Land Use/Land Cover Change Modeling Using Marcov Chain and Cellular Automata (Case Study: Dezful, Iran)

Document Type : Research Article

Authors

• GholamReza Sabzghabaei ¹
• Farhad Salehipour ²
• Soolmaz Dashti ³
• Arezoo Safavian ⁴

¹ University of Technology, Behbahan

² Behbahan Khatam Alanbia University of Technology, Behbahan, Iran.

³ Department of Environment, Ahvaz Branch, Islamic Azad University, Ahvaz, Iran.

⁴ Gorgan University of Agricultural Sciences and Natural Resources, Gorgan, Iran

10.22067/geo.v7i2.64775

Abstract

Introduction

Modeling of land use change is largely a problem. Nevertheless, theoretical and empirical methods have increased in this framework, and studies are carried out on land use change in these areas. Few researchers are able to use models to predict land use variations (Iacono, Levinson, El-Geneidy, Wasfi., 2015).Generally, land use change is a deformation of a piece of land. This change is the foundation of the desired goals, and it is not necessary to have only a change in the coverage of the land. This change may be due to intensity change and management (Verburg, Chen, Soepboer, Veldkamp., 2000). Land use and land cover change is greatly worrying as it reduces biodiversity and impacts human lives as such (Geoghegan, Klepeis, Mendoza, Yelena, Chowdhury, Turner, Vance., 2001). There are a variety of methods for modeling land cover and land use change variations, which can be used to model mathematical equations, system model, statistical model, evolution model, cellular model, and hybrid model. The cellular model includes cellular automatic models (Parker, Manson, Janssen, Hoffmann, Deadman., 2002).An integrated model of automatic and chain cells (CA-Markov) is a combination of the Markov chain model (from the category of empirical estimation models) and the model of automatic cells (from the category of dynamic simulation models). In fact, by adding a spatial contiguity feature, this model simulates land use through a random Markov chain model for the next years. In recent years, due to the easy access to satellite imagery and the capabilities of the GIS, modeling of land cover change and its prediction for the future has been carried out, and many studies have been done in this field (Zare Garizi, Sheikh, Sadodin, Salan Mahini., 2012). Dezful is among the cities of Khuzestan Province, which suffered a lot during the eight years of the 'Holy Defense' in the Iraq war against Iran. The purpose of this research is to investigate the land use change in pre-war and post-war situations and predict the changes for the forthcoming years.

Materials and Methods

In this study the Landsat 5 satellite images was used by Path No.166 and Row No.37 in 1985 before the war. The ETM+ Landsat 7 satellite images was used by Path No. 166 and Row No.38 in 2005 after the war. Satellite data was prepared in GeoTIFF format in seven spectral bands, which were not used from band 6 because of the low spatial resolution (60 m) and the failure to address the thermal properties of the phenomena in this study. In this study land use and land cover maps was used from Landsat images, and simulated land use change by using Marcov chain and cellular automatato in Dezful city. After cutting the satellite images according to the boundaries of the study area, the images were classified to determine the existing use around the region. Of course, by evaluating and analyzing the characteristics of the spectral reflection of different wavelengths, it was possible to separate the different classes of use. Using the IDRISI 15 software, classifying the images was used by using the controlled classification method and maximum likehood algorithm for the two years. This method, using the mean and matrix of covariance of educational data, provides a better analysis than other methods of classification, such as the classification of the smallest distance from image data(Richards & Xiuping, 2006). Finally, four classes of classification were identified, including the city and industrial facilities, agricultural lands, rivers and ranges for all images. Then, matrix of land use and land cover changes probability was calculated in IDRISI 15 software and the Markov method. In the matrix, the rows represent the older layers/cover of the older lands and the columns represent the newer enclosure classes. An automatic method was used to locate these changes and the map of the probability of the conversion of each land use was achieved for 2018-2030.

