地球科学进展 ›› 2023, Vol. 38 ›› Issue (12): 1259 -1270. doi: 10.11867/j.issn.1001-8166.2023.081

研究论文 上一篇    下一篇

基于改进随机森林模型的多要素层次插值技术在土壤硒元素空间分布上的运用
贾黎黎( ), 李婷婷, 朱鑫, 易隆科, 罗思亮   
  1. 广东省地质调查院,广东 广州 510080
  • 收稿日期:2023-06-18 修回日期:2023-10-18 出版日期:2023-12-10
  • 基金资助:
    国家自然科学基金项目(U1911202);广东省地质勘查与城市地质专项项目(2023-25);湛江市财政项目(CLZ0121ZJ01ZC00)

Application of Multifactor Hierarchical Interpolation Technique Based on Improved Random Forest Model in Spatial Distribution of Selenium in Soils

Lili JIA( ), Tingting LI, Xin ZHU, Longke YI, Siliang LUO   

  1. Geological Investigation Institute of Guangdong Province, Guangzhou 510080, China
  • Received:2023-06-18 Revised:2023-10-18 Online:2023-12-10 Published:2023-12-26
  • About author:JIA Lili, Senior engineer, research areas include geochemical exploration, agricultural geological survey and research. E-mail: lichia@126.com
  • Supported by:
    the National Nature Science Foundation of China(U1911202);Guangdong Geological Exploration and Urban Geological Project(2023-25);Zhanjiang Municipal Finance Project(CLZ0121ZJ01ZC00)

地球化学元素的空间分布变异特征复杂,传统区域地球化学调查分析只考虑空间信息要素,无法反映其他要素层次异质性。因此提出采用地质要素、地理要素、空间要素、生态要素和人类活动等层次区域化标签作为机器学习模型的协变量进行硒元素空间分布预测。通过对比随机森林模型、极致梯度提升树模型、特征贡献—随机森林模型、深度神经网络模型以及传统的空间插值模型,发现特征贡献—随机森林模型在硒元素预测方面具有更高的精度,表明基于特征贡献—随机森林模型的多要素层次插值技术在土壤硒元素空间分布预测中具有较高的可行性。根据预测结果对硇洲岛硒元素空间分布进行贝叶斯局部空间自相关分析表明,影响硒元素空间分布的关键因子是基础地质条件和人类活动等。研究可进一步拓展地球化学数据的服务应用维度,使土壤地球化学数据的运用更加精确和科学。

The spatial distribution of geochemical elements is complex, and traditional regional geochemical surveys and analyses consider only spatial information elements, which do not reflect the heterogeneity of other elements. Therefore, this study used hierarchical regionalization labels, such as geological, geographical, spatial, and ecological elements along with anthropogenic activities, as covariates of the machine learning model to predict the spatial distribution of selenium elements. By comparing Random Forest (RF), extreme gradient lifting tree (XGboost), Feature Contribution-Random Forest (FC-RF), Deep learning Neural Network (DNN) models, and traditional spatial interpolation techniques; this study showed that the FC-RF model has higher precision in selenium prediction, which indicates that multilevel interpolation technology based on the FC-RF model is more feasible for predicting the spatial distribution of soil selenium. Based on the predicted results, a Bayesian Laser Interferometer Space Antenna (LISA) spatial correlation analysis was conducted on the spatial distribution of selenium on Naozhou Island, indicating that the key factors affecting the spatial distribution of selenium are basic geological conditions and anthropogenic activities. This study further expands the service application dimensions of geochemical data, thus making the use of soil geochemical data more accurate and scientific.

中图分类号: 

