کارایی سه مدل kNN، RF و SVM و مدل به دست آمده از ترکیب آنها به روش GR برای مدل‌سازی بافت خاک

نوع مقاله : مقاله پژوهشی

نویسندگان

1 دانشجوی دکتری گروه مهندسی علوم خاک، دانشکده کشاورزی، دانشگاه لرستان، خرم آباد، ایران

2 استادیار گروه مهندسی علوم خاک، دانشکده کشاورزی، دانشگاه لرستان، خرم آباد، ایران

3 استادیار گروه مرتع و آبخیزداری، دانشکده کشاورزی و منابع طبیعی، دانشگاه اردکان، اردکان، ایران

4 استاد گروه مهندسی علوم خاک، دانشکده کشاورزی، دانشگاه لرستان، خرم آباد، ایران

10.22092/ijsr.2024.364333.735

چکیده

بافت خاک یکی از مهمترین ویژگی­هایی است که رفتار فیزیکی، شیمیایی و بیولوژکی خاک را کنترل می­کند. روش­های مختلفی برای مدل­سازی بافت خاک استفاده می­شوند. یکی از راهکارهای سود بردن از مزایای این مدل­ها ترکیب تخمین آنها است. با توجه به این که بافت خاک یک داده مرکب است، وقتی اجزاء آن جداگانه تخمین زده می­شوند تضمینی برای اینکه جمع سه جزء برابر 100 شود وجود ندارد، هرچند می­توان از تبدیل­های نسبت لگاریتمی (log-ratio) استفاده کرد. اطلاعات کمی در خصوص کارآیی مدل­های ترکیبی در مدل­سازی داده­های تبدیل­شده و نشده بافت خاک وجود دارد و به نظر می­رسد بر اساس این رویکرد تا کنون مطالعه­ای روی بافت خاک انجام نشده است. در این بررسی، تعداد 200 نمونه خاک­های سطحی از  منطقه کوهدشت برداشت شد. سه مدل جنگل تصادفی (RF)، k نزدیکترین همسایه (kNN) و ماشین­های بردار پشتیبان (SVM) و مدل حاصل از ترکیب آن­ها به روش Granger-Ramanathan (GR) برای مدل­سازی، روش­های نسبت لگاریتمی جمع­پذیر (alr)، نسبت لگاریتمی مرکزی  (clr) و نسبت لگاریتمی ایزومتریک (ilr) برای تبدیل داده­ها و داده­های حاصل از مدل رقومی ارتفاع (DEM) و تصاویر لندست 8 و سنتینل 2 به عنوان ورودی مدل­ها استفاده شد. نتایج نشان داد که متغیرهای استخراج­شده از DEM اهمیت بیشتری در پیش­بینی بافت خاک داشت. به­طور کلی، هر چهار مدل با استفاده از تبدیل­ alr منجر به تخمین­های بهتری نسبت به تبدیل­های clr و ilr و داده­های تبدیل­نشده (UT) گردید. مدل ترکیبی(GR) با مقادیر RMSE برابر با 5/07، 4/21، 5/81 و 6/09 درصد برای رس، مقادیر 7/11، 5/15، 9/04 و 6/70 درصد برای سیلت و 9/20، 7/76، 11/69 و 8/74 درصد برای شن به ترتیب برای داده­های UT و تبدیل­های alr، clr و ilr منجر به بهبود تخمین­ها نگردید. به­طور کلی، کارآیی مدل SVM با داده­های تبدیل شده به روش نسبت لگاریتمی جمع­پذیر کمی­ بیشتر از سایر مدل­ها بود. نتایج نشان داد که ترکیب چند مدل یادگیری ماشین الزاما باعث بهبود تخمین­ها نمی­گردد و می­توان از یک مدل مناسب برای برآورد بافت خاک استفاده کرد.

کلیدواژه‌ها

موضوعات

عنوان مقاله [English]

Using kNN, RF and SVM and their Combination Using GR for Soil Texture Modeling

نویسندگان [English]

  • Fereshteh Mirzaei 1
  • Alireza Amirian-Chakan 2
  • Ruhollah Taghizadeh-Mehrjardi 3
  • Hamid Reza Matinfar 4
1 Ph.D Student, Department of Soil Science, Faculty of Agriculture, Lorestan University, Khorramabad, Iran
2 Assitant Professor, Department of Soil Science, Faculty of Agriculture, Lorestan University, Khorramabad, Iran
3 Assitant Professor, Department of Rangeland and Watersghed Management, Faculty of Agriculture and Natural Resources, University of Ardakan, Ardakan, Iran.
4 Professor, Department of Soil Science, Faculty of Agriculture, Lorestan University, Khorramabad, Iran.
چکیده [English]

