Assessment of the Forest Disturbances Rate Caused by Windthrow Using Remote Sensing Techniques

摘要

Change detection using multi-temporal satellite images data is an important domain with various applications in forestry and can allow an evaluation of areas extended to the same spatial temporal scale. The focus of the study is to assess the changes occurring after catastrophic wind events using Landsat time series data. Estimates of disturbance rates are derived using 8 sample sites selected across the Apuseni Mountains during 2000-2014 periods. Multi-temporal analysis on annual basis has detected the patterns of the changing forest ecosystem and the trends that are occurring and give more accurate results. The satellite images have been calibrated and the root mean square error has been made The satellite images preprocessing is made in order to transform the DN values into the surface reflectance. The approach requires images during the peak growing season. Local knowledge and available ancillary data about windthrow occurrence are required in order to fully understand the nature of these trends. The statistical algorithms are applied to characterize the magnitude of the disturbance. We found evidence of systematic change in the forest ecosystem of the Apuseni Mountains by analyzing multi-temporal surface data. The accuracy of forest disturbance detection diminishes with the decrease of the temporal resolution. Therefore, the approach described in this paper demonstrates that the Landsat time series data can be used operationally for assessing forest cover changes analysis after a windthrow occurrence across a large area.