The Rohingya refugee crisis in Bangladesh: assessing the impact on land use patterns and land surface temperature using machine learning.

Bangladesh, a third-world country with the seventh highest population density in the world, has always struggled to ensure its residents' basic needs. But in recent years, the country is going through a serious humanitarian and financial crisis that has been imposed by the neighboring country Myanmar which has forced the government to shelter almost a million Rohingya refugees in less than 3 years (2017-2020). The government had no other option but to acquire almost 24.1 km2 of forest areas only to construct refugee camps for the Rohingyas which has led to catastrophic environmental outcomes. This study has analyzed the land use and land surface temperature pattern change of the Rohingya camp area for the course of 1997 to 2022 with a 5-year interval rate. Future prediction of the land use and temperature of Teknaf and Ukhiya was also done in this process using a machine learning algorithm for the years 2028 and 2034. The analysis says that in the camp area, from 1997 to 2017, percentage of settlements increased from 5.28 to 11.91% but in 2022, it reached 70.09%. The same drastically changing trend has also been observed in the land surface temperature analysis. In the month of January, the average temperature increased from 18.86 to 21.31 °C between 1997 and 2017. But in 2022. it was found that the average temperature had increased up to 25.94 °C in only a blink of an eye. The future prediction of land use also does not have anything pleasing in store.

Morphology and land use change analysis of lower Padma River floodplain of Bangladesh.

Bangladesh, the largest delta in the world, is situated over the flow of more than 700 rivers. The Ganges is a transboundary river, and it takes the Padma after meeting Jamuna near Aricha. The Padma River's morphology and hydraulic parameters are so dynamic that it erodes a large portion of land every year. However, the situation of erosion has been very threatening since 2014 which is almost the same period when the construction of the Padma Bridge started. Our Study on the erosion-accretion rate and bar dynamics of the selected reach of Padma River shows that the downstream right bank of this river lost around 134.85 sq. km of land between 2003 and 2021. The total bar area has also increased to 76.8%. Land use land classification (LULC) analysis was done for the years 2003, 2009, 2015, and 2021 to predict how the river might act in the future. For this prediction, artificial neural network (ANN) system has been used, and the predicted land use map was generated for 2027. The current kappa validation was 0.869, and the accuracy of the prediction was 87.05%. This study aims at analyzing the current morphological condition and its connection to the construction of the Padma Bridge and also predicts how the lower Padma River might act in near future. According to the study findings, in absence of prompt and appropriate measures, this country is highly susceptible to experiencing catastrophic consequences.