图1 硇洲岛地理位置及采样点布局图
Fig. 1 Geographical location and sampling point layout of Naozhou Island
表1 Se测试数据统计特征
Table 1 Statistical characteristics of Se test data
图2 空间协变量特征重要性分析图
Fig. 2 Analysis of the importance of spatial covariate features
图3 土壤硒元素含量预测研究框架图
Fig. 3 Research framework for prediction of selenium content in soil
图4 模型训练精度评估图
Fig. 4 Evaluation diagrams of model training accuracy
表2 采样点统计区间离散系数与拟合度统计表
Table 2 Statistical table for discrete coefficients and fit of sampling point statistical intervals
图5 基于特征贡献—随机森林(FC-RF)模型、极致梯度提升树(XGBoost)模型、深度神经网络(DNN)模型、克里金(Kriging)插值模型的Se元素空间分布预测
Fig. 5 Prediction of spatial distribution of selenium contents based on Feature Contribution-Random ForestFC-RF), eXtreme Gradient BoostingXGBoost), Deep Neural NetworkDNNand Kriging models
图6 LISA显著性检验判别图
Fig. 6 The discriminant chart of LISA significance test
图7 LISA空间聚类效应分布图
Fig. 7 The distribution of LISA spatial clustering effect
1 HU Q, CHEN L C, XU J, et al. Determination of selenium concentration in rice and the effect of foliar application of Se-enriched fertiliser or sodium selenite on the selenium content of rice[J]. Journal of the Science of Food and Agriculture, 2002, 82: 869-872.
2 QIAO Xinxing, CHAO Xu, REN Rui, et al. Research, development and utilization of selenium-rich soil of Shaanxi: a case study of Sanyuan-Yanliang area[J]. Geophysical and Geochemical Exploration, 2021, 45(1): 230-238.
乔新星, 晁旭, 任蕊, 等. 陕西关中富硒土壤研究及开发利用: 以三原—阎良地区为例[J]. 物探与化探, 2021, 45(1): 230-238.
3 WANG Zhiqiang, YANG Jianfeng, SHI Tianchi. A preliminary study of Se-rich soil in the Shizuishan area, Ningxia and its potential for application[J]. Geophysical and Geochemical Exploration, 2023, 47(1): 228-237.
王志强, 杨建锋, 石天池. 宁夏石嘴山地区富硒土壤及其利用前景[J]. 物探与化探, 2023, 47(1): 228-237.
4 YUAN Hongwei, CHEN Jiangjun, GUO Tengda, et al. Geochemical characteristics and influencing factors of Se-rich soils in Langshan and Xinhua towns, Linhe district, Bayannur City[J]. Geology and Exploration, 2022, 58(5): 1 027-1 041.
袁宏伟, 陈江均, 郭腾达, 等. 巴彦淖尔市临河区狼山镇和新华镇一带富硒土壤地球化学特征及影响因素[J]. 地质与勘探, 2022, 58(5): 1 027-1 041.
5 LIU Jian, WANG Yifan, LIN Zhongyang, et al. Distribution, sources and ecological effects of selenium in topsoil of cultivated land in Jiande City, Zhejiang Province[J]. Geoscience, 2022, 36(3): 953-962.
刘健, 汪一凡, 林钟扬, 等. 浙江建德市耕地表层土壤硒分布、来源及生态效应[J]. 现代地质, 2022, 36(3): 953-962.
6 LI Tingting, JIA Lili, ZHU Xin, et al. Geochemical evaluation of the land quality and classification of the selenium-enriched soils in Chengyue area, Leizhou peninsula[J]. Earth and Environment, 2022, 50(4): 481-489.
李婷婷, 贾黎黎, 朱鑫, 等. 雷州半岛富硒区土地质量地球化学评价及其利用区划研究: 以城月地区为例[J]. 地球与环境, 2022, 50(4): 481-489.
7 STEIN A, HOOGERWERF M, BOUMA J. Use of soil-map delineations to improve (Co-) Kriging of point data on moisture deficits[J]. Geoderma, 1988, 43(2/3): 163-177.
8 HAN Z H, GÖRTZ S. Hierarchical kriging model for variable-fidelity surrogate modeling[J]. AIAA Journal, 2012, 50(9): 1 885-1 896.
9 LIN Y P, CHENG B Y, CHU H J, et al. Assessing how heavy metal pollution and human activity are related by using logistic regression and kriging methods[J]. Geoderma, 2011, 163(3/4): 275-282.
10 YOUNG M T, BECHLE M J, SAMPSON P D, et al. Satellite-based NO2 and model validation in a national prediction model based on universal kriging and land-use regression[J]. Environmental Science & Technology, 2016, 50(7): 3 686-3 694.
