Chinese Journal of Underground Space and Engineering, 2021, 17(S2):1044.
[2]	王克林, 岳跃民, 陈洪松, 等. 科技扶贫与生态系统服务提升融合的机制与实现途径[J]. 中国科学院院刊, 2020, 35(10):1264. WANG Kelin, YUE Yuemin, CHEN Hongsong, et al. Mechanisms and realization pathways for integration of scientific poverty alleviation and ecosystem services enhancement[J]. Bulletin of Chinese Academy of Sciences, 2020, 35(10):1264.
[3]	种国双, 海月, 郑华, 等. 石漠化遥感信息提取方法进展[J]. 应用科学学报, 2021, 39(6):961. CHONG Guoshuang, HAI Yue, ZHENG Hua, et al. Research on temporal and spatial changes of multi-scale rocky desertification based on multi-source data[J]. Chinese Journal of Underground Space and Engineering, 2021, 39(6):961.
[4]	HUANG Xianfei, ZHANG Zhenming, ZHOU Yunchao, et al. Driving factors and prediction of rock desertification of non-tillage lands in a karst basin, southwest China[J]. Polish Journal of Environmental Studies, 2021, 30, 3627.
[5]	FAN Zemeng, LI Saibo, FANG Haiyan. Explicitly identifying the desertification change in CMREC area based on multisource remote data[J]. Remote Sensing, 2020, 12:3170.
[6]	李长春, 陈伟男, 王宇, 等. 基于多源Sentinel数据的县域冬小麦种植面积提取[J].农业机械学报, 2021, 52(12):207. LI Changchun, CHEN Weinan, WANG Yu, et al. Extraction of winter wheat planting area in county based on multi-sensor sentinel data[J]. Transactions of the CSAM, 2021, 52(12):207.
[7]	徐红枫, 王妍, 刘云根, 等. 基于Google Earth Engine的典型峰丛洼地石漠化时空演变与驱动因子分析:以西畴县为例[J]. 水土保持研究, 2022,29(4):407. XU Hongfeng, WANG Yan, LIU Yungen, et al. Analysis of the spatial and temporal evolution and driving factors of rocky desertification in typical cluster depression areas in the past 30years based on Google Earth Engine:Taking Xichou county as an example[J]. Research of Soil and Water Conservation, 2022,29(4):407.
[8]	PU Junwei, ZHAO Xiaoqing, DONG Pinliang, et al. Extracting information on rocky desertification from satellite images:A comparative study[J]. Remote Sensing, 2021, 13:2497.
[9]	QIAN Chunhua, QIANG Hequn, WANG Feng, et al. Optimization of rocky desertification classification model based on vegetation type and seasonal characteristic[J]. Remote Sensing, 2021, 13:2935.
[10]	张雪梅, 祁向坤, 岳跃民, 等. 喀斯特峰丛洼地石漠化治理自然地域分区[J]. 生态学报, 2020, 40(16):5490. ZHANG Xuemei, QI Xiangkun, YUE Yuemin, et al. Natural regionalization for rocky desertification treatment in karst peak-cluster depression regions[J]. Acta Ecologica Sinica, 2020, 40(16):5490.
[11]	王茜, 赵筱青, 普军伟, 等. 滇东南喀斯特区域石漠化时空格局演变研究:以广南县为例[J]. 中国岩溶, 2021,40(4):707. WANG Qian, ZHAO Xiaoqing, PU Junwei, et al. Temporal and spatial pattern evolution of rocky desertification in karst region of Southeast Yunnan:A case study of Guangnan county[J]. Carsologica Sinica, 2021,40(4):707.
[12]	李飞, 伊文超, 阎维巍. 从古地理学角度探讨广西石漠化分布特征[J]. 古地理学报, 2013, 15(1):135. LI Fei, YI Wenchao, YAN Weiwei. Distribution of Guangxi karst rocky desertification by palaeogeographic approach[J]. Journal of Palaeogeography, 2013, 15(1):135.
[13]	JIANG Zhongcheng, LIAN Yanqing, QIN Xiaoqun. Rocky desertification in Southwest China:Impacts, causes, and restoration[J]. Earth-Science Reviews, 2014, 132, 1.
[14]	朱林富. 基于MODIS数据的植被覆盖度时空变化及石漠化监测研究[D].重庆:西南大学, 2018:31. ZHU Linfu. Study on the spatial-temporal variation of vegetation coverage and karst rocky desertification based on MODIS data[D].Chongqing:Southwest University, 2018:31.
[15]	TIAN Yichao, HUANG Hu, ZHOU Guoqing, et al. Aboveground mangrove biomass estimation in Beibu Gulf using machine learning and UAV remote sensing[J]. Science of the Total Environment, 2021, 781:146816.
[16]	KESKIN H, GRUNWALD S, HARRIS W G. Digital mapping of soil carbon fractions with machine learning[J]. Geoderma, 2019, 339:40.
[17]	LIN Kairong, LU Pengyu, XU Chongyu, et al. Modeling saltwater intrusion using an integrated Bayesian model averaging method in the Pearl River Delta[J]. Journal of Hydroinformatics, 2019, 21(6):1.
