Nazmul Haque S M, Uddin Md Jahir
Department of Civil Engineering, Ahsanullah University of Science and Technology, Dhaka, 1208, Bangladesh.
Department of Civil Engineering, Khulna University of Engineering & Technology, Khulna, 9203, Bangladesh.
Sci Rep. 2025 Aug 25;15(1):31263. doi: 10.1038/s41598-025-07443-2.
Sylhet, located in the northeastern part of Bangladesh, is characterized by a unique topography and climatic conditions that make it susceptible to flash floods. The interplay of rapid urbanization and climatic variability has exacerbated these flood risks in recent years. Effective monitoring and planning of land use/land cover (LULC) are crucial strategies for mitigating these hazards. While former studies analyzed LULC in parts of Sylhet using traditional GIS approaches, no comprehensive, district-wide assessment has been carried out using long-term satellite data and cloud computing platforms. This study addresses that gap by applying Google Earth Engine (GEE) for an extensive analysis of LULC changes, transitions, and hot/cold spots across the district. Accordingly, this work investigates the LULC changes in Sylhet district over the past twenty-three years (2000-2023). Using satellite imagery from Landsat 7 Enhanced Thematic Mapper Plus (ETM+) and Landsat 8 Operational Land Imager (OLI), the LULC is classified in six selected years (2000, 2005, 2010, 2015, 2020, and 2023). A supervised machine learning algorithm, the Random Forest Classifier, is employed on the cloud computing platform Google Earth Engine to analyze LULC dynamics and detect changes. The Getis-Ord G statistical model is applied to identify land transformation hot spot and cold spot areas. The results reveal a significant increase in built-up areas and a corresponding reduction in water bodies. Spatial analysis at the upazila level indicates urban expansion in every upazila, with the most substantial increase observed in Beani Bazar upazila, where urban areas expanded by approximately 1500%. Conversely, Bishwanath upazila experienced the greatest reduction in water bodies, with a decrease of about 90%. Sylhet Sadar upazila showed a 240% increase in urban areas and a 72% decrease in water bodies. According to hotspot analysis, Kanaighat upazila has the most amount of unchanging land at 7%, whereas Balaganj upazila has the largest amount of LULC transformation at 5.5%. Overall, the urban area in the Sylhet district has grown by approximately 300%, while water bodies have diminished by about 77%, reflecting trends of urbanization and river-filling. These findings underscore the necessity of ensuring adequate drainage infrastructure to decrease flash flood hazards in the Sylhet district and offer insightful information to relevant authorities, politicians, and water resource engineers.
锡尔赫特位于孟加拉国东北部，其独特的地形和气候条件使其容易遭受山洪暴发。近年来，快速城市化与气候多变性的相互作用加剧了这些洪水风险。有效的土地利用/土地覆盖（LULC）监测和规划是减轻这些灾害的关键策略。虽然以前的研究使用传统地理信息系统方法分析了锡尔赫特部分地区的LULC，但尚未利用长期卫星数据和云计算平台进行全面的全区评估。本研究通过应用谷歌地球引擎（GEE）对全区的LULC变化、转变以及热点/冷点进行广泛分析，填补了这一空白。因此，这项工作调查了锡尔赫特地区在过去二十三年（2000 - 2023年）的LULC变化。利用陆地卫星7号增强型专题制图仪升级版（ETM +）和陆地卫星8号业务陆地成像仪（OLI）的卫星图像，在六个选定年份（2000年、2005年、2010年、2015年、2020年和2023年）对LULC进行分类。在云计算平台谷歌地球引擎上采用监督式机器学习算法——随机森林分类器，来分析LULC动态并检测变化。应用Getis - Ord G统计模型来识别土地转换热点和冷点区域。结果显示建成区显著增加，水体相应减少。在县一级的空间分析表明每个县都有城市扩张，在比尼巴扎尔县观察到的增长最为显著，城市面积扩大了约1500%。相反，比什瓦纳特县的水体减少最多，减少了约90%。锡尔赫特市辖区的城市面积增加了240%，水体减少了72%。根据热点分析，卡奈加特县不变土地的比例最高，为7%，而巴拉甘杰县的LULC转换量最大，为5.5%。总体而言，锡尔赫特地区的城市面积增长了约300%，而水体减少了约77%，反映了城市化和河流淤积的趋势。这些发现强调了确保有足够排水基础设施以减少锡尔赫特地区山洪灾害的必要性，并为相关当局、政治家和水资源工程师提供了有价值的信息。
