Analyzing the Spatial-temporal Changes of Landforms and Land-Use in the Desertification Process of Yazd-Ardakan plain Using Maximum Likelihood Algorithm

جغرافیا و مخاطرات محیطی

Article 2, Volume 7, Issue 1 - Serial Number 25, January 2018, Pages 17-36   

Document Type: Research Article

DOI: 10.22067/geo.v7i1.66078

Authors

Amir Karam^{* 1}; Monireh Rayati Shavvazi¹; Hamid Reza Ghafarian Malamiri²; Adel Sepehr³

¹University of Kharazmi, Tehran

²University of Yazd

³Ferdowsi University of Mashhad

Abstract

1. Introduction
Desertification is associated with the process of destroying and demolishing the natural ecosystems in ultra-arid, arid, semi-humid arid regions, leading to reduced biomass production and the appearance of soil degradation or erosion effects. More than 80% of Iran is located in ultra-arid, arid, semi-humid arid regions of the country with fragile ecological conditions and are influenced by desertification phenomenon.
In order to study the desertification and to analyze the trend of its changes, a wide range of spatial and temporal information such as geomorphology and land-use are needed. In fact, analyzing their changes is the process of identifying differences in the status of an object or a phenomenon by observing it at different times, which provides a basis for better understanding of human relationships and interactions with natural phenomena for managing and making better use of resources. In environmental management, the most common contributions of geomorphologists are to provide a map of landscapes with emphasis on its selected features, to understand the nature and causes of landscape changes, and to organize them according to trend predictions during the formation processes and the transformation of landforms. Human's specific use of territory as land use is also an important consideration in land tenure assessment. These applications change at time intervals. In arid and semi-arid regions, these changes usually lead to increased desertification.
The identification of deserts' landforms and land-use, their classification and mapping at various time intervals are considered as a method for change detection in these areas. Today, using remote sensing (RS) instead of the visual interpretation to identify landforms and land use and their change detection is a necessity. One of the steps in identification of change detection is the classification of remote sensing data. There are various methods for the classification of satellite images. These methods can be used for separation of different geomorphological landforms and land-uses and their variations. Each method has its own advantages and limitations. The choice of the research method in geomorphology depends on the purpose of the study and the available data. Today, the use of satellite data and remote sensing techniques are considered as a modern method in geomorphological studies.
The most common classification methods can be referred to as the maximum likelihood classification. Other classification methods such as the minimum distance, Mahalanobis distance, and neural network have attracted much attention. The main objective of this study is to investigate and analyze the temporal variations of geomorphologic landforms and land-use in desertification part of Yazd-Ardakan Plain. For this purpose, time series satellite observations of Landsat sensor were used from 1986 to 2016 (over 30 years).
2. Materials and Methods
In this research, a part of the Ardakaan-Yazd basin is considered for desertification analysis due to the issues and problems related to land destruction and the critical focus of wind erosion. This district, according to the latest political divisions, includes central zone and Khezerabad zone of Sadoogh county in Yazd. The area of the study area is 1563.11 km2.
The research method in this study is analytical survey. To study the changes of desert's landforms and land uses, two satellite imageries of TM and OLI of Landsat satellite of 1987 and 2016 were used. First, the radiometric and atmospheric corrections was performed using Flaash algorithm, and then the geomorphological landforms and land use were introduced and the training samples were selected by field observations, topography and geomorphology maps, and Google Earth images. To classify the landforms and land use, the supervised classification of maximum likelihood algorithm was used. Then, the accuracy of classified maps was evaluated using the overall accuracy and the Kappa coefficient metrics. Finally, to evaluate the changes of landforms and land-uses, the post classification method was used. To analyze the database, ENVI 5.3, ArcGIS 10.4.1 and Excel 2013 software were used.
3. Results and Discussion
In the present study, 15 landforms were identified in the study area with regard to field observations, available resources and geomorphologic map of Yazd-Ardakan plain.
Totally, 15 landforms were identified in the study area, including Alluvial Fan, Glacis Pediment Plain, Clay Pan, Glacis Dennoyage Plain, Inselberg, Glacis Epandage Plain, Kalut, Erg (barchans, longitudinal dunes, barchanoid), hills, mountains, Salt Dome, Sebkha and Sand Sheet. In addition, 8 land-use classes were identified in the study area. After the image correction, the geomorphological and land use maps were prepared. Finally, to investigate the nature of the changes, the comparison method of post classification was used, which was applied on maximum likelihood algorithm. Then, the changes of landforms were calculated in terms of its area and percentage. The results showed that the dominant class was the Glacis Epandage Plain in the both satellite images of 1987 and 2016. Then, hills with 17/58 percent of the total area are ranked as the highest area in 1987, while this class had downward trend with 11/58 percent in 2016. In 1987, Barkhan class had the lowest area with 0/17 percent, whereas this class had downward trend with 0/11 percent in 2016.
Changes in land use classes over a 30-year period in terms of area and percentage indicate that the predominant class in 1987 was the low-density rangelands of the range of 25-5%, which, remains the prevailing class in this year even with the decrease in the area of this class in 2016.
The saline lands with 22312.63 hectares of the entire region are of the next order of the most area in 1987 but have the decreasing process to 1.14% in 2016. In 1987, the urban class occupied 0.46% of study area with a growth rate of 2.33% by 2016. The construction of forested trees for deforestation also shows a growth rate of 15.36% over 30 years.
4. Conclusions
The results of the landforms and land use classification showed that the maximum likelihood algorithm offers a detailed classification method. The area and the percentage of landforms and land use changes over 30 years showed that landforms such as Barchan, Clay Pan, Longitudinal Dunes, Barchanoid and Kalut had a downward trend, because they were located in the context of the development of the city. The results showed that Sabkha area has followed a downward trend over 30 years. The comparison of our results with the results of previous studies showed that the increase of wells in Yazd-Ardakan plain has helped cultivate the large areas of saline lands. Therefore, the natural and human factors were involved in changing the desert landform and land use in the study area.
It should be noted that this study shows that the separation and analysis of the landform variations, such as the Salt Dome and Inselberg created in long time intervals, cannot be easily accomplished using the maximum likelihood algorithm. Because they were of ascendant trend during 30 years according to the outcomes, and this is because of similarity of their adjacent spectral reflections in which case the pixels of these landforms have been replaced by more pixels. However, according to the results, separating and analyzing the changes of the sandy hills is possible with the maximum similarity algorithm method during 30 years (from 1986 to 2016). In general, the results of this research were in the line of the previous studies that confirm above statement.