[18]	LU Pengyu, LIN Kairong, XU Chongyu, et al. An integrated framework of input determination for ensemble forecasts of monthly estuarine saltwater intrusion[J]. Journal of Hydrology, 2021,598:126225.
[19]	胡军, 赵允坤, 栾长庆, 等. 基于布谷鸟算法优化SVM的尾矿库坝体变形预测[J]. 有色金属工程, 2021, 11(9):123. HU Jun, ZHAO Yunkun, LUAN Changqing, et al. Deformation prediction of tailing pond dam body based on cuckoo search optimized SVM[J]. Nonferrous Metals Engineering, 2021, 11(9):123.
[20]	曲蕴慧, 陈小菊. 基于人工蜂群优化的MR图像分割算法研究[J]. 影像科学与光化学, 2020, 38(3):508. QU Yunhui, CHEN Xiaoju. Research on mr image segmentation algorithm based on artificial bee colony optimization[J]. Imaging Science and Photochemistry, 2020, 38(3):508.
[21]	唐仁华, 罗文兴. 基于万有引力搜索算法优化支持向量机的软土路基参数反演[J]. 铁道科学与工程学报, 2022,19(6):1568. TANG Renhua, LUO Wenxing. Parameter inversion of soft soil subgrade based on support vector machine optimized by gravitational search algorithm[J]. Journal of Railway Science and Engineering, 2022,19(6):1568.
[22]	CAI Mingxiang, HOCINE O, MOHAMMED A S, et al. Integrating the LSSVM and RBFNN models with three optimization algorithms to predict the soil liquefaction potential[J]. Engineering with Computers, 2022,38(4):3611.
[23]	周建伟, 吴一全. MRELBP特征、Franklin矩和SVM相结合的遥感图像建筑物识别方法[J]. 测绘学报, 2020, 49(3):355. ZHOU Jianwei, WU Yiquan. Building area recognition method of remote sensing image based on MRELBP feature, Franklin moment and SVM[J]. Acta Geodaetica et Cartographica Sinica, 2020, 49(3):355.
[24]	晋健, 刘育, 王琴慧, 等. 基于小波去噪和FA-SVM的中长期径流预报[J]. 人民长江, 2020, 51(9):67. JIN Jian, LIU Yu, WANG Qinhui, et al. Mid-long term runoff forecasting based on wavelet de-noising and FA-SVM[J]. Yangtze River, 2020, 51(9):67.
[25]	谢劭峰, 赵云, 李国弘, 等. 基于WPA-WOA-BP神经网络的PM2.5浓度预测[J]. 大地测量与地球动力学, 2021, 41(1):12. XIE Shaofeng, ZHAO Yun, LI Guohong, et al. Prediction of PM2.5 concentration based on WPA-WOA-BP neural network[J]. Journal of Geodesy and Geodynamics, 2021, 41(1):12.
[26]	种国双, 海月, 郑华, 等. 中国西南喀斯特石漠化治理现状及对策[J]. 长江科学院院报, 2021, 38(11):38. CHONG Guoshuang, HAI Yue, ZHENG Hua, et al. Current situation and measures of karst rocky desertification control in Southwest China[J]. Journal of Yangtze River Scientific Research Institute, 2021, 38(11):38.
[27]	于利峰, 乌兰吐雅, 李继辉, 等. 多源遥感数据特征优选的大兴安岭沿麓不同农作物提取[J]. 北方农业学报, 2020, 48(6):119. YU Lifeng, ULAN Tuya, LI Jihui, et al. Multi-source remote sensing data feature optimization for different crop extraction in Daxing Anling along the foothills[J]. Journal of Northern Agriculture, 2020, 48(6):119.
[28]	习慧鹏, 王世杰, 白晓永, 等. 西南典型喀斯特地区石漠化时空演变特征:以贵州省普定县为例[J]. 生态学报, 2018, 38(24):8919. XI Huipeng, WANG Shijie, BAI Xiaoyong, et al. Spatio-temporal characteristics of rocky desertification in typical karst areas of Southwest China:A case study of Puding county, Guizhou province[J]. Acta Ecologica Sinica, 2018, 38(24):8919.
[29]	李素敏, 袁利伟, 甘淑, 等. 基于ETM+的滇东南石漠化区域植被指数与叶面积指数反演关系研究[J]. 昆明理工大学学报(自然科学版), 2015, 40(6):31. LI Sumin, YUAN Liwei, GAN Shu, et al. Study on inversion relationship between vegetation index and leaf area index of rocky desertification area in southeast Yunnan based on ETM+[J]. Journal of Kunming University of Science and Technology (Natural Sciences), 2015, 40(6):31.
[30]	陈彦君, 赵筱青, 普军伟, 等. 滇东南喀斯特山区石漠化程度与坡位因子的关联机制研究[J/OL]. 中国岩溶, (2022-04-16)[2022-05-16]. http://kns.cnki.net/kcms/detail/45.1157.P.20220414.1805.002.html. CHEN Yanjun, ZHAO Xiaoqing, PU Junwei, et al. Study on correlation mechanism between rocky desertification degree and slope position factors in karst mountainous area of Southeast Yunnan[J/OL]. Carsologica Sinica, (2022-04-16)[2022-05-16]. http://kns.cnki.net/kcms/detail/45.1157.P.20220414.1805.002.html.