بررسی خشکیدگی مخاطره آمیز درختان جنگلی استان ایلام با استفاده از تصاویر ماهواره لندست

این تحقیق برای بررسی خشکیدگی درختان جنگلی ناحیه غرب کشور در استان ایلام در یک دوره ی 15 ساله با استفاده از تصاویر ماهواره ای لندست انجام شده است. برای بارزسازی پوشش گیاهی در این تصاویر ، پنج شاخص پوشش گیاهی مورد ارزیابی قرار گرفت و با استفاده از نقاط نمونه برداری شده از روی تصاویر گوگل ارث دقت شاخص ها محاسبه گردید. شاخص گیاهی evi با همبستگی بالا با پوشش جنگلی ، نسبت به سایر شاخص ها، به عنوان شاخص بارز کننده پوشش گیاهی در این تحقیق مورد استفاده قرار گرفت. شاخص evi برای این دوره ی 15 ساله محاسبه شد و میزان همبستگی بین تغییرات evi ناحیه ی جنگلی استان ایلام و تغییرات شاخص هواشناسی spi دوره ی یک ساله و سه ماهه و نه ماهه مورد بررسی قرار گرفت که نتایج حاصل بترتیب برابر با 0.46، 0.09 و 0.73 بدست آمد. در ادامه این تحقیق، با روش pca تغییرات هر سال نسبت به سال 1379 محاسبه شد و همچنین با بررسی رفتار تغییرات هر پیکسل در بازه ی زمانی 15 ساله، تغییرات ناشی از خشکیدگی درختان از سایر تغییرات، تفکیک شد و همچنین میزان این تغییرات برای هر سال محاسبه شد و میزان تاثیرگذاری تغییرات بارش و ریزگردها در افزایش روند خشکیدگی درختان جنگلی استان ایلام، با انجام یک رگرسیون خطی دو متغییره بدست آمد که نتایج سهم 62 درصدی تغییرات بارش و 38 درصدی گردوخاک را نشان می دهند.

Assessment of Hazardous Drought of Ilam Province Forests using Landsat Satellite Images

Forest dieback is a complex and important phenomenon that happened in the world rsquo;s most oak forests nearly a century ago. In recent years, this has occurred in ZAGROS oak forests due to successive droughts. ZAGROS forest region, with an area of about 6 million hectares including provineces West Azerbaijan, Kurdistan, Kermanshah, Ilam, Lorestan, Khuzestan, Fars, Esfahan, Chaharmahal and Bakhtiari, Kohgiluyeh and BoyerAhmad, Hamedan is exposed to several threats. Forests of ILAM province are part of ZAGROS forests, which are located in the west part of the mountain chain. The phenomenon of forest dieback has been seen in this area during recent years. In this study the dieback of forests in ILAM province area has been studied using Landsat satellite imagery and Google earth images, meteorological data and information of the amount of dust in province atmosphere. This study has conducted over a period of 15 years. Also we have to mention that the dust data have been extracted from 550 nm band Modis satellites images. In order to detect vegetation and forest area of ILAM, vegetation indices have been used in Landsat satellite images. In this study, the five vegetation indices, NDVI (Normalized Difference Vegetation Index), EVI (Enhanced Vegetation Index), RVI (Relative Vigor Index), SAVI (Soil Adjusted Vegetation Index), ARVI (Atmospherically Resistant Vegetation Index) have been used. Also to determine the appropriate vegetation index for detecting the changes in vegetation coverage areas, these indicators have been compared to each other. In order to compare these indicators, Google earth imagery has been used. Also with the use of Google earth images, the vegetation coverage of 200 pixels has been estimated in Landsat satellite images. By using these 200 pixels we compared the vegetation indices and carefully calculated the accuracy of each one. In this study, the generalized correlation of the indices with the vegetation coverage changes is respectively: 0.939, 0.953, 0.945, 0.914, 0.925 and according to these results, EVI vegetation index is used as the preferable indicator for vegetation change detection index in ILAMs forests. Landsat Images were obtained for each year and the vegetation changes amount was calculated towards the previous year for each year. Using the rainfall data index SPI (Standardized Precipitation Index), in these periods of one year, ninemonth and threemonth. The correlation between EVI index annual changes examined using SPI index the results were respectively, 0.43, 0.76, 0.09. In this paper, we used PCA (Principal Component Analysis) method to detecte the changed areas in forest vegetation coverage. The rate of changes in pixels amount in each year is calculated respect to year 1379. Also in order to separate the forest area changes from others, the behaviour of each pixel is studied during a period of 15 years. Also with the introduction of two patterns and determination of correlation between the patterns change behaviour, ILAMs vegetation coverage changed areas were isolated from other changes, and by deleting the non ndash;changed pixels, the average annual change was obtained for ILAM province forest. Average entered dust was calculated for each year, using meteorological and Modis satellite images data. Also by applying a two ndash;parameter linear regression, the combined impact of two factors, rainfall and dust was determined. The study implied that precipitation is the most effective parameter on dieback. The influence of two factors, rainfall and dust are respectively 62% and 38%.