A Statistical Assessment of Land Use and Land Cover Transitions of the Urmia Lake Basin Using Transition Matrix

جغرافیاوتوسعه ناحیه ای

Article 5, Volume 15, Issue 1 - Serial Number 28, September 2017, Pages 171-192   

Document Type: scientific-research article

DOI: 10.22067/geography.v15i1.56877

Authors

yaser amini^{* 1}; Asadalah Khoorani²; Masoud Bakhtyarikya²; Saleh Arekhi³

¹hurmozgan

²University of Hormozgan

³Golestan University

Abstract

Extended Abstract
1- Introduction
Monitoring land use changes and understanding its dynamics in a watershed is a special issue in the sustainable management of watershed basins. In recent decades, rapid changes in land use and land cover in Urmia Lake basin is accompanied by important consequences such as the destruction of natural resources, environmental pollution, and rapid growth of cities. Understanding the ratio of changes as well as the systematic and random transitions of Land Use and Land Cover (LULC) over time can be used to determine the extent of degradation, manage these changes in a proper direction, predict future changes, and act properly. Detecting and modeling land use changes by using remotely sent data and GIS technology can provide a solid understanding of changes in land use and therefor can offer suitable management practices. The aim of the present study is the use of remote sensing and GIS for mapping land use and land cover changes and identification of their transitions using a transfer matrix and Landsat images in Urmia Lake basin. Therefor, the performance of pixel-based and object-oriented methods of land use and land cover classification specified in Urmia Lake basin is going to be compared. Also the spatial and temporal dynamics of LULC is going to be analysed for this basin. Another objective of this research is to identify the ratio of vulnerabilities of each land use relative to other land uses.
2- Theoretical Framework
Land-cover transitions can be classified into random and systematic changes. Random transitions are those influenced by coincidental or unique processes of change; for instance, the rapid and abrupt changes of land transformations in an ecosystem depending on resilience and feedback mechanisms. On the other hand, Systematic transitions are those due to regular or common processes of change. (Tucker et al., 1991; Lambin et al., 2003).
3- Methodology
In this study, Landsat TM and ETM + images were used for the period of 1988 to 2015. First, some image pre-processing techniques were done including reducing the brightness of water, strips of pixels in images, and removing the cloud spots. The area of Urmia Lake basin is 5786 Km2 and contains eight Landsat image frames. So, for each period, 8 images were obtained and mosaicked based on band to band method and the coordinate of the region. After running the required pre-processing on the images, training samples were obtained and the images were classified based on vector machine and object-based classification methods. Then, the results of the classification were validated. Based on field observations and vegetation map of Urmia lake basin, the training samples were obtained from 6 land uses including the residential area, forests, agricultural fields, rangelands, water resources, and bare lands. Bias, Gama, Kernel functions, and pyramid levels and Penalty Pyramid Parameters were obtained from a cyclic kernel function for Support Vector Machine. Also, effective parameters of object-oriented approach were obtained including the window width, weighted mean factor, weighted variance factor with error, and similarity tolerance.
In order to evaluate the results, Kappa coefficient and overall accuracy were used. For estimating the rate of transmission and other characteristics of the watershed of Urmia Lake, the transition matrix of object-oriented classification method was extracted for 1988-2015. After that, by using appropriate formula, the rate of gain, loss, persistence, net change, and swap (simultaneous exchange) was calculated for each land use/land cover. The swap represented changes in the location between land covers, whereas the net change was associated with a measurable irreversible change in the surface of one land cover to another; having these two components of change allowed the actual spatial dynamics of LULC change to be determined in the study area. In this manner, it was possible to determine the total change in LULC between 1988 and 2015 and highlight the land cover types that exhibited the greatest variation.
4- Results & Discussion
The overall accuracy of Support Vector Machine (SVM) and object-oriented approach were 94 and 92 respectively. Also, Kappa coefficient was 92 and 89 respectively, showing that although both methods show acceptable results, the object-oriented approach is stronger. The results of the transactions showed that coverage has been persisted in the 59 percent of the land in the catchment area of Lake Urmia during the period of 1988 to 2015, most of which was related to the residential areas. About 14 percent of the basin has experienced swap. Water resources has experienced the most loss and the less swap.
For residential areas, agricultural fields, and water resources classes, the ratio of gain to persistence is more than 1 indicating that the amount of gains are due to the persistence of these classes. The ratio of net changes to persistence was negative for the classes of forest, rangelands, and water resources. The net reduction of water resources was almost contrary to the persistence of this landscape. Also, the net reduction of agricultural fields was almost the same but half the net gain.
5- Conclusions & Suggestions
Urmia Lake basin has experienced rapid changes and transitions during 1988-2015; that is, only 59 percent of the land uses have been stable while the other areas have experienced a kind of transition. Due to the reduction of water resources and rangelands, bare lands and agricultural fields have been increased. The results should be noticed for an integrated watershed management of the basin.

Keywords

Object-oriented approach; Transfer matrix; Vulnerability and persistence; Urmia Lake Basin

References

1. آرخی، ص. (1394). آشکارسازی تغییرات پوشش/کاربری اراضی با پردازش شیءگرای تصاویر ماهواره‌ای با استفاده از نرم‌افزار Idrisi selvi (مطالعۀ موردی: منطقۀ آبدانان). فصلنامۀ علمی پژوهشی اطّلاعات جغرافیایی «سپهر»، 24(95)، 51-62.‎

2. آرخی، ص.؛ اصفهانی، م. (1393). آموزش تصویری نرم‌افزار ایدریسی سلوا. چاپ اوّل. انتشارات دانشگاه گلستان.

