مقایسه کارایی مدلهای رگرسیونی، شبکه عصبی مصنوعی و تلفیق آن با الگوریتم ژنتیک در بررسی فرسایش بادی
محورهای موضوعی : مدیریت آب در مزرعه با هدف بهبود شاخص های مدیریتی آبیاریشاهین ابراهیمی 1 , علی محمدی ترکاشوند 2 , مهرداد اسفندیاری 3 , عباس احمدی 4
1 - دانشجوی دکتری، گروه خاکشناسی، واحد علوم و تحقیقات، دانشگاه آزاد اسلامی، تهران، ایران.
2 - گروه علوم و مهندسی خاک، واحد علوم و تحقیقات، دانشگاه آزاد اسلامی، تهران، ایران.
3 - گروه علوم و مهندسی خاک، واحد علوم و تحقیقات، دانشگاه آزاد اسلامی، تهران، ایران.
4 - گروه خاکشناسی، دانشکده کشاورزی، دانشگاه تبریز، تبریز، ایران.
کلید واژه: ماده آلی, رگرسیون خطی چند متغیره, ریزگرد, رس, پرسپترون,
چکیده مقاله :
زمینه و هدف: فرسایش بادی در بخش بزرگی از پهنه ایران وجود داد که سبب تخریب اراضی و کاهش باروری آنها به همراه اثرات زیست‌محیطی شده است. شناخت مناطق حساس به فرسایش می‌تواند در برنامه‌ریزی‌های حفاظت خاک به کمک مدیران منابع طبیعی و محیط‌زیست آید.روش پژوهش: این تحقیق برای برآورد جزء فرسایش‌پذیر خاک در مقابل باد (EF) از روی ویژگی‌های سهل الوصول خاک دردشت الله آباد واقع در شرق استان قزوین انجام شد. بدین منظور جزء فرسایش‌پذیر خاک در مقابل باد با استفاده از روش‌های رگرسیون چند متغیره (MLR)، شبکه عصبی مصنوعی (ANN) و تلفیق شبکه عصبی مصنوعی با الگوریتم ژنتیک برای بهینه‌سازی اوزان (GA-ANN) با به کار بردن ویژگی‌های سهل الوصول برآورد شد. با بررسی نقشه خاک، تفاوت خاک ها و خصوصیات محیطی دشت الله‌آباد، 103 نمونه خاک طبق یک الگوی تصادفی طبقه بندی شده از 10 سانتی‌متری سطح آنها، جمع‌آوری و به آزمایشگاه ارسال شد. در نمونه های خاک، برخی خصوصیات خاک به‌عنوان ورودی های مدل های برآورد جزء فرسایش‌پذیر خاک در مقابل باد اندازه‌گیری گردید. ورودی های هر مدل شاملpH، ECe، CCE، SAR، جرم مخصوص ظاهری، ذرات شن، سیلت و رس، ذرات درشت خاک با قطر کمتر از 2 میلی متر و ماده آلی بودند. دقت و قابلیت اعتماد نتایج مدل های ایجاد شده با توجه به معیارهای ضریب تبیین، مجذور مربعات خطا، آزمون مورگان-گرنجر- نیوبلد و شاخص آکایک مورد مقایسه قرار گفتند.یافته ها: طبق یافته ها، بیشترین همبستگی جزء فرسایش‌پذیر خاک در مقابل باد (EF) با مقدار رس خاک دیده شد (789/0- r=). همچنین جزء فرسایش‌پذیر خاک با خصوصیات دیگر خاک شامل pH، هدایت الکتریکی، SAR، مقدار ماده آلی و جرم مخصوص ظاهری، همبستگی نشان داد، این همبستگی با سه خصوصیت SAR، ماده آلی و رس در سطح یک درصد همبستگی معنی دار بود. مدلهای ایجاد شده با هر سه روش توانایی بسیار بیشتری در پیش بینی EF در سری داده های آزمون نسبت به داده های سری آموزش داشتند. همچنین نتایج نشان داد که مدل شبکه عصبی از دقت بیشتر و خطای تخمین کمتری در مقایسه با مدل های هیبرید و رگرسیون به‌دست آمده است. نتایج آنالیز حساسیت نیز نشان داد که بیشترین حساسیت مدل به متغیرهای ورودی در مدل ANN، به ترتیب مربوط به ماده آلی و SAR و در مدل GA-ANN مربوط به متغیر میزان رس خاک بود.نتیجه گیری: بر طبق نتایج، تنها مدل رگرسیون در مرحله آموزش دارای R2 بیشتر از 50 درصد (R2=0.56) در برآورد جزء فرسایش پذیری خاک بود که البته این مقدار (R2=0.56) نیز قابل اعتماد نیست. با توجه به نتایج مرحله آزمون، هر سه مدل به کار رفته شامل رگرسیون، شبکه عصبی مصنوعی و تلفیق آن با الگوریتم ژنتیک در برآورد شاخص جزء فرسایش‌پذیر خاک از کارایی مناسبی برخوردار نمی‌باشند به‌طوری که بالاترین ضریب تبیین (R2) در مدل شبکه عصبی در مرحله آزمون (R2 = 0.43)، صحت کمتر از 50 درصد در تخمین EF داشت که نمی تواند صحت مناسبی در پیش بینی جزء فرسایش پذیری بادی خاک باشد.
