برآورد رطوبت های ظرفیت زراعی، پژمردگی دائم و قابل استفاده در خاک های دشت اردبیل با استفاده از مدل های رگرسیونی و شبکه عصبی مصنوعی

نوع مقاله: مقاله پژوهشی

نویسندگان

1 دانشجوی کارشناسی ارشد، گروه علوم خاک، دانشگاه محقق اردبیلی

2 استادیار، گروه علوم خاک، دانشگاه محقق اردبیلی

3 استادیار، گروه مکانیک ماشینهای کشاورزی، دانشگاه محقق اردبیلی

4 استادیار، گروه مرتع و آبخیزداری، دانشگاه محقق اردبیلی

چکیده

اندازه­گیری مستقیم ویژگی­های هیدرولیکی خاک وقت­گیر، پرهزینه­ و گاهی اوقات به دلیل خطاهای آزمایشی و عدم یکنواختی خاک غیر واقعی است. در عوض، این ویژگی­ها می­توانند از روی ویژگی­های زودیافت خاک مانند توزیع اندازه ذرات خاک، جرم ویژه ظاهری، کربن آلی و کربنات کلسیم معادل با استفاده از توابع انتقالی خاک برآورد شوند. هدف از این پژوهش، ارائه مدل­های رگرسیونی و شبکه عصبی مصنوعی بر اساس ویژگی­های زودیافت یاد شده برای برآورد ویژگی­های دیریافت شامل رطوبت ­های ظرفیت زراعی، پژمردگی دائم و قابل استفاده در شماری از خاک­های دشت اردبیل بود.برای این منظور 100 نمونه خاک برداشته شد سپس برخی ویژگی­های فیزیکی و شیمیایی آنها اندازه­گیری شد. داده­ها به دو سری داده­های آموزشی (80 نمونه) و داده­های آزمونی (20 نمونه) تقسیم شدند. برای ایجاد مدل­های شبکه عصبی از نرم­افزار 5 Neurosolution و برای ایجاد مدل­های رگرسیونی از نرم افزار SPSS استفاده شد. مقادیر ضریب تبیین (R2) و مجذور میانگین مربعات خطا (RMSE) در تخمین پارامترهای دیریافت شامل رطوبت­های ظرفیت زراعی، پژمردگی دائم و قابل استفاده به ترتیب برابر 82/0 و 29/2، 82/0 و 38/1، 57/0 و 97/1 برای بهترین مدل­ رگرسیونی و به ترتیب برابر 87/0 و 9/1، 90/0 و 02/1، 73/0 و 56/1 برای بهترین مدل­ شبکه عصبی مصنوعی بود. مقادیر R2 و RMSE برای نتایج مدل­های رگرسیونی و شبکه عصبی مصنوعی نشان داد که هر دو روش می­توانند ضرایب رطوبتی خاک را با دقت مناسبی برآورد کنند. با این حال، مدل­های رگرسیونی در برآورد رطوبت قابل استفاده کارآیی لازم را نداشتند. دقت تخمین ضرایب رطوبتی توسط مدل­های شبکه عصبی مصنوعی بیشتر از مدل­های رگرسیونی بود. 

کلیدواژه‌ها


عنوان مقاله [English]

Estimating of Field Capacity, Permanent Wilting and Available Water Content in Ardabil Plain Soils using Regression and Artificial Neural Network Models

نویسندگان [English]

  • Hamed Amir Abedi 1
  • Shokrollah asghari 2
  • Tarahhom Mesri Gandoshmin 3
  • Farshad Keivan behjo 4
1 MSc. Student of Soil Science, College of Agriculture, Mohaghegh Ardabili University
2 Assis. Prof. Department of Soil Science, Mohaghegh Ardabili University
3 Assis. Prof, Department of Mechanical Engineering, Mohaghegh Ardabili University
4 Assistant Professor, Faculty of Natural Resources, Mohaghegh Ardabili University
چکیده [English]

Direct measurement of soil hydraulic properties is consuming, costly and sometimes unreliable because of soil heterogeneity and experimental errors. These properties can be estimated from surrogate data such as particle size distribution, bulk density, organic carbon and CaCO3 using pedotransfer functions (PTFs). The objective of this research was presentation of regression and neural network models for estimation of missing soil properties including field capacity, permanent wilting and available water contents from above-cited surrogate soil properties in some soil of  the Ardabil Plain. Total 100 soil samples were taken and then some physical and chemical properties of them measured. Soil samples were divided into two groups as 80 for the development and 20 for the validation of PTFs. Neural network and regression models were made using Neurosolution5 and SPSS softwares, respectively. The values of determination coefficient (R2) and root mean square error (RMSE) for the estimation of field capacity, permanent wilting and available water contents were obtained 0.82 and 2.29, 0.82 and 1.38, 0.57 and 1.97 in the best regression models and 0.87 and 1.9, 0.90 and 1.02, 0.73 and 1.56 in the best neural network models, respectively. The values of R2 and RMSE for the results of regression and artificial neural network PTFs showed that both models can be applied to predicting missing soil properties. Regression models hadn’t any efficiency to predict available water content. Artificial neural networks were performed better than regression models in this case.

کلیدواژه‌ها [English]

  • artificial neural networks
  • field capacity water content
  • permanent wilting water content
  • Regression
  • Soil Pedotransfer Function
References

Bauer A and Black AL. 1992. Organic carbon effects on available water capacity of three soil textural groups. Soil Sci. Soc. Am. J. 56׃ 248-254.

