Assessing urban growth exposure to sea level rise using a machine learning based model with Alexandria as a case study

Abstract

Alexandria governorate, like most coastal developing regions around the world, suffers from uncontrolled urban sprawl through high-risk areas vulnerable to threats related to climate change. This could result in significant consequences such as an accelerated sea level rise, which is expected to directly impact most low-lying coastal cities worldwide. This paper examines the impact of sea level rise on existing and future predicted urban growth areas in the Alexandria governorate as a case study. Additionally, it compares the total estimated economic losses for urban areas under both scenarios. For predicting future urban areas exposed to the threat of sea level rise, a multilayer perceptron (MLP) based model was used. The results show that if urban areas continue to expand into regions at risk of sea level rise, total urban exposure will be at least doubled. This will almost triple the expected direct economic losses for urban areas by the end of the century. Using remotely sensed, open data and geographic information system-based applications for data processing, this approach can be easily transferred to other regions to aid decision making toward urban resilience.