Background and Aim: Wind erosion has occurred in a large part of Iran, which has caused land degradation and reduced fertility along with environmental effects. Identifying erosion-sensitive areas can help natural resource and environmental managers in soil conservation planning.Methods:This study is a step to estimate the erodible component of soil against the wind (EF) from soil accessibility characteristics in Allahabad plain located in the east of Qazvin province. For this purpose, the soil erodibility component, which is closely related to soil erosion versus wind, using multivariate regression (MLR), artificial neural network (ANN), and artificial neural network with genetic algorithm for weight optimization (GA-ANN) were estimated using accessible characteristics. Regarding soil map, soil differences, and environmental characteristics of Allahabad plain, 103 soil samples were collected according to a stratified random pattern of 10 cm of soil surface. In soil samples, some soil properties were measured as inputs of models for estimating erodible soil components against the wind. The inputs of each model included pH, ECe, CCE, SAR, bulk density, sand particles, silt and clay, coarse soil particles with a diameter of more than 2 mm, and organic matter. Accuracy and reliability of the results of the created models were compared with each other according to the criteria of coefficient of determination, square of error, Morgan-Granger-Newbold and Akaike information criterion.Results: Based on data, the highest correlation between soil erodible fraction to wind erosion (EF) was observed with soil clay content (r = -0.789). Also, soil erodible components showed a correlation with other soil properties including pH, electrical conductivity, SAR, organic matter, and the should be omitted density. This correlation was significant with three properties of SAR, organic matter, and clay at a should be added 1% level. The models created by the three methods were much more capable of predicting EF in the test data series than the training series data. The results also showed that the neural network model had a should be omitted more accuracy and less estimation error compared to hybrid and regression models. The results of sensitivity analysis of the models also showed that the highest sensitivity of the model to input variables in the ANN model, related to organic matter and SAR, respectively, and in the model GA-ANN was related to soil clay content variable.Conclusion: According to the results, R2 in the regression model of training data was more than 50% in estimating EF, but this value (R2 = 0.56) is not reliable. According to the test data, all three models, including regression, artificial neural network, and its combination with genetic algorithm had not been efficient enough in estimating EF, so that can be omitted the highest R2 in the neural network model in the test data (R2 = 0.43) had an accuracy of less than 50% in estimating the EF, which cannot be an appropriate accuracy in predicting EF.
Ahmadi A. 2011. Artificial neural networks applicability in erosion and runoff simulation using fractal dimensions. Thesis of Department of Soil Science, Agricultural College, University of Tabriz. P.197 (in Persian with English abstract).
Akaike, H (1974). "A new look at the statistical model identification". IEEE Transactions on Automatic Control. 19(6): 716-723.
Ananani, M., Amirian Chakan, A., Faraji, M. Yousefi Khaneghah, S.H. (2018). Use of erodibility indices and surface ridges in the study of soil sensitivity to wind erosion. 15th Iranian Soil Science Congress, Isfahan University of Technology, Isfahan, Iran.
Ayoubi S, Shahri AP, Karchegani PM, Sahrawat KL (2011) Application of artificial neural network (ANN) to predict soil organic matter using remote sensing data in two ecosystems (pp. 181-196). InTech Open Access.
Azimzadeh, H.R., Ekhtesasi, M.R., Hatami, M., Qalibaf, M.A. (2002). Wind erosion: erodibility relation to soil physical and chemical properties in Iran central plain (yazd-ardakan plain - case study). Journal of Agricultural Sciences and Natural Resources, 9(1): 139-151.