3. زهتابیان، غ.؛ طباطبایی، غ. (1387). بررسی روند بیابان-زائی در استان خوزستان با استفاده از پردازش تصاویر ماهواره‌ای. همایش منطقه‌ای توسعۀ پایدار در زیست‌بوم‌های بیابانی، یزد. 57ـ70.

4. علوی‌پناه، س. (1392). کاربرد سنجش از دور در علوم زمین(علوم خاک). چاپ چهارم. انتشارات دانشگاه تهران.

5. علوی‌پناه، ک.؛ متین‌فر، ح. ر.؛ و سرمدیان، ف. (1383). ارزیابی کاربری داده‌های ماهواره‌ای ازنظر صرفه‌جویی وقت. کنفرانس ملّی بهره‌وری، فرهنگستان علوم ایران، 425ـ439.

6. فاطمی نصیرآبادی، س. (1385). آموزش نرم‌افزار ENVI (نرم‌افزار پردازش تصاویر ماهواره‌ای). چاپ اوّل. تهران: سازمان نقشه‌برداری کشور.

7. Alo, C. A., & Pontius, R. G. (2008). Identifying systematic land-cover transitions using remote sensing and GIS: The fate of forests inside and outside the protected areas of southwestern Ghana. Environment and Planning B: Planning and Design, 35(2), 280-295.

8. Bauer, T., & Steinnocher, K. (2001). Per-parcel land use classification in urban areas applying a rule-based technique. GeoBIT/GIS, 6, 24-27.

9. Braimoh, A. K. (2006). Random and systematic land-cover transitions in northern Ghana. Agriculture, Ecosystems & Environment, 113(1), 254-263.

10. Briones, P. S., & Sepúlveda-Varas, A. (2017). Systematic transitions in land use and land cover in a pre-Andean subwatershed with high human intervention in the Araucania region, Chile. Ciencia e Investigacion Agraria, 43(3), 396-407.‏

11. Definiens the image intelligence company. (2006). Definiens professional user guide. Retrieved from http://www.definiens.com/Userguide.pdf

12. Hussain, M., Chen, D., Cheng, A., Wei, H., & Stanley, D. (2013). Change detection from remotely sensed images: From pixel-based to object-based approaches. ISPRS Journal of Photogrammetry and Remote Sensing, 80, 91-106.

13. Karami, A., Khoorani, A., Nohegar, A., Shamsi, S. R. F., & Moosavi, V. (2015). Gully erosion mapping using object-based and pixel-based image classification methods. Environmental & Engineering Geoscience, 21(2), 101-110.

14. Lambin, E. F., Geist, H. J., & Lepers, E. (2003). Dynamics of land-use and land-cover change in tropical regions. Annual Review of Environment and Resources, 28(1), 205-241.

15. Lu, D., Mausel, P., Brondizio, E., & Moran, E. (2004). Change detection techniques. International Journal of Remote Sensing, 25(12), 2365-2401.

16. Mackie, R. I. (2013). Dynamic analysis of structures on multicore computers: Achieving efficiency through object-oriented design. Advances in Engineering Software, 66, 3-9.

17. Manandhar, R., Odeh, I. O., & Pontius, R. G. (2010). Analysis of twenty years of categorical land transitions in the Lower Hunter of New South Wales, Australia. Agriculture, Ecosystems & Environment, 135(4), 336-346.‏

18. Matinfar, H. R., Sarmadian, F., Alavipanah, S. K., & Heck, R., (2008). Characterizing land use/land cover types by Landsat7 data based on object-oriented approach in Kashan region. Iranian Journal of Range and Desert Research, 14(4), 589-602.

19. Petropoulos, G. P, Kalaitzidis, C., Vadrevu, K. P. (2012). Support vector machines and object-based classification for obtaining land use/cover cartography from Hyperion hyperspectral imagery. Computers & Geosciences, 41, 99-107.

20. Pontius, R. G., Shusas, E., & McEachern, M. (2004). Detecting important categorical land changes while accounting for persistence. Agriculture, Ecosystems & Environment, 101(2), 251-268.

21. Puissant, A., Rougier, S., & Stumpf, A. (2014). Object-oriented mapping of urban trees using random forest classifiers. International Journal of Applied Earth Observation and Geoinformation, 26, 235-245.‏

22. Puissant, A., Rougier, S., & Stumpf, A. (2014). Object-oriented mapping of urban trees using Random Forest classifiers. International Journal of Applied Earth Observation and Geoinformation, 26, 235-245.

23. Roostaei, S., Alavi, S., Nikjoo, M., & Kamran, K. V. (2012). Evaluation of object-oriented and pixel-based classification methods for extracting changes in urban areas. International Journal of Geomatics and Geosciences, 2(3), 738-749.

24. Singh, A. (1989). Digital change detection techniques using remotely-sensed data. International Journal of Remote Sensing, 10(6), 989-1003.‏

25. Tso, B., & Mather, P. M. (2009). Classification methods for remotely-sensed data. New York: Taylor and Francis.

26. Tucker, C. J., Dregne, H. E., & Newcomb, W. W. (1991). Expansion and contraction of the Sahara Desert. Science, 253, 299-301.