Predicting future urban growth scenarios and potential urban flood exposure using Artificial Neural Network-Markov Chain model in Miami Metropolitan Area

Due to the increased coastal population growth and urbanization along with rising sea levels , more people and properties will be under increased flood risk globally. Urban growth prediction models are now used to simulate potential future urban growth scenarios but impacts of future flood risks due to sea level rise (SLR) on future development have not been studied well. Owing to its higher predictive accuracy, this study employs the Multi-layer perceptron (MLP) based Artificial Neural Network-Markov Chain (ANN-Markov) model to simulate three future urban growth scenarios (business as usual (BAU), planned growth (PG), and sustainable growth (SG)) in Miami Metropolitan Area (Miami MSA), and three SLR scenarios (1 ft, 2 ft, and 3 ft) were spatialized with the current high-risk flooding (HF) zone to delineate future flood risk areas. Then the flood risks of future urban development in each growth scenario were assessed at both regional (MSA) and local (County) scales. Results show that current land use plan (PG) slightly decreases flood risks at the regional scale but not always at the local scale compared to the without growth regulation scenario (BAU). Nevertheless, flood risks in the PG scenario are significantly higher compared to the without growth in the HF zone scenario (SG). Urban growth scenario predictions can help prepare for and understand the SLR impacts. • An MLP based ANN-Markov model was used to simulate different urban growth scenarios. • Flood risks in each growth scenario were assessed at both regional and local scales. • SLR scenarios were utilized to assess potential flood risks on future urban growth. • A regional land use planning is not always effective at the local level. • Incorporating SLR with future growth scenarios helps in flood preparedness.