Bayat, H. 2009. Establishment of transfer functions for predicting moisture curve through artificial neural networks (ANNs) and group data management (GMDH) using fractal parameters and principal component analysis. PhD Thesis, Department of Soil Science, Faculty of Agriculture, University of Tabriz.
Besalatpour AA, Ayoubi S, Hajabbasi MA, Mosaddeghi MR, Schulin R (2013) Estimating wet soil aggregate stability from easily available properties in a highly mountainous watershed. Catena 111:72-79.
Besalatpour, A. A., S. Ayoubi, M. A. Hajabbasi, A. Yousefian Jazi, and A. Gharipour. 2014. Feature selection using parallel genetic algorithm for the prediction of geometric mean diameter of soil aggregates by machine learning methods. Arid Land Research and Management 28:383-94.
Black, G. R., and K. H. Hartge. 1986. Bulk Density. In: Methods of Soil Analysis, Part 1. Physical and Mineralogical Methods, Agronomy Monograph no. 9 (2nd Edition). 363-375.
Blanka, V., Mezõsi, G., and Meyer, B. 2013. Projected changes in the drought hazard in Hungary due to climate change, Idõjárás: Quarterly J. Hungarian Meteorol. Serv. 117: 219-237.
Bouma J, 1989. Using soil survey data for quantitative land evaluation Advances in soil science. Springer, 177-213.
Canasveras JC, Barrón V, Del Campillo MC, Torrent J, Gómez JA. 2010. Estimation of aggregate stability indices in Mediterranean soils by diffuse reflectance spectroscopy. Geoderma. 158(1):78-84.
Chandler D.G., K.E. Saxton and A.J. Busacca. 2005. Predicting wind erodibility of loessial soils in the Pacific Northwest by particle sizing. Arid Land Research and Management, 19(1): 13-27.
Chepil W, 1958. Soil conditions that influence wind erosion. USDATech, Bui, No, 1185.
Chepil W, 1960. Conservation of relative field erodibility to annual soil loss by wind. Soil Science Society of American Journal, 24(2):143-148.
Colazo, J.C. and Buschiazzo, D.E. 2010. Soil dry aggregate stability and wind erodible fraction in a semiarid environment of Argentina. Geoderma, 159 (1-2): 228-236.
Coppinger K D, Reiners W A, Burke I C and Olson R K, 1991. Net erosion on a sagebrush steppe landscape as determined by cesium-137 distribution. Soil Science Society of America Journal, 55: 254-258.
Crowe A.M., McClean C.J., and Cresser M.S. 2006. An application of genetic algorithms to the robust estimation of soil organic and mineral fraction densities. Environ. Model. Software. 21: 1503-1507.
Eshtehardian E., Afshar A., and Abbasnia R. 2006. A genetic algorithm-based optimizing approach for project time-cost trade-off with uncertain measure. 3th International Project Management Conference, Tehran, Iran. (in Persian with English abstract).
Gee GW and Or D, 2002. Particle-size analysis. In: Warren AD (ed.) Methods of Soil Analysis. Part 4. Physical Methods. Pp.255-295, Soil Sci. Soc. Am. Inc.
Gomez, L., J.L. Arrue, M.V. Lopez, G. Sterk, D. Richard, R. Gracia, M. Sabre, A. Gaudichet and J.P. Frangi. 2003. Wind erosion in a semiarid agricultural area of Spain the WELSONS project. Catena, 52: 235-256.
Harvey, M.A.D.I., 1997. The evaluation of economic forecasts. Ph.D. Thesis, University of Nottingham. P.276.
Haykin, S. 1999. Neural Networks andLearning Machines. 3rd edition, Pearson, Prentice Hall, Upper Saddle River, New Jersey 07458, United States.
Hojjatnooghi, F., Shirani, H., Pazira, E., Besalatpour, A. 2019. Identification of Soil Properties Influencing Some Soil Physical Quality Indicators Using Hybrid PSO-ICA-SVR Algorithm in Some Agricultural Land Uses of Kerman Province, Iran. Communications in Soil Science and Plant Analysis 50 (16), 1986-2002.
Huang, M., Peng, G., Zhang, J., Zhang, S. 2006.Application of artificial neural networks to the prediction of dust storms in Northwest China. Global and Planetary Change, 52 (1-4): 216-224.
Huawei P, David R. Huggins, Brenton. 2020. Wind erosion of soil influenced by clay amendment in the inland Pacific Northwest, USA. Land Degradation and Development, 32 (1): 241-255.