Results and Discussion

The results of the classification of the images using the error matrix indicate that the accuracy of the overlapping of the classification for 1985 and 2005 was 99.6% and 99.5%, respectively, and the Kappa coefficient for both years was 99.99, which is acceptable at an appropriate level. The results of the classification of the images showed that in 1985, the total surface area of the region was about 1475.7 hectares covered by urban areas, while the non-urban lands constituted an area of 10555 hectares. In 2005, the area of urban areas was about 218.29 hectares, while the non-urban land was 12.91 million hectares. The results of the comparison of the two maps of the beginning and the end of the given period showed that during this period, 6172,200 hectares (about 87.8%) were added to urban areas and 1040.58 hectares (about 4.61%) were added to agricultural lands, which indicates the expansion of the region in urban and agricultural development during the postwar period. The probability of converting each of the cabrivals to another user in the years 2018 and 2030 was derived using the transmission probability matrix via the automatic network method. In the next step, the probability transformation matrix was obtained using Markov chain analysis based on observed changes in 2018-2030.By detecting land use and land cover changes, we can determine growth and destruction natural resources and mange them. Markov generally processes the future status of a system based on pre-model status. Coverage and usage changes from one time to another. The use of Markov analysis is the basis for estimating the likelihood of future system changes. In this type of analysis, a matrix of the probability of land use change is created from time t of time t+1. Based on the maps of land cover/land use obtained from satellite imagery (1985 and 2005) and the Markov model, the changes in the coverage/land use and urban growth rate were successfully simulated in the city of Dezful. In this paper, the analysis and prediction of urban growth and land use change were considered using remote sensing data and Markov chain analysis. Accordingly, in the years 2018 and 2030 urban lands will increase with an increase of 130.71 and 334 hectares respectively, and agriculture will be reduced to 320.22 and 273.24 hectares, respectively. In the period of 20 years, the residential lands increased and the agricultural land decreased. Simulation of land use change and land cover for the years 2018 and 2030 indicates a decrease in agricultural areas and an increase in urban areas. Dezful, like other cities of Khuzestan Province, suffered a lot during the eight years of the Holy Defense, but after that has started to grown and improved.

Conclusions

Based on the results of land use planning and landslide modeling in Dezful, it can be confirmed that agricultural, rangeland and river utilization will be likely to decrease in the forthcoming years and land use will be further enhanced. Considering the growing human population and the urgent need to exploit the resources of these changes, But these changes must be taken into account in order to ensure that resource utilization is carried out in a systematic manner near future, and that it will not lead to the destruction of resources. Thus, with the development of sustainable development policies, the growing trend of urban areas will not reduce agricultural land and other limited lands in the future and transform them into lands built. Simulated user-mapped maps can be used as a system to alert the actions taken and their possible future effects in one area for other locations. Of course, this simulation is based on the current situation and the continuation of the future changes that may affect a series of changes in policies in a cross-sectional and specific manner, such as government policies in Mehr Housing and other cases. Using time-domain remote sensing data to identify changes and see how it was used in the past, and simulation of land cover and land use in the future can help planners and designers manage, plan, and apply developmental policies. In fact, the simulation and modeling of urban growth for the future is considered as the future management and planning requirements of the city. Since land use maps are of particular importance as the basic information for different planning, the results of this research can be used in the future planning of the region concerning its land use change.
 

Keywords

• Urban growth
• simulation
• Markov chain
• Remote sensing
• Dezful

---

References

خاکپور، براتعلی؛ ولایتی، سعدالله؛ کیانژاد، قاسم؛ 1386. الگوی تغییر کاربری اراضی شهر بابل طی سال‌های 1362- 1378. مجله جغرافیا و توسعه ناحیه‌ای، شماره 9، 64-45.

زارع گاریزی، آرش؛ ‌شیخ، عبدالواحدبردی؛ سعدالدین، امیر؛ سلمان‌ماهینی، رسول؛ 1391. شبیه‌سازی مکانی- زمانی تغییرات گستره جنگل در آبخیز چهل‌چای استان گلستان با استفاده از مدل تلفیقی سلول‌های خودکار و زنجیره مارکوف. فصلنامة علمی پژوهشی تحقیقات جنگل و صنوبر ایران. شماره 2، 273- 285.

شکوئی، حسین؛ 1373. دیدگاه‌های نو در جغرافیای شهری. چاپ تهران. موسسه انتشارات سمت. صدرموسوی، میرستار؛ یزدانی‌‌چهاربرج، رسول. 1392. ارزیابی و شبیه‌سازی رشد شهری میانه اندام با استفاده از مدل CA-Markov (نمونه موردی: شهر مراغه)، همایش ملی معماری پایدار و توسعه شهری، یزد.

قربانی، رسول؛ پورمحمدی، محمدرضا؛ محمودزاده، حسن. 1392. رویکرد زیست‌محیطی در مدل‌سازی تغییرات کاربری اراضی محدوده کلانشهر تبریز با استفاده از تصاویر ماهواره‌ای چند زمانه‌ای، ارزیابی چند معیاری و سلول‌های خودکار مارکوف. فصلنامه علمی پژوهشی مطالعات شهری. شماره 8، 30- 23.