Keywords

Spatial-temporal changes; Landform; land use; desertification; Maximum Likelihood; Yazd-Ardakan Plain

References

اختصاصی، محمدرضا؛ احمدی، حسن؛ باغستانی، ناصر؛ خلیلی، علی؛ فیض نیا، سادات؛ 1375. منشأیابی تپه‌های ماسه‌ای در حوزه دشت یزد-اردکان. انتشارات موسسه تحقیقات جنگل‌ها و مراتع، تهران.

اختصاصی، محمدرضا؛ سپهر، عادل؛ 1390. روش‌ها و مدل‌های ارزیابی و تهیه نقشه بیابان‌زایی، انتشارات دانشگاه یزد.

آرخی، صالح؛ 1393. تهیه نقشه کاربری اراضی دشت عباس ایلام با استفاده از روش‌های شبکه عصبی مصنوعی. ماشین بردار پشتیبان و حداکثر احتمال. نشریه مرتع داری. شماره 2، 30-43.

آرخی، صالح؛ 1394. آشکارسازی تغییرات پوشش-کاربری اراضی با پردازش شیءگرای تصاویر ماهواره‌ای مواد و روش‌ها با استفاده از نرم افزار Idrisi selvi (مطالعه موردی: منطقه آبدانان). فصلنامه اطلاعات جغرافیایی (سپهر). شماره 95، 51-62.

اکبری، الهه؛ زنگنه اسدی، محمدعلی؛ تقوی مقدم، ابراهیم؛ 1395. پایش تغییرات کاربری اراضی با استفاده از روش‌های مختلف تئوری آموزش آماری منطقه نیشابور. مجله آمایش جغرافیایی فضا، شماره 20، 35-49.

جعفری، مهدی؛ زهتابیان، غلامرضا؛ احسانی، امیرهوشنگ؛ 1392. بررسی تأثیر باند حرارتی و الگوریتم‌های طبقه‌بندی نظارت شده داده‌های ماهواره‌ای در تهیه نقشه‌های کاربری اراضی کاشان (مطالعه موردی: کاشان). فصلنامه تحقیقات مرتع و بیابان و ایران، شماره 1، 72-87.