Soil texture is one of the most important soil properties that govern soil physical, chemical and biological behaviors. In modeling soil textural fractions, different models are used. To combine the benefits from different models, one approach is combining their predictions. Since soil texture is a compositional data, when its fractions are estimated separately there is no guarantee that the estimates will sum to 100. Log-ratio transformations before modeling are ways to deal with the problem. Little is known about modeling transformed and untransformed (UT) soil texture data using a combination of different models. In the present study, 200 surface soil samples (0-30 cm) were collected from Kuhdasht region. Random forest (RF), k-nearest neighbors (kNN) and support vector machines (SVM) and their combination using Granger-Ramanathan (GR) method were used to model soil texture data. Additive log-ratio (alr), centroid log-ratio (clr) and isometric log-ratio (ilr) transformations were used to transform texture data. Environmental variables derived from Landsat 8 and Sentinel-2 images and a digital elevation model (DEM) were used as input for all models. Results indicated that covariates derived from DEM were more important in modeling soil texture. All models improved the estimates of soil texture fractions when alr transformed data was compared to UT, clr, and ilr transformed data. The combined model (i.e. GR) did not show superiority over other models. Using GR model RMSE values for alr, clr, ilr transformed clay data and UT were 5.07%, 4.21%, 5.81%, and 6.09%, respectively. For silt RMSE values (in the same order as clay) were 7.11%, 5.15%, 9.04%, and 6.70%, and for sand were 9.20%, 7.67%, 11.69% and 8.74%, respectively. Generally, SVM using alr transformed data showed a slightly higher potential for modeling soil texture. Generally, results indicated that combining different machine learning algorithms did not necessarily improve the estimates. Therefore, it is possible to use a single appropriate model for modeling soil texture.  

کلیدواژه‌ها [English]

  • Compositional data
  • Ensemble model
  • Log-ratio transformation
  • Random forest

