Mapping, Dynamics, and Future Change Analysis of Sundarbans Delta Using Cellular Automata and Artificial Neural Network Modeling

Abstract

Over the past century, anthropogenic activities have had a significant impact on low-lying deltas. In this article, we utilized Landsat satellite images and sea-level data from the past thirty years to analyze the effects of land use/land cover (LULC) change, vegetation cover, and terrain temperature distribution on the Sundarbans delta. Anticipated LULC change has been simulated using a cellular automata algorithm coupled with an artificial neural network for 50 years into the future, considering climate change and other ecological drivers. The results of our analysis show that the thick, dense mangrove coverage in the Sundarbans has decreased from 65% in 1970 to 59% in 2020. However, the intricate network of channels has increased in width, and their coverage increased from 33% to 43% between 1990 and 2015. The simulations for 2030–2050 show that the mangrove coverage is expected to decline from 63.5% to 56.7% (a decrease of 628 km2), while the total area of waterbodies will increase from 33.9% to 39.8%. By 2080, we predict that the mangrove coverage will decrease by about 269 km2, the waterbodies will increase by 159 km2, the built-up areas will increase by approximately 3.89 km2, and the barren areas or beaches will cover 371 km2 (3.9% of the total area). For the simulation of 2100, we predict that the mangroves will decrease by approximately 48% (4511 km2). The waterbodies will increase to 45%, while the built-up areas and barren beaches will reach up to 117 and 517 km2, respectively. There has been an increase in sea level in the Sundarbans by an average of +17.3, +9.6, + 6.1, +0.6, +1.6, and −4.01 mm/year at Khepupara, Charchanga, Hiron Point, Dladia, Diamond Harbor, and Sagar, respectively. The findings of this article can assist the Bangladesh and Indian governments and policymakers in developing optimal land use plans and implementing successful mangrove forest management and conservation plans in the future.

Keywords

• Cellular automata algorithm coupled with an artificial neural network (CA-ANN)
• land use/land cover (LULC)
• QGIS
• sea level rise
• Sundarbans delta