Monitoring inundated infrastructure and assets with high resolution satellite imagery to enable financial protection

An estimated 80% of the 21 million people exposed to flooding annually are uninsured or underinsured, and live in developing countries (DFID, 2017, WRI, 2015). New markets, financing, technology, and social innovation will be required to meet this challenge to protect vulnerable populations. One obstacle for developing innovative financial protection is inadequate data to identify built infrastructure, locate livelihood assets and flood vulnerable populations, and monitor flood risk in rapidly urbanizing areas. For existing financial protection schemes, such as parametric of index based insurance, inaccurate relief payouts can fail to protect vulnerable populations and threaten financial solvency of insurance schemes. Physically based models and public remote sensing satellites often poorly represent flood risk in urban areas, where most people live and assets are located. Yet advances in the development of satellite sensors with very high spatial resolution and temporal cadence, and in the algorithms to extract inundation and impacted infrastructure, could begin to fill identified data gaps. This talk will showcase these advances, first presenting methods and results to map infrastructure and flood events in several African cities, including Dar es Salaam, Abidjan, Kinshasa, and Beira. Next, we will demo an open source, globally scalable, demonstrative tool to assess flooded infrastructure using data from Planet, public satellite sensors, and flood recurrence estimates from Cloud to Street. We will discuss the potential and limits of this technology to develop insurance and other forms of protection in the context of the existing markets in low and middle income countries.