عنوان مقاله [English]
Background: Identifying the relationship between conservation of biodiversity and ecosystem processes is one of the main topics in ecological research. Forests are one of the valuable natural resources of the planet, which plays an important role in the ecological balance and the lives of human societies. Diversity of tree species is one of the key parameters for describing forest ecosystems in the management of nature. Modeling and preparing a map of tree diversity is a useful tool for conservation and management of forests. In terms of tree diversity, the Caspian forests are the richest forests in Iran that have undergone severe changes in recent years. One of the most important and most effective ways to learn about tree diversity is the use of satellite imagery. The aim of this study was to determine the capability of GeoEye images in monitoring of tree diversity in circular forest forests in Mazandaran province.
Materials and Methods: For this purpose, 150 landfills of 30 to 30 meters in length were used for field surveying. Then, the Shannon-Weiner, Simpson and Simpson indexes were calculated for each sample piece. Pre-processing and necessary processing, such as principal component decomposition, making vegetation indices and texture analysis were performed on the images. For modeling, classification and regression tree methods, random forest, different variants of the nearest neighbor and different kernels of backup machine were used. Of the 70% of the teaching samples used for modeling. Then the best bands were selected for modeling. Models were evaluated using 30% of the samples. Then the best models were specified for each part.
Results: The results showed that the indices produced, the infrared band and the resulting bands were identified as the best band for modeling. The RBF kernel has the best result from a backup machine with a 58% explanatory factor and a root mean square error of 46% for modeling the Shannon Wiener Diversity Index among the above models. Also, the method of random forest with the coefficient of explanation of 54 and 57 percent, and the root mean square error margin of about 48 and 14 percent, respectively, for the Simpson and Simpson images have the best results.
Conclusion: The results showed that GeoEye satellite data has a relatively good ability to estimate tree diversity in circular forest forests. The models used by the random forest method for two modes and the RBF kernel were the best-performing vector-path vector technique in one state. Overall, the results showed that these data can be used for management, conservation and monitoring of tree diversity in the northern forests of the country.
Keywords: tree diversity, Shannon Wiener, Simpson and Circulation Series