EmmanuelSAkosah/Flood-Prediction-Accra

Fine-grained Flood Prediction in Accra with a Deep Learning based Hydrologic Model ## Flood Prediction In Accra, Ghana Summary of Project Description: ### Overview We are a team of students and professionals interested in combining expertise in machine learning, weather/flood prediction and public policy to create an accurate flood prediction tool to warn locals of fatal floods. We also aim to guide public policy through the research findings we'll uncover over the course of the project. It is our hope that such an undertaking will provide early warnings to locals during flooding and put an end to Accra's annual fatal floods. ### The prediction model Current flood prediction models rely mainly on weather signals mined from weather databases while almost disregarding environmental signals. The disregard of environmental signals such as soil types and their permeability, building proximity and human-induced drainage blockages etc, can only result in coarse or inaccurate flood predictions. We aim to incorporate such signals into a deep-learning based flood prediction model specialized for the Accra Metropolitan Area. We are actively looking to expand our team and invite all interested students. All skill levels are welcome. Interested students and professionals with experience in -Machine Learning -Weather Forecasting -GIS Tools -or a passion for the project should reach us at manny.19@dartmouth.edu. See the detailed, preliminary project description here