Early warnings: Floods prediction with climate models
The Columbia Water Center is developing an initiative to predict floods events around the world by using climate models to perform seasonal predictions.
The Columbia Global Flood Initiative, a new joint initiative of Columbia Water Center, the International Research Institute for Climate & Society (IRI), CICAR (the Cooperative Institute for Climate Applications and Research) and the Center for Climate Systems Research, seeks to better understand, predict and plan for extreme floods. The project is based on the conviction that while human beings may not have direct control of where and how much rain falls (the long-term effects of human-caused global warming notwithstanding) there is a great deal more that can be done to manage the risk of extreme flooding around the world. […]
Today, scientists have a much better understanding of how the global climate works than they did even a few years ago. As a result, phenomena such as flooding—once thought of as essentially random events–are increasingly understood as the result of predicable (if complex) climate patterns.
What this means is for any given part of the world it may be possible to forecast when and where the next extreme flood will occur, anywhere from a season to a year ahead of time. Global climate patterns that can affect where and when extreme floods will occur include El Nino/La Nina-Southern Oscillation (ENSO), the Pacific Decadal Oscillation (PDO) and other “climate precursors” such as ocean temperatures, or the amount of regional snowpack. […]
The implications are vast. Understanding when and where an extreme flood is likely to occur a season or even a year ahead of time could allow everyone from policymakers to reservoir managers to emergency responders better plan for what is coming.
This group of researches have been using a top-down approach to floods frequency, where they are mainly determined by climate. And of course they are. It would be interesting, however, to check how the ability of the ecosystem to deal with high precipitation discharges has been reduced by deforestation, or more precisely, by land use and land cover change. If such relation is strong, then it would offer a way to increase insurance to flood events.
Here you can read the complete note on the State of the Planet blog:
and here you can find an interesting video of a successful application of the early warning system to flooding events in Western Africa in 2008. If you feel like diving into the literature in your Easter holidays, there is a couple of interesting papers: