نوع مقاله : پژوهشی

نویسندگان

1 دانشکده مهندسی نقشه برداری و اطلاعات مکانی پردیس دانشکده های فنی دانشگاه تهران

2 دانشجوی کارشناسی ارشد هیدروگرافی دانشکده‌ی مهندسی نقشه‌برداری و اطلاعات مکانی، پردیس دانشکده‌های فنی، دانشگاه تهران، تهران، ایران.

3 ـ استادیار گروه مهندسی­ آب، دانشکده کشاورزی، دانشگاه تبریز. تبریز، ایران.

چکیده

چکیده
شوری آب مشخص‌کننده­ی پراکندگی حیات جانوری و گیاهی در دریاچه­ها، دریاها و اقیانوس­هاست. در این مقاله با مطالعه بازتابش از سطح آب، شاخص­های شوری و همچنین داده­های میدانی نقشه­ی شوری دریاچه­ی ارومیه تهیه گردید. سپس مدل­سازی شوری این دریاچه با استفاده از رگرسیون بردار پشتیبان و تصاویر لندست-8 انجام گردید. جهت انتخاب ویژگی­های مناسب از میان هفده ویژگی ورودی اولیه­ی مدل از دو الگوریتم ژنتیک و انتخاب ویژگی ترتیبی به کمک نرم­افزار متلب استفاده شد. در نهایت میزان شوری آب دریاچه­ی ارومیه با خطا و دقت نسبتاً مناسبی تخمین گردید. به­طوری­که مدل رگرسیون بردار پشتیبان با تمام ویژگی­ها با RMSE=24.55psu و R2=41%، مدل رگرسیون ­بردار پشتیبان مبتنی بر الگوریتم ژنتیک با RMSE=21.97psu و R2=54% و مدل رگرسیون ­بردار پشتیبان مبتنی بر انتخاب ویژگی ترتیبی با مقادیر RMSE=21.93psu و R2=53% توانستند میزان شوری دریاچه­ی ارومیه را تخمین بزنند.

تازه های تحقیق

-

کلیدواژه‌ها

عنوان مقاله [English]

Urmia Lake Salinity Mapping Using Support Vector Regression and Landsat-8 Imagery

نویسندگان [English]

  • Mahdi Hasanlou 1
  • Meysam Jamshidi 2
  • Mohammad Taghi Sattari 3

1 School of Surveying and Geospatial Engineering, College of Engineering, University of Tehran

2 M.SC Student, School of Surveying and Geospatial Engineering, College of Engineering, University of Tehran.

3 Assistant Professor, Department of Water Engineering, Faculty of Agriculture, University of Tabriz

چکیده [English]

Introduction
Urmia Lake is located in the North West of Iran and its area between 4750 to 6100 square kilometers at an altitude of 1250 meters above sea level. This lake is a permanent lake in Iran. In fact, Urmia Lake is one of the lowest parts of the catchment area North West of Iran. The total surface area of Urmia Lake is 51,876 km square, which is 3.15% of the total area of Iran and 7% of all the water’s surface in the country. The depth varies between 6 and 16 meters, the length of the lake is 50 km and its width varies between 128 km to 140 km.  In the catchment area of the lake, there is the main river with annual input about 2 billion cubic meters. Annual rainfall in the catchment area is variable between 200 and 300 mm. Air temperature the area around the lake in winter to 20°C and 40°C in summer increases. Urmia Lake is important in terms of economically, transport, exploitation of the mineral wealth of biodiversity, mitigating climate, and tourism. This unique Lake addition to the previous is habitat for kind of native artemia its name is urumiana artemia that this artemia is unique to this lake. Also, Urmia Lake is the world's second largest habitat for Artemia. According to the research, the main elements in the Urmia Lake include Cl-, Na +, Ca2+, Mg2+, HCo3-, K+, Li, So42- and F.
Methodology
In this study, newly launched Landsat series (Landsat-8) was used for monitoring Urmia Lake salinity and retrieving the salinity map. By incorporating the Landsat-8 datasets, this study determined the salinity changes and created a model to estimate the salinity in Urmia Lake with processing Landsat-8 satellite images as a result; we can obtain salinity map regularly without ground operations. We can also monitor the health of the habitat in terms of salinity and examine the impact of increasing salinity on the plants, animals, and ecosystems of the region. This study applied remote sensing techniques to develop a salinity prediction model for Urmia Lake. In this study, we use Landsat-8 satellite images radiances of Urmia Lake and some salinity indices and in-situ data so we have 17 features to make water surface salinity model with support vector regression (SVR) with all features. After that, we use two algorithms; GA and SFS for selecting suitable features and make models with those features.
Result
Results with all features model show RMSE=24.55 and R2=41% and result with GA feature selection model shows RMSE=21.97 and R2=54% and results with SFS feature selection model shows RMSE=21.93 and R2=53%.
 
