Global Flood Observatory
Coordinator: Dartmouth Flood Observatory and World Food Program
The Dartmouth Flood Observatory has been monitoring floods for over 30 years using a variety of techniques and methods, most of them developed for the purpose. The European Commission, the World Food Program, IFRC and other humanitarian actors have operational units that monitor and assess emergencies worldwide, including floods. International forecasting organizations are able to produce flood warnings with increasing lead times.
Based on guidelines developed by DFO and JRC in 2012, the partnership intends to provide structure and a technological platform to allow distributed organizations to collectively assess floods and their impact in a systematic and continuous way. Operational partners are expected to contribute in-kind with time of expert analysts.
Figure 4. Schematic representation of Global Flood Observatory
The 5 operational needs are:
· 10-30 day advanced lead time weather forecasts are increasingly reliable. These need to be translated into advanced lead time flood prediction at a dense, globally distributed array of forecast points, and whose location can be guided in part by observed flooding in the past. This will require not only global models, but also interoperability among regional models.
· Near real time, operational, flood advice capability. Latency should be under 3 hours for operational purposes. One needs to track early onset of flooding, as it occurs. Monitoring floods is also essential for improvement of the model forecasts.
· Near real time, moderate spatial resolution, high temporal resolution (daily time steps) flood mapping and immediate public data distribution in formats such as GIS that allow variety of end user access and comparisons from different systems. A combination of optical and SAR sensors is needed for robustness. Data products could include value-added versions such as a per-event total flooded area map, a flood duration map, and/or water depth maps.
· High spatial resolution flood imaging and data product dissemination, such that results can be inter-compared. Multiple data sources are required such that there is a robust capability to guarantee that the flooding WILL be imaged and mapped and data made available.
· Long term preservation of the record of inundated lands, at some standard spatial resolutions (e.g. at 250-500 m, and also, more limited coverage, at 10-50 m). We need a global public data base of such information.