Bouma J. 1989. Using soil survey data for quantitative land evaluation. Adv. Soil Sci. 9:177­-213.

Campbell GS. 1985. Soil Physics with Basics, Elsevier press, Amsterdam.

Ganbarian Alavijeh  B and  Liaghat A.  2011. Evaluation of pedotransfer functions and effect of organic matter in prediction of soil saturated water content.J. Water Soil. 25(5): 1016-1024. (in Farsi with English Summary)

Gardner, WH. 1986. Water content. In: Methods of soil analysis. Part1. Physical and mineralogical methods. 2nd Ed. (ed. A. Klute). Madison, WI: Am. Soc. Agron. 493-544.

Haghverdi A, Ghahraman B, Khoshnood Yazdi AA and Arabi Z. 2010. Estimating of water content in FC and PWP in north and north east of Iran's soil samples using k-nearest neighbor and artificial neural networks. Journal of Water and Soil, 24 (4): 804-814. (In Farsi with English Summary)

Hillel D. 2004. Introduction to environmental soil physics. Elsevier Academic Press. 494 p.

Hutson JL and Cass A. 1987. A retentivity function for use in soil-water simulation models. J. Soil Sci. 38: 105–113.

Khodaverdiloo H, Homaee M, Th. van Genuchten M and Ghorbani Dashtaki Sh. 2011. Deriving and validating pedotransfer functions for some calcareous soils. J. Hydro. 399: 93-99.

Klute A. 1986. Methods of soil analysis. Part 1. Physical and mineralogical methods. 2nd  Ed. Agron. Monog. 9. ASA and SSSA, Madison, WI.

McBratneyAB, Minasny B, Cattle SR and Vervoort RW. 2002. From pedotransfer function to soil inference system. Geoderma, 109: 41–73.

Merdun H, Cinar O, Meral R and Apan M. 2006. Comparison of artificial neural network and regression pedotransfer functions for prediction of soil water retention and saturated hydraulic conductivity. Soil Till. Res. 90: 108–116.

Mesri Gundoshmian T. 2009. The use of intelligent systems to optimize drop Combine. PhD Thesis, Engineering of Agriculture Machinery Mechanization Department, University of Tabriz. (In Farsi with English Summary)

Minasny B and Mcbartney AB. 2002. The neuro-m method for fitting neural network parametric pedotransfer functions. Soil Sci. Soc. Am. J. 66: 352-361.

Minasny B, McBratneyAB and Bristow KL. 1999. Comparison of different approaches to the development of pedotransfer functions for water retention curves. Geoderma, 93: 225–253.

Mosaddeghi MR and Mahboubi AA. 2011. Point pedotransfer functions for prediction of water retention of selected soil series in a semi-aired region of western Iran. Archives of Agronomy and Soil Science. 57(4): 327-342.

Navvabian M, Liaghat EM and Homaie M. 2004. Rapid estimation of hydraulic conductivity using neural networks. In: Proceedings of the Second National Student conference on Water and soil Resources. Shiraz Uni. Press, Pp: 203-211.

Pachepsky YA, Timlin DJ and Varallyay G. 1996. Artificial neural networks to estimate soil water retention from easily measurable data. Soil Sci. Soc. Am. J. 60: 727–773.

Page AL (ed.).1985. Methods of soil analysis. Part 2. Chemical and microbiological methods. Agronomy No. 9. American Society of Agronomy, Madison, WI.

Ramezani M, Ganbarian B, Liaghat AM and Salehi Khoshkroudi Sh. 2011. Developing pedotransfer functions for saline and saline- alkali soils. J. Water and Irrigation Manag. 1(1): 99-110. (In Farsi with English Summary)

Sarmadian F, Taghizade R, Asghari H and Akbarzade A. 2010. Comparison neuro-fuzzy, neural network and regression stepwise methods in predicting some soil properties. J. Water Soil Res. 41 (1): 211-220. (In Farsi with English Summary)

Schaap M G, Leij FJ and Van Genuchten MTh. 2001. Rosetta: a computer program for estimating soil hydraulic parameters with hierarchical pedotransfer functions. J. Hydrol. 251: 163–176.

Schaap MG and Bouten W. 1996. Modeling water retention curves of sandy soils using neural networks. Water Resour. Res. 32: 3033–3040.

Schaap MG, Leij FJ and Van Genuchten MTh. 1998. Neural network analysis for hierarchical prediction of soil hydraulic properties. Soil Sci. Soc. Am. J. 62:847–855.

Shirani H and Rafienejad N. 2012. Estimating of some missing soil properties with regression pedotransfer functions and neural network in the Kerman. J. Soil Res. 25. (4): 349-359. (In Farsi with English Summary)

Shirazi MA and Boersma L. 1984. A unifying quantitative analysis of soil texture. SSSA. J. 48: 142–147.

Tamari, S, Wo¨sten, JH M and Ruiz-Suarez JC. 1996. Testing an artificial neural network for predicting soil hydraulic conductivity. Soil Sci. Soc. Am. J. 60: 1732–1741.

Van Genuchten, MTh. 1980. A closed-form equation for predicting the hydraulic conductivity of unsaturated soils. Soil Sci. Soc. Am. J. 44: 892–898.

Warrick AW. 2002. Soil physics companion. CRC Press. 389 p.