PIRE Fellows Race Clark and Zac Flamig (meteorology PhD candidates at the University of Oklahoma) traveled to Windhoek, Namibia in the spring of 2015 to work with government officials and other local stakeholders to improve forecasting and monitoring of flood and drought impacts within the country. Namibia is a stable and prosperous middle-income country in an arid and semi-arid region with highly variable rainfall. Many citizens are heavily dependent on agriculture, which increases the need for accurate flood and drought monitoring. For all these reasons, Namibia has been an ideal place to build local capacity to effectively use hydrological models, remote sensing platforms, and cloud computing.
Here’s what Race and Zac had to say:
We have been working in Namibia for several years now and have been a part of the OSDC PIRE program since 2013. Each year we’ve conducted hydrological model training, meetings about the Namibia Flood Dashboard, and field campaigns. In 2015, the focus was on training government hydrologists in Namibia to use a new hydrological model – the Enhanced Framework for Flash Flood Forecasting (EF5). EF5 uses remotely sensed topographical data, satellite precipitation estimates, and estimated evapotranspiration to forecast stream flow anywhere in the world. Users can add forecast rainfall from a global weather model to extend the lead-time of a forecast. Where stream flow observations exist, the model can be calibrated to produce more accurate simulations. It will be possible soon to produce seasonal drought and flood forecasts based on seasonal climate forecasts from extremely long-range atmospheric models. Right now, some of these outputs are available on the Namibia Flood Dashboard, which is a website (matsu-namibiaflood.opensciencedatacloud.org) that contains a host of useful information related to flooding and water in Namibia. The Dashboard is hosted by the Open Science Data Cloud and Project Matsu and is maintained by our partners at NASA.
So where do we go from here? We fully expect to continue to be invited back to Namibia to work with their government hydrologists. They have expressed interest in using EF5 in multiple capacities across their country. On our end, we continue to develop more complex and advanced model training for use in future trips to the country. We also are exploring the use of MOOCs and remote training strategies to supplement, but not replace, in-person training activities. The Dashboard remains a critical part of our efforts, as it hosts the event images taken from various NASA satellites in past floods over Namibia. Finally, we envision using EF5’s stream flow outputs to produce flooding inundation maps, and validating these against NASA satellite observations on the Dashboard.