Flood model for Kerala, India, including Climate Change projections

Over the last decade, UNEP established its Ecosystem-based Disaster Risk Reduction (Eco-DRR) portfolio to promote global awareness of the role of ecosystems in disaster risk reduction and mainstream it in the global DRR agenda. One of the study fields is the State of Kerala, India. Flood is one of the main hazards in Kerala, with almost 15% of the total area prone to floods. In August 2018, devastating floods displaced almost 1 million people and killed 350 men and women. The riverine flooding has been attributed to the excessive monsoon rainfall, which were reported to be 164% above normal levels. The lack of a hazard flooding model with recent river discharge values at high resolution undermines the capacities of the disaster management authorities to take efficient anticipatory measures. In addition, climate change projections indicate an increase in frequency and severity of rainfall events for Kerala. In order to reinforce the adaptive capacities of the Kerala State to floods a downscaled flood model that considers climate change projections was developed jointly by the CIMA Foundation and UNEP/GRID-Geneva, under the directives of UNEP Crisis Management Branch. CIMA implemented the hydrological model in the Kerala region and calibrated it according to the available observed discharge data. The hydrological model is a distributed one and uses as meteorological input: rainfall, temperature, humidity, radiation and wind velocity. Future climate change based on CORDEX experiment according to the RCP8.5 scenario. The model results provide the discharge statistics of present and of future climate. UNEP/GRID-Geneva subsequently introduced these discharge statistics, together with other data such as hydraulic morphology, land cover, historical climatology, in a hydrological model in order to produce detailed maps of the water heights according to different return periods (from 10 to 500 years). The outputs of this model are a set of twelve maps, including several return periods, currently available on the MapX "Kerala floods" workspace. Initial results show a good correlation between the extent of the flooded areas in the models and those related to the 2018 flood.