اقلیم و بوم سازگان مناطق خشک و نیمه خشک

اقلیم و بوم سازگان مناطق خشک و نیمه خشک

مدل‌سازی تخریب جنگل با‌ استفاده از رگرسیون لجستیک (مطالعه موردی: روستای اولادقباد، شهرستان کوهدشت)

نوع مقاله : مقاله پژوهشی

نویسندگان
1 literacy movement street, third culture alley
2 استادیار اقتصاد جنگل، دانشکده علوم کشاورزی و منابع‌طبیعی، دانشگاه لرستان، خرم آباد، ایران
3 ستادیار دانشکده کشاورزی و منابع‌طبیعی، دانشگاه لرستان، خرم آباد، ایران.
چکیده
زمینه و هدف: این پژوهش با‌ هدف تعیین سطح تخریب جنگل، طی دوره زمانی 1372 تا 1392 و امکان مدل‌سازی تغییرات گستره جنگل و ارتباط آن با‌عوامل فیزیوگرافی و برخی عوامل انسانی با استفاده از رگرسیون لجستیک انجام گرفت.
روش بررسی: بعد از تصحیحات هندسی تصاویر، طبقه‌بندی آن با ‌استفاده از الگوریتم حداکثر احتمال به دو کلاسه جنگل و غیر جنگل از تصاویر ‌سال‌ 1372 سنجنده TM و 1392 سنجنده OLIبا صحت کلی برای هر سال به ترتیب 81، 82 درصد و ضریب کاپا 72 و 71 درصد تهیه‌‌ شد. نقشه تغییرات گستره جنگل از تقاطع دو نقشه حاصل به دست آمد.
یافته ها: نتایج‌ نشان داد که در طول 20 سال حدود 528 هکتار و به میزان 8/9 درصد از سطح جنگل‌های منطقه کاهش یافته ‌است. در مدل‌سازی، نقشه تغییرات متغییر وابسته و متغیرهای توپوگرافی شیب، ارتفاع از سطح دریا و متغیرهای انسانی فاصله از جاده و روستا به عنوان متغییر مستقل وارد مدل ‌شدند. نتایج مدل‌سازی رگرسیون لجستیک با Pseudo R2 برابر با 22/0 و ضریب ROC برابر با 73/0 نشان دهنده توافق نسبی مدل با‌ نقشه تغییرات گستره جنگل است.
بحث نتیجه گیری: متغیرهای فاصله از روستا، ارتفاع از سطح دریا و شیب با‌ مقدار تخریب رابطه عکس دارند. اما با ‌افزایش فاصله از جاده در این منطقه مقدار تخریب افزایش می‌یابد. از آن جایی که فاکتور جمعیت در روند تخریب جنگل تأثیر مهمی دارد، توصیه می‌شود در مدل‌سازی تخریب استفاده شود.
کلیدواژه‌ها
موضوعات

عنوان مقاله English

Modeling Deforestation Using Logistic Regression (Case Study: Kuhdasht, Lorestan Province)

نویسندگان English

zohreh bazvand 1
kamran Adeli 2
Ziaedin Badehian 3
1 literacy movement street, third culture alley
2 Assistant professor of College of Agriculture and Natural Resources,Lorestan University, Iran
3 Assistant professor of College of Agriculture and Natural Resources,Lorestan University, Iran.
چکیده English

Background and Objectives: This study aims to determine the forest distribution and area over the period 1993-2013 and model the possibility of changes in forest. For this purpose, the relationship of forest changes with physiographic factors and many human factors using logistic regression were studied.
Methods: After geo-correction of the images and their classification using maximum likelihood algorithm, forest land use map related to 1993-2013 period was prepared. The map of forest changes was derived from intersection of the two maps.
Findings: To investigate the spatial relationship between forest changes and physiographic and human factors, logistic regression was used with slope and elevation as topographic variables, and distance from roads and village as human variables. Forest area has been 13250 ha which has decreased about 528 ha (equal to 9.8%) during the 20 years. Relative agreement between obtained model and the map of forest changes map by Pseudo R2 and ROC coefficient was equal to 0.22 and 0.73, respectively.
Conclusion: Distance from village, elevation and slope variables had negative relationship with the rate of destruction. However, the rate of destruction increases with increasing distance from the roads. Since the population influences the process of deforestation, it is recommended to use this factor in destruction evaluation and modeling.