خلیلی، علی؛ 1371. اقالیم خشک و فراخشـک. چاپ در مجموعـه مقـالات سـمینار بررسی مسائل مناطق بیابانی و کویری ایران - یـزد، مرکـز تحقیقـات کـویری و بیابانی ایران وابسته به دانشگاه تهران. صص 14-32.

ربیعی، حمیدرضا؛ ضیائیان، پرویز؛ محمدی، عباسعلی؛ 1384. کشف و بازیابی تغییرات کاربری و پوشش اراضی شهر اصفهان به کمک سنجش‌ازدور وسیستم اطلاعات جغرافیایی. فصلنامه مدرس علوم انسانی، شماره 4، 19-32.

سفیانیان، علیرضا؛ مدنیان، ملیحه السادات؛ 1390. مقایسه روش‌های طبقه‌بندی کننده حداکثر مشابهت و حداقل فاصله از میانگین در تهیه نقشه پوشش اراضی (مطالعه موردی: استان اصفهان). مجله علوم و فنون کشاورزی و منابع طبیعی، علوم آب و خاک، شماره 57، 253-264.

شایان، سیاوش؛ یمانی، مجتبی؛ فرج زاده، منوچهر؛ احمدآبادی، علی؛ 1391. طبقه‌بندی نظارت شده لندفرم‌های ژئومورفولوژیکی مناطق خشک با استفاده از پارامترهای ژئومورفومتریک، مطالعه موردی: منطقه مرنجاب. سنجش‌ازدور و GIS ایران. سال چهارم، شماره 2، 19-28.

صفری، عبدالرضا؛ آوریده، حمیدرضا؛ همایونی، سعید؛ خزایی، صفا؛ 1393. برآورد عمق آب‌های ساحلی به کمک تصاویر سنجش‌ازدور فراطیفی. نقشه برداری و اطلاعات مکانی، شماره 1، 1-5.

عطاپورفرد، عباس؛ حسینی، مجید؛ خسروشاهی، محمد؛ 1383. شناسایی و تفکیک مناطق بیابانی از دیدگاه ژئومورفولوژیکی، مطالعه موردی: استان تهران. فصلنامه پژوهشی تحقیقات مرتع و بیابان ایران، شماره 3. 275-286.

علوی پناه، ک؛ 1382. کاربرد سنجش‌ازدور در علوم زمین (علوم خاک). انتشارات دانشگاه تهران.

فاطمی، سید باقر؛ رضایی، یوسف؛ 1384. مبانی سنجش‌ازدور. انتشارات آزاده.

قنواتی، عزت اله؛ ضیائیان فیروزآبادی، پرویز؛ علوی نژاد، سیدنظام؛ 1386. آشکارسازی تغییرات ژئومورفولوژیک و کاربری با استفاده از GIS و RS اراضی خورموسی. نشریه زمین شناسی مهندسی، شماره ٢، 413-430.

کرینسلی، دانیل؛ 1350. مطالعه ژئومورفولوژی و آب و هوای گذشته (پلایای ایران)، سازمان جغرافیای کشور.

گیویی اشرف، زهرا؛ سر کارگر اردکانی، علی؛ 1390. پایش کاربری اراضی با استفاده از سنجش‌ازدور به‌منظور ارزیابی بیابان‌زایی، مطالعه موردی: دشت مروست، استان یزد. چاپ در مجموعه مقالات همایش ملی ژئوماتیک. صص 1-7.

مرادی، حمیدرضا؛ فاضل پور، محمدرضا؛ صادقی، حمیدرضا؛ حسینی، زین العابدین؛ 1387. بررسی تغییر کاربری اراضی در بیابان‌زایی محدوده شهر اردکان با استفاده از سنجش‌ازدور. فصلنامة تحقیقات مرتع و بیابان ایران. شماره 1، 1-12.

مقیمی، آرمین؛ عبادی، حمید؛ صادقی، وحید؛ 1395. مروری بر روش‌های شناسایی تغییرات از تصاویر ماهواره‌ای چند زمانه با رویکرد پیکسل پایه و شئ پایه. نشریه علمی-ترویجی مهندسی نقشه برداری و اطلاعات مکانی، شماره 2، 99-110.

