A three decade assessment of forest cover changes in Nijhum dwip national park using remote sensing and GIS

Spatiotemporal change detection of land use land cover (LULC) in Fashiakhali wildlife sanctuary (FKWS) impact area, Bangladesh, employing multispectral images and GIS

“…To avert the effect of cloud imaging in the land cover classification process, the satellite images used in this study were obtained during the winter season of the considered years: November-February. This choice aligns with previous remote sensing studies in Bangladesh, where most Landsat images from the winter period in the country exhibited negligible or no clouds (Islam et al 2018(Islam et al , 2021Chowdhury et al 2020;Billah et al 2021). Since the study area is primarily an old-growth forest environment, seasonal change, such as sun sensor geometric variation, was considered low or insignificant in this study.…”

“…The post-processing tools of the SCP plugin in QGIS were used to perform the change detection analysis. Furthermore, the classification accuracy of the images was assessed to evaluate the validity of the information obtained from the data using stratified random sampling (Chowdhury et al 2020;Islam et al 2021). To reflect a sizable amount of validation data for each year, 149 randomly selected validation points from across all categories were purposefully sampled.…”

“…Another common image classification accuracy measure is Kappa coefficient (K). Typically, its values range from 0 to 1, where the higher the value, the higher the agreement and accuracy (Billah et al 2021;Islam et al 2021;Hasnat 2021). This K statistic was also computed in this study using the following formula:…”

“…Remote sensing is a fundamental approach for studying spatial and temporal changes in LULC (Pastor-Guzman et al 2018;Islam et al 2018;Mamnun and Hossen 2020). Using medium to high-resolution satellite imageries in remote sensing and GIS applications, dynamic changes in the surface of the planet can periodically and easily be detected (Mallupattu and Reddy 2013;Rawat and Kumar 2015;Islam et al 2021). Landsat satellite images, which have a spatial resolution of 30 ⋅ 30 m, can be a valuable economic data source to collect information from a particular area and identify changes in LULC (Gounaridis et al 2018).…”

“…Detecting and monitoring changes in forested land using widely available remote sensing data and methods also contribute to implementing climate change mitigation initiatives like Reducing Emissions from Deforestation and Forest Degradation (REDD+) and Clean Development Mechanism (CDM) (Sangermano et al 2012;Potapov et al 2014). The use of vegetation indices, such as the Normalized Difference Vegetation Index (NDVI) and Soil Adjusted Vegetation Index (SAVI), has frequently been reported for mapping forest cover (Nath and Acharjee 2013;Nath 2014;Bera and Prakash 2018;Islam et al 2021;Oluwajuwon et al 2021). Studies have also demonstrated how land surface temperature (LST) affects vegetation indices, such as NDVI (Alam et al 2022;Hussain et al 2022).…”