Discussion and Conclusion
Satellite images show that from 1995 to 2003, the lake water surface dropped and proportionate to the dropping water salinity increased to 220 to 300 grams per liter. Also although Artemia is resistant to salt, appropriate salinity is below 100 grams per liter. When water salt is more than 100 grams per liter contents of his tiny body lost and die. Now because of reduction in salinity, the lake has arrived at about 300 grams per liter. Dissolved salt in water has a direct effect on the electrical conductivity of water. In this regard, incorporating high spatial resolution satellite like Landsat-8 images is inevitable. Also, the proposed modeling methods show these changes in multi-data and in widespread Urmia Lake very well.

کلیدواژه‌ها [English]

  • Keyword: Water salinity modeling
  • Urmia lake
  • remote sensing
  • Landsat-8
  • SVR
  • GA
  • SFS
منابع
ـ علوی­پناه، سیدکاظم؛ کمال، خدایی و جعفر بیگلو (1384)، مطالعه­ی کارایی داده­های ماهواره­ای در بررسی کیفیت آب در دو سوی میانگذر دریاچه­ی ارومیه، پژوهش­های جغرافیایی، 38، شماره­ی 1، صص 57-69.
ـ رشیدنیقی، علی؛ و مجنونی، ابولفضل و امیرحسین ناظمی (1391)، اثرات زیانبار خشک شدن دریاچه­ی ارومیه بر محیط زیست منطقه،همایش بین­المللی بحران­های زیست محیطی و راهکارهای بهبود آن، صص 1-10.
ـ مرادی، نسرین (1395)، بررسی و مدل‌سازی رنگ اقیانوس با استفاده از تصاویر ماهواره‌ای با توان تفکیک مکانی بالا، پایان‌نامه کارشناسی ارشد در رشته مهندسی عمران  -نقشه‌برداری گرایش هیدرو گرافی- دانشگاه تهران.
ـ نیازی، یعقوب؛ اختصاصی، محمدرضا؛ ملکی­نژاد، حسین؛ حسینی، سیدزین­العابدین و جعفر مرشدی (۱۳۸۹)، مقایسه­ی دو روش طبقه­بندی حداکثر احتمال و شبکه­یعصبی مصنوعی در استخراج نقشه­ی کاربری اراضی- مطالعه­ی موردی: حوضه­ی سد ایلام، جغرافیا و توسعه، شماره­ی ۲۰، صص ۱۱۹-۱۳۲.
-Agh, Naser, Theodore J. Abatzopoulos, Ilias Kappas, et al. (2007), Coexistence of Sexual and Parthenogenetic Artemia Populations in Lake Urmia and Neighbouring Lagoons, International Review of Hydrobiology, 92(1): PP,48-60.
-AghaKouchak, A., Norouzi, H., Madani, K., Mirchi, A., Azarderakhsh, M., Nazemi, A., & Hasanzadeh, E. (2015), Aral Sea syndrome desiccates Lake Urmia: call for action, Journal of Great Lakes Research, 41(1): PP,307-311.
-Aha, David W., and Richard L. Bankert. (1996), A comparative evaluation of sequential feature selection algorithms, Learning from Data, Springer New York: PP,199-206.
-Alipour, S.  (2006), Hydrochemistry of Seasonal Variation of Urmia Salt Lake Iran, Saline Systems, 2(9): PP,1-19.
-Baban, S.M. (1993), Detecting water quality parameters in the Norfolk Broads, UK, using Landsat imagery, International Journal of Remote Sensing, 14(7): PP,1247–1267.
-El-Askary, H. S. Abd El-Mawla, J.Li, M. El-Hattab, and M. El-Raey, (2014), Changedetection of coral reef habitat using Landsat-5 TM, Landsat 7 ETM+ and Landsat 8 OLI data in the Red Sea (Hurghada, Egypt), International Journal of Remote Sensing, 35(6): PP,2327-2346.
-Bingham, Frederick M, Stephan D Howden, and Chester J Koblinsky (2002), Sea Surface Salinity Measurements in the Historical Database, Journal of Geophysical Research 107(C12): SRF 20-1-SRF, PP, 20-10.