11 TAN Z, YANG Q, ZHENG Y. Machine learning models of groundwater arsenic spatial distribution in Bangladesh: influence of Holocene sediment depositional history[J]. Environmental Science & Technology, 2020, 54(15): 9 454-9 463.
12 HENGL T, NUSSBAUM M, WRIGHT M N, et al. Random forest as a generic framework for predictive modeling of spatial and spatio-temporal variables[J]. PeerJ, 2018. DOI:10.7717/peerj.5518 .
13 DEMYANOV V, KANEVSKY M, CHERNOV S, et al. Neural network residual Kriging application for climatic data[J].Journal of Geographic Information and Decision Analysis,1998,2:215-232.
14 ZHAO W H, MA J, LIU Q Y, et al. Accurate prediction of soil heavy metal pollution using an improved machine learning method: a case study in the Pearl River Delta, China[J]. Environmental Science & Technology, 2023, 57(46): 17 751-17 761.
15 HU Y A, CHENG H F. Application of stochastic models in identification and apportionment of heavy metal pollution sources in the surface soils of a large-scale region[J]. Environmental Science & Technology, 2013, 47(8): 3 752-3 760.
16 ZHANG H, YIN A J, YANG X H, et al. Use of machine-learning and receptor models for prediction and source apportionment of heavy metals in coastal reclaimed soils[J]. Ecological Indicators, 2021, 122. DOI:10.1016/j.ecolind.2020.107233 .
17 BREIMAN L. Random forests[J]. Machine Learning, 2001, 45: 5-32.
18 ZHANG H, WU P B, YIN A J, et al. Prediction of soil organic carbon in an intensively managed reclamation zone of Eastern China: a comparison of multiple linear regressions and the random forest model[J]. Science of the Total Environment, 2017, 592: 704-713.
19 CHEN T Q, GUESTRIN C. XGboost: a scalable tree boosting system[C]// Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining, 2016: 785-794.
20 WONG P Y, LEE H Y, CHEN Y C, et al. Using a land use regression model with machine learning to estimate ground level PM2.5 [J]. Environmental Pollution, 2021, 277. DOI:10.1016/j.envpol.2021.116846 .
21 XIA Weiping, TAN Jianan. A comparative study of selenium content in Chinese rocks[J]. Acta Scientiae Circumstantiae, 1990, 10(2): 125-131.
夏卫平, 谭见安. 中国一些岩类中硒的比较研究[J].环境科学学报, 1990, 10(2): 125-131.
22 QIN Jianxun, FU Wei, ZHENG Guodong, et al. Selenium distribution in surface soil layer of Karst area of Guangxi and its affecting factors: a case study of Wuming County[J]. Acta Pedologica Sinica, 2020, 57(5): 1 299-1 310.
覃建勋, 付伟, 郑国东, 等. 广西岩溶区表层土壤硒元素分布特征与影响因素探究: 以武鸣县为例[J]. 土壤学报, 2020, 57(5): 1 299-1 310.
23 WEI Zhenshan, TU Qijun, TANG Shuhong, et al. A discussion on the geochemical features and origin of selenium-rich soil on the northern slope of the Tianshan Mountains from Urumqi to Shawan County[J]. Geophysical and Geochemical Exploration, 2016, 40(5): 893-898.
魏振山, 涂其军, 唐蜀虹, 等. 天山北坡乌鲁木齐至沙湾地区富硒土壤地球化学特征及成因探讨[J]. 物探与化探, 2016, 40(5): 893-898.
24 LI Yigen, DONG Yanxiang, ZHENG Jie, et al. Selenium: abundant soil survey and assessment in Zhejiang[J]. Quaternary Sciences, 2005, 25(3): 323-330.
郦逸根, 董岩翔, 郑洁, 等. 浙江富硒土壤资源调查与评价[J]. 第四纪研究, 2005, 25(3): 323-330.
25 ZHOU Mo, CHEN Guoguang, ZHANG Ming, et al. Geochemical characteristics and influencing factors of selenium in soils of South Jiangxi Province: a typical area of Qingtang-Meijiao[J]. Geoscience, 2018, 32(6): 1 292-1 301.
周墨, 陈国光, 张明, 等. 赣南地区土壤硒元素地球化学特征及其影响因素研究: 以青塘—梅窖地区为例[J]. 现代地质, 2018, 32(6): 1 292-1 301.
26 LIU Daorong, XU Hong, ZHOU Yi, et al. Characteristics and genetic analysis of selenium-rich soil in Changshan County, western Zhejiang Province[J]. Geophysical and Geochemical Exploration, 2019, 43(3): 658-666.
