Journalism and Numeracy

This is the third time I've heard of this type of error in the press, but is perhaps my favorite:

Average temperatures in the US have risen by 1.5F (-17C) over the last 50 years, the report said. Rainfall in major storms has increased 20% over the last 100 years - with the heaviest downpours in the north-east. Sea levels have risen up to eight inches along some parts of the east coast.

Emphasis added. The temperature increased by negative 17 Celsius. Perhaps that is why it's so hot these days in Texas. Er, or something.

Joe Romm picked up the Guardian story in an earlier incarnation, and without comment on this aspect approvingly quotes the following:

The final draft of today’s report uses climate models to map out starkly different futures if the current generation of Americans fails to act to reduce the carbon emissions that cause global warming.

If today’s generation acts on climate change, the average US temperature will rise 0.4C-1.83C (4-6.5F) by the end of this century, said the draft, which was finalised in April.

If it does not, average temperatures could rise by about 2.1C-4.3C (7-11F) with catastrophic consequences for human health and the economy.

Americans have already been living with evidence of changing climate, the report said. Over the last 30 years winters have grown shorter and milder, with a 2.1C (7F) rise in winter temperatures in the midwest and northern Great Plains. Hurricanes have become deadlier.

Forgive me, but I expect more attention to data from a physicist. "Tim R." points the problem out in comments at Romm's site but no correction appears.

It's an unfortunate blemish actually. Romm's article is worth reading. He raises the question of where the press is on this issue. (Pretty much absent actually, but what else is new?)