Climate Change Adaptation and Mitigation using Deep Learning for Urban Environments

Rising global temperatures need swift action. Adaptation and mitigation strategies for climate change have been the subject of several scholarly works. To better understand climate change and develop adaptation and mitigation plans, new research methodologies are required. Due to technological advances, deep learning (also known as DL) and machine learning (ML) are being employed in more fields, including climate change. This study reviews the state-of-the-art deep learning and machine learning approaches for adapting to and preventing the effects of climate change. Another goal is to identify the most prevalent ecological mitigation and adaptation strategies in urban environments that have been studied using ML and DL approaches. This project goal is secondary. This study evaluates climate change mitigation-related applied machine learning research focusing on geographical data and satellite imagery, transportation in cities, and buildings. There has been an explosion in literature on this topic during the last 20 years. This study demonstrates how climate solution research may provide policy suggestions at the street, urban, development, along with home scales via the use of big data and machine intelligence. These suggestions for policy are tailored to specific situations yet scalable to global mitigation. This research work employs a hybrid Deep Learning technique that combines word frequency analysis with topic modeling. Specifically, it employs the Convolutional Neural Network (CNN) as well as the Gated Recurrent Unit. Finally, this study suggests that CNN-GRU machine learning is leading the pack of approaches to adapt and mitigate the climate change. When it comes to analyzing the ways to adapt and mitigate climate change, DL algorithms have been mainly used in geoengineering with surface temperature readings.