Khanbabakhani, L., Mohammadi Torkashvand, A. and Mohmoodi, M.A. 2018. Predictionof soil texture using artificial neural networks. Soil and Water Resources Conservation, 8(1): 1-10.
Kim, M. and Gilley, J.E. 2018. Artificial Neural Network estimation of soil erosion and nutrient concentrations in runoff from land application areas. Computers and Electronics in Agriculture, 64 (2): 268-275.
Klute, A.1986. Methods of Soil Analysis: Part 1. Physical and Mineralogical Methods (Sssa Book Series No 5)2nd Edition, American Society of Agronomy and Soil Science Society of America, Madison, WI.
Kokelj, S. V., Lantz, T. C., Solomon, S., Pisaric, M. F. J., Keith, D., Morse, P., Thienpont, J, R., Smol, J. P., Esagok, D., 2012. Using multiple sources of knowledge to investigate northern environmental change: regional ecological impacts of a storm surge in the outer Mackenzie delta, N.W.T. Arctic, 65(3): 257 – 272.
Leenders J.K., Visser S.M., and Stroosnijder L. 2005. Farmers' perceptions of the role of scattered vegetation in wind erosion control on arable land in Burkina Faso. Land Degradation & Development 16(4): 327-337.
Liu S.h., Butler D., Brazier R., Heathwaite L., and Khu S. 2007. Using genetic algorithm to calibrate a water quality model. Sci. Total Environ. 374: 260-272.
Lo C.H., Chan P.T., Wong Y.K., Rad A.B., and Cheung K.L. 2007. Fuzzy-genetic algorithm for automatic fault detection in HVAC systems. Applied Soft Computing 7: 554–560.
Lyles, L. and R. L. Schrandet. 1971. Wind erodibility as influence by rainfall and salinity. Soil Science, 114: 367-372.
Majdi, H., Karimian Iqbal, M., Karimzadeh, H.R., Jalalian, A. (2006). The effect of different types of clay mulch on the amount of wind eroded materials. Journal of Agricultural Science and Technology and Natural Resources, 10(3): 137-148.
Marashi, M., Mohammadi Torkashvand, A., Ahmadi, A., Esfandiari, M. 2017. Estimation of soil aggregate stability indices using artificial neural network and multiple linear regression models. Spanish Journal of Soil Sciences. 7(2): 89-99.
Marashi, M., Mohammadi Torkashvand, A., Ahmadi, A., Esfandiari, M. 2019. Adaptive neuro-fuzzy inference system: estimation of soil aggregates stability. Acta Ecologica Sinica. 39(1): 95-101.
Merdun, H., O. Cinar, R. Meral, and M. Apan. 2006. Comparison of artificial neural network and regressionpedotransfer functions for prediction of soil water retention and saturated hydraulic conductivity. Soil and Tillage Research 90:108–16.
Mezosi, G., Blanka, V., Bata1, T., Kovács, F., and Meyer, B. 2013. Estimation of regional differences in wind erosion sensitivity in Hungary. J. Natur. Hazards Earth Syst. Sci. 1: 4713-4750.
Minasny, B., Hopmans, J.W., Harter, T., Eching, S.O., Tuli, A. and Denton, M.A., 2004. Neural networks predication of soil hydraulic functions for alluvial soils using multistep outflow data. Soil Sci. Soc. Am. J., 68:417-429.
Mohammadi Torkashvand, A. and Nikkami, D. 2008. Investigating some methodologies of preparing erosion features map by using RS and GIS. International Journal of Sediment Research, 23 (2): 124-132.
Mohammadi Torkashvand, A., Haghighat, N. 2009. Investigation of some models derived from data layers integration in geographic information system with slope layer for providing water-soil erosion types maps. Research Journal of Environmental Sciences 3 (2), 202-209.
Mokhtari P, Ayoubi SH, Honarju N, Jalalian A (2011) Predicting soil organic matter by artificial neural network in landscape scale using remotely sensed data and topographic attributes. Geophysical Res Abs 13: 10-75.
Munns, R., 2003. Comparative physiology of salt and water stress. Plant Cell Environment, 25: 239-50.
Munson, S. M., Belnap, J. and Okin, S. G., 2011. Responses of wind erosion to climate-induced vegetation changes on the Colorado Plateau. Proceedings of the National Academy of Sciences of the United States of America, 108(10): 3854–3859.
