Life Science Journal 2012;9(4) http://www.lifesciencesite.com 1475 An Exploration into Spatial - Temporal Variations Trend Focusing on Forest Classification and Adoption of Classified Error Matrix (Case Study: Central Zagros) Behnaz Moradi Ghiyasabadi 1 (Corresponding author), Masoud Monavari 2 , Alireza Ghragozloo 3 , Nematollah Khorasani 4 , and Borhan Riyazi 5 1- PhD student in Department of Environmental and energy, Science and Research branch, Islamic Azad University, Tehran, Iran 2- Assistant professor in Department of Environmental and energy, Science and Research branch, Islamic Azad University, Tehran, Iran 3- Assistant professor in Geomantic collage of NCC OF Iran. 4- Assistant professor in Department of Environmental and energy, Science and Research branch, Islamic Azad University, Tehran, Iran 5- Assistant professor in Department of Environmental and energy , Science and Research branch , Islamic Azad University , Tehran, Iran. Abstract: In order to study on the rate of variations in Zagros central region, LANDSAT satellite data from years 1976-2008 were used to prepare a map comprising five classes (of land use) including agriculture, urban, forest, rocks and ranch. Accordingly, the satellite imageries relating to the given period were prepared, interpreted and mapped and variations trend was quantitatively computed, and obtained data were compared to each other while this task was done at three levels i.e. Macro-level (Central Zagros), Meso-level (Dena Protected Area) and Micro-level (Western Dena). On the other hand, measurement of canopy has been introduced as an appropriate factor in forest management and classification at micro level (11). To prepare the given map, distribution of canopy was classified by means of arithmetic interpretation of aerial photographs so that scanned aerial photos 1:20000 (1968) and 1:40000 (2001) were prepared by application of PCI Geometrica orthographically at first step and mosaic pattern was arranged from them. In the next step, the arithmetic orthophotomosaic was classified in three classes (dense canopy, semi-open canopy and open canopy forests) and the resulting map from this classification was prepared using ArcGIS System. To calculate canopy, dotted network with 0.5mm intervals was used as arithmetic layer. Statistical analyses have been adopted for a 30-year period by means of Maximum Likelihood Supervised Technique and in order to determine variation in contrast method after classification as well as Maximum Likelihood Algorithm. The obtained results indicated the rates of total accuracy in images classification for 1976 and 2008 as 90.22% and 94.09% respectively. The computations suggest that areas for farming lands, residential use and open canopy forest have increased, while the dense, semi-open and open canopy forest areas as well as ranches and rocks have been reduced. [Behnaz Moradi Ghiyasabadi, Masoud Monavari, Alireza Ghragozloo, Nematollah Khorasani, and Borhan Riyazi. An Exploration into Spatial - Temporal Variations Trend Focusing on Forest Classification and Adoption of Classified Error Matrix (Case Study: Central Zagros). Life Sci J 2012;9(4):1475-1486] (ISSN:1097-8135). http://www.lifesciencesite.com. 225 Keywords: Variation Trend, Error Matrix, Remote Sensing, Satellite Imageries, Aerial Photographs Introduction With respect to today’s importance of ecosystems survey, preparation of applied maps for lands and evaluation and study on variation trend of eco-systemic structure have turned into a crucial approach in the field of resources management (31, 34). In this sense, remote sensing together with aerial photographs and ground control systems are employed as powerful tools to determine and analyze land use planning and land coverage (28, 30). Assessment of land coverage variations is a process that may lead to a proper concept from way of interaction between humans and the environment, and is of greater importance for mountainous- forestial regions (46). Central Zagros and Dena Protected Zone are one of the biospheric deposits in Iran that are also registered in UNESCO, and solely possess several endemic plant varieties as much as 46 countries throughout of the world. The investigation may show that there are 2000 plant varieties and several medicinal types of plants throughout Kohkilouyeh and Boyer-Ahmad Province in Iran, out of which one thousand plant varieties grow in Dena Protected Zone and these numbers are related the indigenous and unique varieties in the Dena region (8). However, for some period of time, this region has been exposed to many threats; as a result, in order to manage and protect green areas and tree coverage, it requires depiction of these variations and planning and