[1]          Amini, M. et al. (2008). Investigating the possibility of modeling the probability of destruction of western forests   of forest degradation using logistic regression, GIS and remote sensing. Geography and Development Magazine, 29: 31-42.
[2]          Bagheri, R., & Shtaei, Sh. (2010). Modeling of widespread forest degradation using logistic regression (Case study: Chehel-Chay watershed, Golestan province). Iranian Forestry Association, 2(3), 243-252.
[3]          Kamran, K., & Bayram, C. H. (2015). Monitoring and estimating the trend of land use change/land cover using Markov Chain model (Case study: Bilastag plain- South Khorasan). Remote Sensing and Geographic Information Systems in Natural Resources, 6(2), 75-88.
[4]          Clark, W. A., & Hosking, P. L. (1986). Statistical methods for geographers. Wiley, New York.
[5]          Coppin, P., & Jonckheere, I. (2004). Digital change detection methods in ecosystem monitoring. International Journal of Remote Sensing, 25(9), 1565-1596.
[6]          Darvish Sefat, A., & Pirbabakvar, M. (2010). Remote sensing for GIS managers. Compilation: Aronof, Stan. Tehran Publishing House, first edition, 432 p.
[7]          Eastman, J. R. (2006). IDRISI Andes tutorial. Clark Labs, Worcester, MA.
[8]          Faraj Allah, A., et al. (2011). Monitoring and predicting the spatial and temporal changes in land use/land cover. Remote Sensing and Geographic Information System in Natural Resources, 6(4), 6-14.
[9]          Gessler, P., Chadwick, O., Chamran, F., Althouse, L., & Holmes, K. (2000). Modeling soil-landscape and ecosystem properties using terrain attributes. Soil Science, 6, 2046-2056.
[10]      Great, A. (2006). Modeling of rainfall and runoff in Kachik watershed, Golestan province. MSc. Thesis, Faculty of Grassland and Watershed Management, Gorgan University of Agricultural Sciences and Natural Resources, 41 p.
[11]      Huang, Q. H., Cai, Y. L., & Peng, J. (2007). Modeling the spatial pattern of farmland using GIS and multiple logistic regression: A case study of Maotiao River Basin, Guizhou Province, China. Environmental Modeling & Assessment, 1, 55-61.
[12]      Lee, S. (2004). Application of likelihood ratio and logistic regression models to landslide susceptibility mapping using GIS. Environmental Management, 34(2), 223-232.
[13]      Mahiny, A., & Clarke, C. K. (2012). Guiding SLEUTH land-use/land-cover change modeling using multicriteria evaluation: Towards dynamic sustainable land-use planning. Environment and Planning B: Planning and Design, 39, 925-944.
[14]      Mahmoudi, M., & Ismail, C. (2010). Investigating the possibility of modeling forest cover destruction in the Northern Zagros forests using aerial photos and GIS (Case study of Sardasht forests). MSc. Thesis, Gorgan University of Agricultural Sciences and Natural Resources, pp. 14-19.
[15]      Mas, J. F., Puig, H., Palacio, J. L., & Sosa-Lopel, A. (2004). Modeling deforestation using GIS and artificial neural networks. Environmental Modeling & Software, 19(5), 461-471.
[16]      Masrori, E. (2008). Spatial modeling of forest cover extension in Khorasar Khorasabad region using GIS. PhD Dissertation, Gorgan University of Agricultural Sciences and Natural Resources, pp. 65-81.
[17]      Miranda, E. J., Treviño, M., Pompa, C., & Aguirre, A. (2012). Modeling susceptibility to deforestation of remaining ecosystems in North Central Mexico with logistic regression. Journal of Forestry Research, 23, 345-354.
[18]      Mirzai, V., & Niknejad, M. (2010). Identifying the effective factors on reducing forest coverage using landsat satellite images (Case study: Beaver Forest Area- Maleshahi County). Quarterly Journal of Conservation and Exploitation of Natural Resources, 1(2), 91-97.
[19]      Oliad Qadikelyai, J. (2004). Remote sensing satellite images. Compiled by: Thomas, Lilsand and Ralph, Keeper, Mazandaran University Press, 835 p.
[20]      Pontius, R. G., & Schneider, L. (2001). Land-use change model validation by a ROC method for the Ipswich watershed, Massachusetts, USA. Agriculture, Ecosystems & Environment, 85(3), 239-248.
[21]      Ranjbar, A. (2002). Assessment and evaluation of forest degradation process using GIS and remote sensing data. MSc. Thesis, K. N. T. University of Technology, 63 p.

  • تاریخ دریافت 26 اسفند 1401
  • تاریخ بازنگری 02 دی 1402
  • تاریخ پذیرش 25 دی 1402