ملامهر علیزاده، فاطمه؛ جنتی، مهدی؛ شایان، سیاوش؛ 1383. کارایی داده‌های سنجش‌ازدور (RS) در تهیه نقشه‌های لندفرم و نقش آن در برنامه ریزی محیطی. فصلنامه مدرس علوم انسانی، شماره 4، 111-148.

میرباقری، وهب؛ برادران نصیری، مهران؛ زارع ارنانی، محمد؛ 1387. بررسی امکان پایش تغییرات پوشش گیاهی دشت یزد-اردکان با استفاده از تصاویر ماهواره‌ای لندست بین سال‌های 1976-2002. همایش ژئوماتیک 87. تهران، صص 1-10.

نعیمی نظام آباد، علی؛ قهرودی تالی، منیژه؛ ثروتی، محمدرضا؛ 1388. آشکارسازی تغییرات لندفرم‌های ساحلی منطقه عسلویه. فصلنامه جغرافیایی سرزمین، علمی– پژوهشی، شماره 22، 65-59.

نیازی، یعقوب؛ اختصاصی، محمدرضا؛ ملکی نژاد، حسین؛ حسینی، زین العابدین؛ مرشدی، جعفر؛ 1389. مقایسه دو روش حداکثر احتمال و شبکه عصبی مصنوعی در استخراج نقشه کاربری اراضی (مطالعه موردی: حوضه سد ایلام). مجله جغرافیا و توسعه، شماره 20، 119-132.

Chen KS, Yen SK, Tsay DW., 1997. Neural classification of SPOT imagery through integration of intensity and fractal information. Remote Sensing, 18: 763–783.

Clarke KC, Parks BO, Crane M., 2002. Geographic Information Systems and Environmental Modeling," Prentice hall upper saddle river.

Congalton RG., 1991. A review of assessing the accuracy of classifications of remotely sensed data. Remote Sensing of Environment. 37: 35-46.

DeFries RS, Townshed JG., 1994. NDVI derived land cover classification at a globe scale. Remote Sensing. 5: 3567-3586.

Foody GM., 2002. Hard and soft classifications by a neural network with a non exhaustively defined set of classes. Remote Sensing. 23: 3853–3864.

Howald KJ., 1989. Neural network image classification. Proceedings of the ASPRS-ACSM Fall Convention (Falls Church, VA: American Society for Photogrammetry and Remote Sensing), 207-215.

https://earthexplorer.usgs.gov/

Kavzoglu T, Mather PM., 2004. The use of backpropagating artificial neural networks in land cover classification. International Journal of Remote Sensing, 24: 4907–4938.

Kenneth M, Gunter M., 2012. Monitoring Land-Use Change in Nakuru Kenya Using Multi-Sensor Satellite Data. advance remote sensing, 3: 74-78.

Lillesand TM, Kiefer RW., 1994. Remote sensing & image interpretion. John wiley and sons Inc, New York.

Lu D, Mausel P, Brondizio E, Moran E., 2004. Change detection techniques, Remote Sensing, 25: 2365 – 2401.

Mannan B, Roy J, RAY AK., 1998. Fuzzy ARTMAP supervised classification of multi-spectral remotely-sensed images. Remote Sensing, 19: 767–774.

Masoud AA, Kike K., 2006. Arid land salinization detected by remotely –sensed land cover changes, A case study in the Siwa region. Nw Egypt. Arid Environments, 66: 151-167.

Richards JA, Jia X., 2006. Remote Sensing Digital Image Analysis an Introduction, 4th Edition, Springer, Germany, Berlin, Heidelberg.

Ridd MK, Liu J., 1998. A comparison of four algorithms for change detection in an urban environment. Remote Sensing of Environment, 63: 95-100.

Soffianian A, Madanian MA., 2012. Comparison of Maximum likelihood and Minimum Distance to Mean classifiers in preparing land cover map (A case study: Isfahan area), 57: 253-264.

Sunar F., 1998. An analysis of changes in Multi-dataset, a case study in the Ikitelia area, Istanbul, Turkey. Remote sensing, 19: 225-235.

Tso B, Mather PM., 2009. Classification Methods for Remotely Sensed Data. Taylor and Francis Pub, America.

Yang M, Shoaling W, Tandong Y, Xiaohua G, Anxin L, Xuejun G., 2004. Desertification and its relationship with permafrost degradation in Qinghai-xizang (Tibet) plateau, Cold Reg. Sci. Technol. 39: 47-53.