-Bhargava, D.S., and D.W. Mariam (1992), Cumulative Effects of Salinity and Sediment Concentration on Reflectance Measurements, International Journal of Remote Sensing 13(11): PP,2151–2159.
-Bowers, D.G., G.E.L. Harker, P.S.D. Smith, and P. Tett (2000), Optical Properties of a Region of Freshwater Influence (the Clyde Sea), Estuarine, Coastal and Shelf Science 50(5): PP,717–726.
-Chang and C.-J. Lin, (2011), LIBSVM: A Library for Support Vector Machines, ACM Trans. Intell. Syst. Technol., 2(3): PP,271–272.
-Ghalibaf, Mohammad Bagher, and Zahra Moussavi (2014), Development and Environment in Urmia Lake of Iran, European, Journal of Sustainable Development 3(3): PP,219-226.
-Kang, Kyung Chan (2014), Seawater desalination by gas hydrate process and removal characteristics of dissolved ions, Desalination, 353(1): PP,84-90.
-Khorram, Siamak (1985), “Development of Water Quality Models Applicable throughout the Entire San Francisco Bay and Delta, Photogrammetric Engineering and Remote Sensing 51(1): PP,53–62.
-Marghany, MAGED, and MAZLAN Hashim (2011), A Numerical Method for Retrieving Sea Surface Salinity from MODIS Satellite Data, International Journal of Physical Sciences, 6(13): PP,3116–3125.
-Marghany, M.M. Hashim, and A.P. Cracknell (2010), Modelling sea surface salinity from MODIS Satellite data, in International Conference on Computational Science and Its Applications: PP,545–556.
-M.A. Tahir, A. Bouridane, F. Kurugollu, and A. Amira (2004), Feature selection using tabu search for improving the classification rate prostate needle biopsies, ICPR 2004, Proceedings of the 17th International Conference on, PP,335–338.
-Marghany, Maged, Mazlan Hashim, and Arthur P. Cracknell (2010), Modelling Sea Surface Salinity from MODIS Satellite Data, In International Conference on Computational Science and Its Applications, Springer: PP,545–556.
-Nikraftar, Z., M. Hasanlou, and M. Esmaeilzadeh (2016), Novel Snow Depth Retrieval Method Using Time Series SSMI Passive Microwave Imagery, ISPRS-International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences: PP,525-530.
-Nikraftar, Z. and M. Hasanlou (2015), Snow Depth Estimation Using Time Series Passive Microwave Imagery via Genetically Support Vector Regression (Case Study Urmia Lake Basin), The International Archives of Photogrammetry, Remote Sensing and Spatial Information Sciences: PP,555-558.
-Palandro, S. Andréfouët, P. Dustan, and F. Muller-Karger (2003), Change detection in coral reef communities using Ikonos satellite sensor imagery and historic aerial photographs, International Journal of Remote Sensing, 24(4): PP,873-878.
-Wang, Fugui, and Y. Jun Xu (2008), Development and Application of a Remote Sensing-Based Salinity Prediction Model for a Large Estuarine Lake in the US Gulf of Mexico Coast, Journal of Hydrology 360(1): PP.184-194.
-Wong, Man-Sing, Kwon-Ho Lee, Young-Joon Kim, et al. (2007), Modelingof Suspended Solids and Sea Surface Salinity in Hong Kong Using Aqua/MODIS Satellite Images, Korean Journal of Remote Sensing 23(3): PP,161-169.