|
چکیده
|
Fires affect the structure, composition, and functioning of ecosystems. Monitoring post-fire vegetation recovery is critical for evaluating the ecological impacts of wildfires effects on ecosystems, and supporting ecosystem restoration after fires. Each year, many low- to moderate-intensity surface fires occur in the Zagros forests. This study aimed to assess vegetation recovery in a burned area of the northern Zagros forests by analyzing the spectral behavior of vegetation post-fire. Temporal OLI-Landsat 8 images from 2013 to 2018 were used. In the first step, accurate atmospheric correction was applied to the images. Five vegetation indices (VIs), NBR, NDVI, NDII, EVI, and MSAVI1 were used to monitor the vegetation recovery from two years pre-fire to three years post-fire. Both qualitative (i.e. visual analysis) and a quantitative (i.e. M-statistic separability index) approaches were used to evaluate the performance of VIs in distinguishing between burned and unburned areas. Surveys were conducted simultaneously in the burned and control areas with similar physiographic conditions to minimize phenological differences and isolate the effects of meteorological variations such as rainfall and drought, on vegetation regeneration and spectral changes. Visual assessment showed that all VIs responded well in detecting post-fire recovery, highlighting their effectiveness. Among the VIs, NBR demonstrated the highest discrimination performance, while NDVI showed the lowest sensitivity. The results showed that all VIs nearly returned to their pre-fire values within one year post-fire. This rapid recovery is attributed to the low intensity of surface fires and the fast coppice regeneration of oak trees post-fire which leads to increased canopy cover. Therefore, identifying burned areas becomes challenging just one year after the fire in some parts of Zagros forests due to the rapid recovery of vegetation.
|