سنجش و پيش بيني تغييرات پوشش گياهي حوزه استان مازندران طي بازه زماني 2017-2005 با استفاده از زنجيره ماركوف و GIS

Evaluation and prediction of vegetation changes of Mazandaran, Iran from 2005 to 2017 using Markov chain method and Geographical Information Systems GIS

عموماً جهت ارزيابي فرآيندهاي طبيعي، از قبيل اثرات بلندمدت تغيير اقليم كه متأثر از اندركنش مؤلفه‌هاي سازنده سامانه اقليمي از قبيل بيوسفر،ليتوسفر و يا عواملي كه خارج از سامانه اقليمي،تغييرات آب و هوايي را در بازه زماني درازمدت كنترل مي‌نمايند، و همچنين در خصوص فرآيندهاي كوتاه مدت كه شامل توالي پوشش گياهي و فرآيندهاي ژئومورفولوژيكي است، پايش تغيير صورت مي‌گيرد. همچنين، به منظور ارزيابي اثرات ناشي از فعاليت‌هاي انساني از قبيل جنگل‌زدايي، كشاورزي و شهرسازي، پايش تغيير مورد استفاده قرار مي‌گيرد. همانگونه كه تغييرات محيطي انعكاس دهنده وضعيت مديريت اراضي است، روش هاي پايش تغيير مي‌تواند به ارزيابي اين عمليات كمك كند. در اين راستا هدف از پژوهش‌حاضر سنجش و پيش بيني تغييرات پوشش‌گياهي حوزه استان مازندران طي دوره 2017-2005 با استفاده از زنجيره ماركوف و GIS مي‌باشد. براي بررسي و تجزيه تحليل تغييرات از روش طبقه‌بنديdecision tree با توجه به استانداردهاي ناسا ابتدا براي هر valu16 يك كلاس تعريف شد. بر اين اساس مشخص شده است كه آستانه ي تغيير در منطقه ي مورد مطالعه با 1 انحراف از ميانگين قرار داشته است. پس از تعيين آستانه ي تغيير، مناطق داراي تغييرات كاهشي، افزايشي و بدون تغيير مشخص گرديده است. جهت ارزيابي دقت تكنيك هاي سنجش تغيير پس از برداشت واقعيات زميني كه از طريق بازديد ميداني و تصاوير ماهواره اي Google Earth به دست آمد از دقت كل و ضريب كاپا استفاده شد. بر اساس نتايج به دست آمده مشخص گرديد كه داده‌هاي ارزيابي شده با ميانگين دقت كل 91 ، ضريب كاپاي 0/88 را در ارزيابي پايش تغييرات پوشش‌گياهي منطقه‌ي مورد مطالعه به خود اختصاص داده‌اند.

change monitoring is generally used to evaluate natural processes such as the long-term effects of climate change, which is affected by the interaction of the climatic system’s constructive components such as the biosphere, lithosphere, or factors that control the climate changes outside the climatic system, over a long period of time, as well as the short-term processes that include vegetation sequence and geomorphological processes. Change monitoring is also used to evaluate the effects derived from human activities such as deforestation, agriculture and urban development. Remote sensing is a very useful technology, which can be used to obtain information layers from the soil and vegetation. Materials and Methods Land Cover Product was used to process the MODIS1 Satellite data which is one of the most frequently used products designed relating to MODIS Satellite, and is used annually. This Sensor with 250-500 meter and also 1-kilometer spatial resolution has 36 spectral bands in the range of visible, reflectional infrared and thermal infrared wavelengths, which can well be used for various applications of the surface, the Earth surface, atmosphere and the oceans. MOD12Q1, which is one of the MODIS products, was used to investigate and analyze the profile of the vegetation changes in Mazandaran province using the NDVI and EVI indicators from 2005 to 2017. The related images have been prepared annually with 500-meter resolution and sine coordinate system in the form of a combination of Terra and Aqua data. Given the standards provided by NASA, the changes were investigated using the “decision tree” classification method, and the map for the prediction of its changes was calculated using the Markov Chain Model. The ArcGis software was then used to analyze these changes in order to determine which use of land with what percentage of changes has been allocated to which area. Results and Discussion In 2005, land-uses associated with dense vegetation dominated an area of 398.77 m2. These land-uses include wasteland, dense vegetation and scattered vegetation. The estimation of the changes occurring in the aforementioned land-uses showed that the maximum changes relating to the low density vegetation with an average of 55.62% are in the northwestern and the eastern parts, and the minimum changes relating to the in dense vegetation with an average of 77.21% are in the central parts of the region, respectively. Furthermore, the observations of the images of the year 2005 show that the use of dense vegetation which has turned into low density vegetation in the image of the year 2017, has had the minimum changes. Finally, considering the prediction of the observed changes, it can be concluded that these changes were more related to the altitude range of 1400 m to 2260 m with the slope coefficients of 15% to 99%. The prediction carried out using the Markov Chain also suggests that the low-density land cover, which was over 864/80 km2 in 2017, will turn into barren lands in proportion to the changes occurringin 2022. Conclusion A major part of the vegetation changes in the area is due tothelack of job opportunities, extra labor attraction and the economic poverty of the inhabitants.In addition,the pressure on the meadow fields hasreached its highest limit by ranchers,which has resulted inthe reduction of grasslands. Eventually, it could be stated that the evaluationmethods and modelsof the vegetation changes have their own featuresand no method on its own is usable andappropriate for all cases, hence,the identification of an appropriate method for evaluating thevegetation changesneeds to be examined quantitatively and qualitativelyin order to provide the best result.