pyaj0/near-real-time-drought-exposure

Near Real-Time Drought Exposure: Case Study of the Central Region, Ghana # near-real-time-drought-exposure Near Real-Time Drought Exposure: Case Study of the Central Region, Ghana :3rd_place_medal: 3rd place winner! zindi.africa Note: This document is the framework document for Hack the Continent Open Buildings Challenge organized by Google Research, Accra on the Zindi platform. The goal of this competition was to put Google’s Open Buildings dataset to good use. ## Overview In Africa, the frequency of natural disasters has substantially increased over the past four decades and at a faster pace than across the world (ie more than a third of the world’s droughts during 2010–2019 took place in Sub-Saharan Africa). Although droughts are less frequent compared to other natural hazards (floods, storms, ...), they had the largest human toll in terms of the number of people affected [1]. In the 2022 IPCC report, experts reaffirm that climate change is expected to dramatically increase the intensity and occurrence of extreme events in the Africa. A rise in drought occurrence would highly increase adverse impacts on water scarcity, crop production, infectious diseases, malnutrition and displacement [2]. Therefore, solutions to anticipate drought disasters become a priority in planning better humanitarian responses and executing disaster risk reduction programs efficiently. For this competition, the goal was to identify the latest locations particularly exposed to droughts using precipitation and building data. To achieve this goal: - buildings were clustered to group dense areas together, - the Standardized Precipitation Index (SPI) was calculated and - drought exposure locations were found. Ghana’s Central region was used as test area. ## Data ### Building Data Open Buildings is a large-scale open dataset that contains the location of buildings in Africa derived from high-resolution satellite imagery. More than 500 million buildings were detected on 67% of …