刘道荣, 徐虹, 周漪, 等. 浙西常山地区富硒土壤特征及成因分析[J]. 物探与化探, 2019, 43(3): 658-666.
27 HONG Wanhua, SU Te, TU Feifei, et al. Study on the distribution and influencing factors of selenium based on the geochemical method of land quality: a case study of Tongren area[J]. Metal Mine, 2021(12): 160-168.
洪万华, 苏特, 涂飞飞, 等. 基于土地质量地球化学方法的硒元素分布规律和影响因素研究: 以铜仁地区为例[J]. 金属矿山, 2021(12): 160-168.
28 CAO Ronghao. Study on selenium content of surface soils in Longhai, Fujian and its influencing factors[J]. Rock and Mineral Analysis, 2017, 36(3): 282-288.
曹容浩. 福建省龙海市表层土壤硒含量及影响因素研究[J]. 岩矿测试, 2017, 36(3): 282-288.
29 CHEN Chunliang, BAO Kaiqiang, WANG Mengying, et al. Effects of vegetation removal on soil organic matter and nutrients in an erosive environment[J]. Research of Soil and Water Conservation, 2022, 29(5): 131-136.
陈春良, 鲍凯强, 王梦莹, 等. 植被去除对侵蚀环境土壤有机质和养分的影响[J]. 水土保持研究, 2022, 29(5): 131-136
30 HOBBIE S E. Effects of plant species on nutrient cycling[J]. Trends in Ecology & Evolution, 1992, 7(10): 336-339.
31 GUSTAFSSON J P, JACKS G, STEGMANN B, et al. Soil acidity and adsorbed anions in Swedish forest soils—long-term changes[J]. Agriculture, Ecosystems & Environment, 1993, 47(2): 103-115.
32 GUSTAFSSON J P, JOHNSSON L. Selenium retention in the organic matter of Swedish forest soils[J]. Journal of Soil Science, 1992, 43(3): 461-472.
33 SHANG Jingmin, LUO Wei, WU Guanghong, et al. Spatial distribution of Se in soils from different land use types and its influencing factors within the Yanghe watershed, China[J]. Environmental Science, 2015, 36(1): 301-308.
商靖敏, 罗维, 吴光红, 等. 洋河流域不同土地利用类型土壤硒(Se)分布及影响因素[J]. 环境科学, 2015, 36(1): 301-308.
34 YU T, YANG Z F, LV Y Y, et al. The origin and geochemical cycle of soil selenium in a Se-rich area of China[J]. Journal of Geochemical Exploration, 2014, 139: 97-108.
35 HOU Shaofan, LI Dezhu, WANG Lizhen, et al. The differentiation characteristics of soil sel-enium in warm-temperate geographic lands-cape[J]. Acta Geographica Sinica, 1992, 47(1): 31-39.
侯少范, 李德珠, 王丽珍, 等. 暖温带地理景观中土壤硒的分异特征[J]. 地理学报, 1992, 47(1): 31-39.
36 ZHANG Miaomei. Study on seawater intrusion in Naozhou Island, Guangdong Province[J]. Underground Water, 2018, 40(5):138-139, 159.
张妙美. 广东省硇洲岛海水入侵研究[J]. 地下水, 2018, 40(5):138-139, 159.
[1] 钟庆祥, 张豫, 陶贞, 贺一聪, 吴迪, 林培松. 土壤—植物系统硒的迁移转化机制研究进展[J]. 地球科学进展, 2023, 38(1): 44-56.
[2] 雒恺, 马金龙. 花岗岩风化过程中稀土元素迁移富集机制研究进展[J]. 地球科学进展, 2022, 37(7): 692-708.
[3] 段美铃, 宋昊, 胡伟, 廖昕. 川北韩家店组页岩风化过程的矿物学与元素地球化学研究[J]. 地球科学进展, 2022, 37(6): 641-659.
[4] 闫雅妮, 张伟, 张俊文, 任亚雄, 赵志琦. 大陆硅酸盐岩石风化过程中镁同位素地球化学研究进展[J]. 地球科学进展, 2021, 36(3): 325-334.
[5] 储著银, 许继峰. 铼—锇同位素和铂族元素分析方法及地学应用进展[J]. 地球科学进展, 2021, 36(3): 245-264.
[6] 张为,周丽,唐红峰,李和平,王力. 水热体系中 Na2SO4/K2SO4 溶解度的热力学计算[J]. 地球科学进展, 2019, 34(4): 414-423.
[7] 林祖苇,赵新福,熊乐,朱照先. 胶东三山岛金矿床黄铁矿原位微区微量元素特征及对矿床成因的指示[J]. 地球科学进展, 2019, 34(4): 399-413.
[8] 康健,陈列锰,宋谢炎,戴智慧,郑文勤. 金川超大型 Ni-Cu-( PGE)矿床橄榄石微量元素特征及地质意义[J]. 地球科学进展, 2019, 34(4): 382-398.
[9] 张瑞刚, 高雪, 杨立强. 岩浆混合作用的识别:以义敦岛弧稻城岩体为例[J]. 地球科学进展, 2018, 33(10): 1058-1074.
[10] 唐志敏, 侯青叶, 游远航, 杨忠芳, 李括. 珠三角平原区第四系剖面重金属分布特征及其影响因素[J]. 地球科学进展, 2017, 32(8): 885-898.
[11] 王的, 冯海艳, 景慧敏. 北京市冬季、春季PM 10和PM 2.5中元素地球化学特征[J]. 地球科学进展, 2017, 32(8): 850-858.
[12] 穆延宗, 乜贞, 卜令忠, 王云生, 伍倩. 我国油(气)田水钾资源研究进展[J]. 地球科学进展, 2016, 31(2): 147-159.
[13] 董爱国, 朱祥坤. 表生环境中镁同位素的地球化学循环[J]. 地球科学进展, 2016, 31(1): 43-58.
[14] 黄从俊, 李泽琴. 拉拉IOCG矿床萤石的微量元素地球化学特征及其指示意义[J]. 地球科学进展, 2015, 30(9): 1063-1073.
[15] 侯荣娜, 王淑华, 张翔, 侯克选, 张铖, 王金荣. 中祁连西段花岗岩类的地球化学特征及构造意义[J]. 地球科学进展, 2015, 30(9): 1034-1049.
阅读次数
全文


摘要