کامیاب، حمیدرضا؛ سلمان‌ماهینی، رسول؛ حسینی، سیدمحسن؛ غلامعلی‌فرد، مهدی؛ 1389. اتخاذ رهیافت اطلاعات محور با کاربرد روش رگرسیون لجستیک برای مدل‌سازی توسعه شهری گرگان، فصلنامه محیط‌‌شناسی. شماره 54، 96-89.

Anderson, W.P., Kanaroglou, P.S., Miller, E.J., 1996. Urban form energy and the environment, a review of issues, evidence and policy. Urban Studies 33(1), 7-35.

Arsanjania, J., Helbich, M., Kainz, K., Darvishi Boloorani, A., 2012. Integration of logistic regression, Markov chain and cellular automata models to simulate urban expansion. International Journal of Applied Earth Observation and Geo information 21, 265-275.

Barredo, I., Demicheli, L., 2003. Urban sustainability in developing countries megacities: modelling and predicting future urban growth in Lagos. Journal of Urban sustainability in developing countries’ megacities 5, 297-310.

Dongjie, G., HaiFeng, L., Takuro, I., Weici, S., Tadashi, N., Kazunori, Kazunori., 2011. Modeling urban land use change by the integration of cellular automaton and Markov model. Ecological Modeling 222 (20), 3761–3772.

Fan, F., Wang, Q., Wang, Y., 2007. land use and land cover change in Guangzhou, Chaina, from 1998 to 2003. Based on landsat TM/ETM+ imagery, Sensors, 7, 1323-1342.

Geoghegan, J. Klepeis, S., Mendoza, P.M., Yelena, O.R., Chowdhury, R. Turner, B.L. and Vance C., 2001. Modeling tropical deforestation in the southern Yucatan Peninsular region: comparing survey and satellite data Agriculture. Ecosystems and Environment 85, 25-46.

Healey, S. P., Cohen, W. B., Zhiqiang, Y., Krankina, O. N., 2005. Comparison of Tasseled Cap-based Landsat data structures for use in forest disturbance detection. Remote Sensing of Environment, 97, 301-310.

Iacono, M., Levinson, D., El-Geneidy, A., 2012. A Markov Chain Model of Land Use Change in the Twin Cities, University of Minnesota: Nexus Research Group. Working Papers 107.

Islam, Sh., Ahmed, R., 2011. Land use change prediction in Dhaka City using GIS aided markov chain modeling. J. Life Earth Sci, 6, 81-89.

Parker, D.C., Manson, S.M., Janssen, M.A., Hoffmann, M.J., Deadman, P., 2002. Multi agent systems for the simulation of land use and land cover change: A Review, 43.

Richards John A., Xiuping Jia., 2006. Remote Sensing Digital-Image Analysis. Electronics & Electrical Engineering, 439.

Rogan, J., Franklin, J., Roberts, D. A., 2002. A comparison of methods for monitoring multi-temporal vegetation change using Thematic Mapper imagery. Remote Sens. Environ, 80, 143−156.

Seto, K. C., Woodcock, C. E., Song, C., Huang, C., Lu, J., Kaufman, R.K., 2002. Monitoring, Land-use change in the Pearl River delta using Landsat TM. Int. J Remote Sens, 23, 1985-2004.

Subedi, P., Subedi, K., Thapa, K., 2013. Application of a Hybrid Cellular Automaton Markov (CA-Markov) Model in Land-Use Change Prediction: A Case Study of Saddle Creek Drainage Basin, Florida. Applied Ecology and Environmental Sciences, 6, 126-132.

Verburg, P.H., Chen, Y., Soepboer, W., and Veldkamp, A., 2000. GIS based modeling of human-environment interactions for natural resource management. Applications in Asia. Proc. 4th Int. Conf. Integrat. GIS Environ. Modeling (GIS/EM4): Problems, Prospects and Research Needs, Banff, Alberta, Canada, Sept 8, 1-18.

Wu, Q., Li, H.Q., Wang, R.S., Paulussen, J., He, Y., Wang, M., Wang, B.H., Wang, Z., 2006. Monitoring and predicting land use change in Bijing using remote sensing and GIS. Landscape and urban planning, 78, 322-333.

Xian, G., Crane, M., 2005. Assessments of urban growth in the Tampa Bay watershed using remote sensing data. Journal of Remote Sensing of Environment, 97, 203-215.