Naghizade Asl F, Asgari HR, Emami H, Jafari M. 2019. Combined effect of micro silica with clay, and gypsum as mulches on shear strength and wind erosion rate of sands. International Soil and Water Conservation Research, 7(4): 388-394.
Nelson D W, and Sommers L E, 1982. Total carbon, organic carbon, and organic matter: 539-579, In: Page AL (ed.) Methods of Soil Analysis. Part 2. 2nd ed. American Society of Agronomy, Madison,WI .
NeuroSolutions, 2005. Getting Started Manual Version 4. Neurodimension, Inc.1800 N. Main Street, uite D4, Gainesville, FL 32609.
Nosrati, H. and Eftekhari, M. (2014). A new approach for variable selection using fuzzy logic. Computational Intelligence in Electrical Engineering,4, 71 -83.
Palizvanzand H. and Ahmadi A. 2016. Comparison of linear regression, Fuzzy and Fuzzy-genetic models to predict soilcation exchange capacityHabib. Applied Soil research, 3(2): 86-100. In Persian.
Panayiotopoulos KP, Barbayiannis N, Papatolios K (2004) Influence of electrolyte concentration, sodium adsorption ratio, and mechanical disturbance on dispersed clay particle size and critical flocculation concentration in Alfisols. Commun Soil Sci Plant Anal 35:1415–1434.
Qiang, M., Chen, F., Zhou, A., Xiao, S., Zhang, J., & Wang, Z. (2007). Impacts of wind velocity on sand and dust deposition during dust storm as inferred from a series of observations in the northeastern Qinghai–Tibetan Plateau, China. Powder Technology, 175(2), 82-89.
Refahi, H. 2006. Soil Water Erosion and Control. Tehran University Press, Tehran, Iran.
Sarkar, T., Mishra, M. 2018. Soil erosion susceptibility mapping with the application of logistic regression and artificial neural network. Journal of Geovisualization and Spatial Analysis, 2: 1-17.
Shahabinejad N, Mahmoodabadi M, Jalalian A, Chavoshi E. 2019. The fractionation of soil aggregates associated with primary particles influencing wind erosion rates in arid to semiarid environments. Geoderma, 356, Article 113936.
Shahabinejad N, Mahmoodabadi M, Jalalian A, Chavoshi E. The Influence of Soil Properties on the Wind Erosion Rate at Different Regions of Kerman Province. JWSS. 2020; 24 (3) :209-222.
Sirjani E, Sameni A, Moosavi AA, Mahmoodabadi M, Laurent B. 2019. Portable wind tunnel experiments to study soil erosion by wind and its link to soil properties in the Fars province, Iran. Geoderma, 333: 69-80.
Sparks, D. and Bartels, J. 1996. Methods of soil analysis: Part3. Chemical methods. Soil Science Society of America Book Series. Madison, Wisconsin. USA. Pp 1390.
Stout, J. E., & Zobeck, T. M. (1996). Establishing the threshold condition for soil movement in wind-eroding fields. Paper presented at the Proceedings of International Conference on Air Pollution from Agricultural Operations.
Tung C., Hsu S., Liu C.M., and Li Sh.Jr. 2003. Application of the genetic algorithm for optimizing operation rules of the LiYutan reservoir in Taiwan. J. Am. Water Resour. Assoc. 39 (3): 649–657.
Wösten, J.H.M., Pachepsky, Y.A., Rawls, W.J., 2001. Pedotransfer functions: bridging the gap between available basic soil data and missing soil hydraulic characteristics. J. Hydrol. 251(3),123-150.
Yan H, Wang H, Wang C, Zhang G, Patel N. 2005. Losses of soil organic carbon under wind erosion in China. Global change biology, 11, 5: 828-840.
Yulevitch G, Danon B, Krasovitov B, Nitzan A, Swet A, Tsesarsky M, Katra I. 2020. Evaluation of wind-induced dust-PM emission from unpaved roads varying in silt content by experimental results. Atmospheric Pollution Research, 11(2): 261-268.
Zare Abyaneh H., Bayat Varkeshi M. Marofi S. Ildromi A.R. 2010. Simulation of Malayer Plain Groundwater Level Based on Weather Data Using Artificial Neural Network. Natural Geographic Research Journal, 78, 17-38.
Zobeck TM, Baddock M, Scott R, Pelt V, Tatarko J, Acosta-Martinez V. 2013. Soil property effects on wind erosion of organic soils. Aeolian Research, 10, 43-51.
_||_