مدل سازی مناسب ترین شاخص های پوشش گیاهی تحت تأثیر عوامل اقلیمی با استفاده از تصاویر سنتینل 2، مطالعه موردی: جنگل فریم (مقاله علمی وزارت علوم)
درجه علمی: نشریه علمی (وزارت علوم)
آرشیو
چکیده
تغییر اقلیم تأثیر قابل ملاحظه ای بر محیط زیست دارد و منجر به حساسیت متفاوت پوشش گیاهی به عوامل آب و هوایی در مقیاس های مکانی- زمانی مختلف می شود. آگاهی از وضعیت پوشش گیاهی به دلیل کاربرد در برنامه ریزی های خرد و کلان در حال حاضر از ارکان مهم در تولید اطلاعات است .با توجه به پرهزینه و زمان بر بودن استفاده از روش های مبتنی بر مشاهدات، امروزه فناوری سنجش از دور به عنوان راهکار جدید در بهبود این روش ها مطرح شده است. در پژوهش پیش رو هدف، بررسی اثر عوامل اقلیمی بر روند پوشش گیاهی جنگل فریم در استان مازندران با استفاده از تصاویر سنتینل 2 و تعیین مناسب ترین شاخص برای این منطقه است. به منظور مدل سازی از فاکتورهای اقلیمی (درجه حرارت و بارندگی) مربوط به منطقه به دست آمده از نزدیک ترین ایستگاه هواشناسی مربوط، استفاده شد. بعد از پیش پردازش و پردازش تصاویر سنتینل 2 ارزش های رقومی متناظر از باندهای طیفی استخراج و به عنوان متغیرهای مستقل در نظر گرفته شد. رابطه درجه حرارت و بارندگی با شاخص های پوشش گیاهی با ضریب همبستگی 0.43 و 0.56 و میزان AIC و BIC به ترتیب (565 و 3209) و (739 و 3383) به دست آمد. همچنین نتایج نشان داد بیش ترین اثرگذاری در رابطه با هر دو فاکتور درجه حرارت و بارندگی مربوط به شاخص پوشش گیاهی تفاضلی (DVI) است، که کارائی بالای این شاخص در منطقه را نشان می دهد. با توجه به نتایج فوق، می توان بیان کرد که شاخص مذکور به منظور بررسی تأثیر متغیرهای اقلیمی بر جنگل مورد مطالعه، انطباق و همبستگی مناسبی دارد.Modeling the most appropriate vegetation indices under the influence of climatic factors using sentinel 2 images - Case study: Farim Forest
Extended IntroductionVarious Climate factors considerably affect the environment and different vegetation covers show different levels of sensitivity to climate factors in the spatial-temporal scale. Data specifically collected from vegetation cover plays an important role in micro and macro planning and information generation. Methods using air temperature recorded in weather stations to estimate the relative heat in urban areas are considered to be both time-consuming and costly. On the other hands, data with relatively high spatial resolution are capable of measuring ground surface parameters more efficiently and accurately. Thus, remote sensing technology is now considered to be a solution used to improve previously mentioned methods. Remotely sensed data are now widely used to find the quantitative relationship between patterns of vegetation cover and the elements of climate. Predicting the conditions of vegetation cover is considered to be essential for planners seeking an efficient plan for its exploitation and protection.Materials & MethodsThe present study seeks to investigate the effects of climatic factors on the vegetation trend observed in Frame forest in Mazandaran province using Sentinel 2 images and to determine the most suitable index for this area. Climatic Data collected from the nearest weather station in Farim City have been used to model climate factors (temperature and precipitation). Changes in the height above mean sea level were also considered. Following the pre-processing and processing of the Sentinel 2 images, the corresponding digital values were extracted from the spectral bands and applied as independent variables. ENVI software was used for image processing and STATISTICA and R software were used for modeling. 70% of the resulting data were used for training and the rest were used for testing or evaluating the model. Mean square error, correlation, Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) were used to evaluate the presented models. Models with the highest correlation and the lowest standard error, the mean square error, the Akaike information evaluation criterion and the Bayesian evaluation criterion were selected as the best models for the studied variables.Results & Discussion A correlation coefficient of 0.43 and 0.56 was observed between temperature and precipitation and vegetation indices. AIC and BIC values equaled (565 and 3209) and (739 and 3383) respectively. Differential Vegetation Index (DVI) has proved to be the most effective parameter in relation to both temperature and precipitation factors in the region. Results indicated that differential vegetation index, green normalized difference vegetation index (GNDVI) and green difference vegetation index (GDVI) have a positive correlation with temperature, while there is a negative correlation between temperature and normalized vegetation index. Precipitation is considered to be one of the most important factors affecting vegetation. Results indicate that differential vegetation index, green difference vegetation index, green normalized difference vegetation index, non-linear vegetation index and normalized difference vegetation index have the highest impact on precipitation. In forest ecosystems, changes in climatic factors may affect trees differently. ConclusionCollecting information about the state of vegetation cover in forests is considered to be very important. Thus, the present study has endeavored to investigate the relationship between indices of vegetation cover and climatic variables. To reach this aim, satellite data are used as a suitable and efficient tool for investigating forest ecosystems with a relatively low cost. This provides the possibility of continuously monitoring land surface. Results indicated that climatic factors affect vegetation indices in the study area. Vegetation cover protects and stabilizes the environment and thus, many researchers have tried to investigate the growth and spatial patterns of vegetation cover in different regions. It is also suggested to study the effects of climatic factors on the vegetation cover of the study areas in different geographical directions. In addition, using other climatic factors such as relative humidity, wind speed, evaporation, transpiration, and higher resolution images can increase the accuracy of the study.
تبلیغات
