Horus: Pilot for a generic damage evaluation methodology based on remote-sensing data
Horus is a pilot project that explores the combination of high-resolution building inventory data from OpenStreetMap and other local sources with image-processing algorithms for the detection of earthquake damage and flood extents using remote-sensing data, along with supplementary geospatial datasets as inputs to a machine learning (ML) classification model. The ML model is trained using detailed building damage datasets from past events in a supervised learning framework, and the trained model is intended to be used to estimate the extent of damage and loss in events previously unseen by the model. The proposed framework is applied in three case study applications: March 2020 Mw5.3 Zagreb earthquake; January 2020 Mw6.4 Puerto Rico earthquake and August 2016 Louisiana floods. The GEM Risk Team is responsible for the overall coordination of the project involving multiple partners, compilation of building-level earthquake damage datasets, development of the ML models, and training and testing of the models. Duration: 2020
This pilot project, initiated and funded by the World Bank, aims to develop a framework for semi-automated damage and loss assessment due to earthquake and floods from Earth Observation (EO) data and other supplementary datasets, while also leveraging recent advances in machine-learning algorithms. These damage and loss estimates can potentially be used for transparent financial compensation and to target the distribution of resources geographically and temporally.
Collaborators: Advanced Rapid Imaging and Analysis (ARIA) team at the National Aeronautics and Space Administration’s (NASA) Jet Propulsion Laboratory (JPL) and California Institute of Technology (Caltech), the Global Earthquake Model (GEM) Foundation, JBA Risk, and the Humanitarian OpenStreetMap team (HOT) Funding partner: World Bank Group