Climate disasters increasing with waistlines?
Aid charity Oxfam is warning that the number of people impacted by climate-related disasters will rise 54% in the next six years, reaching 375 million.
Using data on 6,500 droughts, floods and other disasters dating back to 1980, Oxfam predicts another 133 million people will be in peril by 2015. Dealing with this will require an increase in aid spending from 2006 levels of $14.2 billion to £25 billion a year.
“Any such projection is not an exact science, but what is clear is that substantially more people may be affected by disasters in the very near future, as climate change and environmental mismanagement create a proliferation of droughts, floods and other disasters,” says the new report. “And more people will be vulnerable to them because of their poverty or location.”
In addition to those people directly impacted by disasters, others will be put at risk by climate-related conflicts when climate change exacerbates more traditional security threats, says the ‘Right to Survive’ report. The projections are made by smoothing out the extremes of the historical disaster data and fitting a straight trend-line up to 2015 (graph below).
“The world barely copes with the current level of disasters,” says Oxfam International’s executive director Jeremy Hobbs (press release). “A big increase in the numbers of people affected will overwhelm it unless there is fundamental reform of the system that puts those in need at its centre.”
The UK’s Climate and Energy Minister Joan Ruddock says, “Oxfam rightly points out that climate change is not tomorrow’s crisis and is already affecting millions of people across the world.” (AP, Daily Telegraph)
In other global warming news: it’s all the fault of fat people, according to a new study in the International Journal of Epidemiology.
Continue reading 'Climate disasters increasing with waistlines?' on Nature's The Great Beyond blog.
Comments
This looks fishy ... the confidence bounds must be derived from the smoothed data, to be so narrow, and thus the true unsmoothed confidence bounds would be much larger. Plus, a long smooth (needed to get rid of the spikes) on a short data series is fairly meaningless.
Posted by: Tom Fiddaman | April 21, 2009 06:34 PM