مراجع

  1. Abedi, F., Amirian-Chakan, A., Faraji, M., Taghizadeh-Mehrjardi, R., Kerry, R., and Razmjoue, D., and Scholten, T. 2021. Salt dome related soil salinity in southern Iran: Prediction and mapping with averaging machine learning models. Land Degradation and Development, 32: 1540-1555.
  2. Adhikari, K., Minasny, B., Greve, M.B., and Greve, M.H. 2014. Constructing a soil class map of Denmark based on the FAO legend using digital techniques. Geoderma, 214: 101-113.
  3. Aitchison, J. 1986. The statistical analysis of compositional data. London: Chapman and Hall.
  4. Aitchison, J. 1992. On criteria for measures of compositional difference. Mathematical Geology, 24: 365–379.
  5. Akpa, S.I.C., Odeh, I.O.A., and Bishop, T.F.A. 2014. Digital mapping of soil particle-size fractions for Nigeria. Soil Science Society of America Journal, 78: 1953-1966.
  6. Banaei, M. 2000. Soil resources and use potentiality map of Iran, 1: 1000000. Karaj: Soil and Water Research Institute.
  7. Behrens, T., Schmidt, K., Viscarra Rossel, R.A., Gries, P., Scholten, T.,  and MacMillan, R.A. 2018. Spatial modelling with Euclidean distance fields and machine learning. European Journal of Soil Science, 69: 757-770.
  8. Breiman, L. 2001. Random forests. Machine Learning, 45: 5-32.
  9. Chagas, C.S., Junior, W.C., Bhering, S.B., and Filho, B.C. 2016. Spatial prediction of soil surface texture in a semiarid region using random forest and multiple linear regressions. Catena, 139: 232-240.
  10. Chen, S., Arrouays, D., Mulder, V.L., Poggio, L., Minasny, B., Roudier, P., Libohova, Z., Lagacherie, P., Shi, Z., Hannam, J., Meersmans, J., Richer-de-Forges, A.C., and Walter, C. 2022. Digital mapping of GlobalSoilMap soil properties at a broad scale: A review. Geoderma, 409, 115567.
  11. Chen, S., Mulder, V.L., Heuvelink, G.B.M., Poggio, L., Cabuet, M., Román Dobarco, M., and Arrouays, D. 2020. Model averaging for mapping topsoil organic carbon in France. Geoderma, 366, 114237.
  12. Conrad, O., Bechtel, B., Bock, M., Dietrich, H., Fischer, E., Gerlitz, L., Wehberg, J., Whichmann, V., and Bohner, J. 2015. System for automated geoscientific analyses (SAGA) v. 2.1.4. Geoscientific Model Development, 8: 1991-2007.
  13. Diks, C.G.H., and Vrugt, J.A. 2010. Comparison of point forecast accuracy of model averaging methods in hydrologic applications. Stochastic Environmental Research and Risk Assessment, 24: 809-820.
  14. Egozcue, J.J., Pawlowsky-Glahn, V., Mateu-Figueras, G., and Barcelo-Vidal, C. 2003. Isometric logratio transformations for compositional data analysis. Mathematical Geology, 35: 279-300.
  15. Filzmoser, P., Hron, K., and Reimann, C. 2009. Principal component analysis for compositioal data with outliers. Evironmetrics, 20: 621-632.
  16. Gallan, J.C., and Austin, J.M. 2015. Derivations of terrain covariates for digital soil mapping in Australia. Soil Research, 53: 895-906.
  17. Gallant, J.C., and Dowling, T.I. 2003. A multiresolution index of valley bottom flatness for mapping depositional areas. Water Resources Research, 39: 1347-1359.
  18. Gee, G.W., and Bauder, J.W. 1986. Particle size analysis. In A. Klute (ed). Methods of soil analysis: Part 1. American Society of Agronomy, Madison.
  19. Granger, C.W., and Ramanathan, R. 1984. Improved methods of combining forecasts. Journal of Forecasting, 32: 197-204.
  20. Greve, M.H., Kheir, R.B., Greve, M.B., and Bøcher, P.K. 2012. Quantifying the environmental parameters to predict soil texture fractions using regression-tree model with GIS and LIDAR data: The case study of Denmark. Ecological Indicators, 18: 1-10.
  21. Hengl, T., Rossiter D.G., and Stein, A. 2003. Soil sampling strategies for spatial prediction by correlation with auxiliary maps. Australian Journal of Soil Research, 418: 1403-1422.
  22. Kaya, F., Başayiğit, L., Keshavarzi, A., and Francaviglia, R. 2022. Digital mapping for soil texture class prediction in northwestern Türkiye by different machine learning algorithms. Geoderma Regional, 31, e00584.
  23. Lantz, B. 2015. Machine learning with R. Packt Publishing Ltd., Birmingham.
  24. Lark, R.M., and Bishop, T.F.A. 2007. Cokriging particle size fractions of the soil. European Journal of Soil Science, 583: 763-774.
  25. Liu, F., Geng, X., Zhu, A.X., Fraser, W., and Waddell, A. 2012. Soil texture mapping over low relief areas using land surface feedback dynamic patterns extracted from MODIS. Geoderma, 171- 172: 44-52.
  26. Mallah, S., Delsouz Khaki, B., Davatgar, N., Scholten, T., Amirian-Chakan, A., Emadi, M., Kerry, R., Mosavi, A.H., and Taghizadeh-Mehrjardi, R. 2022. Predicting soil textural classes using random forest models: learning from imbalanced dataset. Agronomy, 2613.
  27. Malone, B.P., Minasny, B., Odgers, N.P., and McBratney, A.B. 2014. Using model averaging to combine soil property rasters from legacy soil maps and from point data. Geoderma, 232: 34-44.
  28. Mehrabi-Gohari, E., Matinfar, H.R., Jafari, A., Taghizadeh-Mehrjardi, R., and Triantafilis, J. 2019. The spatial prediction of soil texture fractions in arid regions of Iran. Soil Systems, 3: 65.
  29. Metternicht, G.I., and Zinck, J.A. 2003. Remote sensing of soil salinity: potentials and constraints. Remote Sensing of Environment, 85: 1–20.
  30. Minasny, B., and McBratney, A.B. 2018. Limited effect of organic matter on soil available water capacity. European Journal of Soil Science, 69: 39-47.
  31. Odeh, I.O.A., Todd, A.J., and Triantafilis, J. Spatial prediction of soil particle-size fractions as compositional data. Soil Science, 168: 501-515.
  32. Pahlavan-Rad, M.R., and Akbarimoghaddam, A. 2018. Spatial variability of soil texture fractions and pH in a flood plain (case study from eastern Iran). Catena, 160: 275-281.
  33. Poggio, L., and Gimona, A. 2017. 3D mapping of soil texture in Scotland. Geoderma Regional, 9: 5-16.
  34. Román Dobarco, M., Arrouays, D., Lagacherie, P., Ciampalini, R., and Saby, N.P.A. 2017. Prediction of topsoil texture for Region Centre (France) applying model ensemble methods. Geoderma, 298: 67-77.
  35. Sun, Y., Wong, A.K.C., and Kamel, M.S. 2009. Classification of imbalanced data: A review. International Journal of Pattern Recognition and Artificial Intelligence, 23: 687-719.
  36. Swain, S.R., Chakraborty, P., Panigrahi, N., Vasava, H.B., Reddy, N.N., Roy, S., Majeed, I., and Das, B. S. 2021. Estimation of soil texture using Sentinel-2 multispectral imaging data: An ensemble modeling approach. Soil and Tillage Research, 213: 105134.
  37. Taghizadeh-Mehrjardi, R., Minasny, B., Toomanian N., Zeraatpisheh, M., Amirian-Chakan, A., and Triantafilis, J. 2019. Digital mapping of soil classes using ensemble of models in Isfahan region, Iran. Soil Systems, 3: 37.
  38. Taghizadeh-Mehrjardi, R., Nabiollahi, K., Minasny, B., and Triantafilis, J. 2015. Comparing data mining classifiers to predict spatial distribution of USDA-family soil groups in Baneh region, Geoderma, 253–254: 67–77
  39. Taghizadeh-Mehrjardi, R., Toomanian, N., Khavaninzadeh, A.R., Jafari, A., and Triantafilis, J. 2016. Predicting and mapping of soil particle-size fractions with adaptive neuro-fuzzy inference and ant colony optimization in central Iran. European Journal of Soil Science, 67: 707–725.
  40. Umali, B.P., Oliver, D.P., Forrester, S., Chittleborough, D.J., Hutson, J.L., Kookana, R.S., and Ostendorf, B. 2012. The effect of terrain and management on the spatial variability of soil properties in an apple orchard. Catena, 93: 38-48.
  41. Van Looy, K., Bouma, J., Herbst, M., Koestel, J., Minasny, B., Mishra, U., Montzka, C., Nemes, A., Pachepsky, Y.A., Padarian, J. and Schaap, M.G. 2017. Pedotransfer functions in Earth system science: Challenges and perspectives. Reviews of Geophysics, 55: 1199-1256.
  42. Vapnik, V.N. 1995. The nature of statistical learning theory. Springer, New York.
  43. Waiser, T.H., Morgan, C.L.S., Brown, D.J., Hallmark, C.T. 2007. In situ characterization of soil clay content with visible near-infrared diffuse reflectance spectroscopy, Soil Science Society of America Journal, 71: 389–396.
  44. Wang, C., Zhao, L., Fang, H., Wang, L., Xing, Z., Zou, D., Hu, G., Wu, X., Zhao, Y., Sheng, Y., Pang, Q., Du, E., Liu, G., and Yun, H. 2021. Mapping surficial soil particle size fractions in alpine permafrost regions of the Qinghai–Tibet plateau. Remote Sensing, 13: 1392.
  45. Wang, D., Yang, H., Qian, H., Gao, L., Li, C., Xin, J., Tan, Y., Wang, Y., and Li, Z. 2023. Minimizing vegetation influence on soil salinity mapping with novel bare soil pixels from multi-temporal images. Geoderma, 439, 116697
  46. Wang, Z., and Shi, W. 2017. Mapping soil particle-size fractions: A comparison of compositional kriging and logratio kriging. Journal of Hydrology, 546: 526-541.
  47. Wilding, L.P. 1985. Spatial variability: Its documentation, accommodation and implication to soil survey. P. 166–189. In D.R. Nielsen and J. Bouma (ed). Soil spatial variability. Pudoc, Wagenigen.
  48. Wu, W., Li, A.D., He, X.U., Ma, R., Liu, H.B., and Lv, J.K. 2018. A comparison of support vector machines, artificial neural network and classification tree for identifying soil texture classes in southwest China. Computers and Electronics in Agriculture, 144: 86-93.
  49. Wulder, M.A., White, J.C., Loveland, T.R., Woodcock, C.E., Belward, A.S., Cohen, W.B., Fosnight, E.A., Shaw, J., Masek, J.G., and Roy, D.P. 2016. The global Landsat archive: status, consolidation, and direction. Remote Sensing of Environment, 185: 271–283.
  50. Xuemei, L., and Jianshe, L. 2013. Measurement of soil properties using visible and short wave-near infrared spectroscopy and multivariate calibration. Measurement, 46: 3808–3814.
  51. Zhang, M., Shi, W., and Xu, Z. 2020. Systematic comparison of five machine-learning models in classification and interpolation of soil particle size fractions using different transformed data. Hydrology and Earth System Sciences, 24: 2505-2526.

 

 

پژوهش های خاک
دوره 38، شماره 1
مرداد 1403
صفحه 77-97

فایل ها

هم رسانی

ارجاع به این مقاله

آمار