The website YouTube has made available an interview of John Hirst, who is Head of the UK Met Office (see). In this interview, John Hirst incorrectly indicates that multi-decadal climate forecasts are easier to make than seasonal forecasts. He must assume that if we could skillfully predict the magnitude of the global average surface temperature trend (which has not even been convincingly demonstrated; see), that necessarily provides skillful knowledge of the weather decades into the future to be experienced in the UK and elsewhere.
It would be nice if such the relationship between the global average surface temperature trend and regional weather extremes was so simple. However, as we have seen in the recent extreme cold and snowy weather across large areas of the Northern Hemisphere, even if the global average surface temperature anomalies (and even the lower tropospheric temperatures anomalies) are above its multi-decadal average , record cold and snow can occur. Moreover, the UK Met Office seasonal weather prediction (often called “seasonal climate prediction“) has shown a clear lack of skill. Their failures to accurately predict weather patterns for the following seasons have been notable; e.g.
Comment By The UK Met Office On Their Seasonal Weather Predictions
Comments On UK Met Office 2008/2009 Winter Forecast
2007 – Forecast by the UK Met Office To Be The Warmest Year Yet – What Is The Basis For This Claim?
One of the messages from the recent extreme weather is that only if the major circulation features can be skillfully predicted (such as ENSO; the PDO; the NAO; the AO, ect) can we expect seasonal prediction skill.
What this means is that multi-decadal model forecasts must be able to skillfully predict these circulation features. However, they have no demonstrated skill with respect to this aspect of the climate system. Indeed, since there has been only limited success with respect to seasonal predictions; e.g. see), we expect that there are no accurate multidecadal predictions when the number of forcings and feedbacks that affect the climate system are even larger than on the seasonal scale; see
Pielke, R.A., 1998: Climate prediction as an initial value problem. Bull. Amer. Meteor. Soc., 79, 2743-2746.
John Hirst should be more candid on the lack of scientific understanding that we have with respect to regional weather patterns and our ability to skillfully forecast extreme events even a month or two ahead of time. To claim that the UK Met Office can provide skillful forecasts of the likelihood of such extreme events decades from now based just on the knowledge of a subset of human climate forcings (i.e. primarily added atmospheric carbon dioxide) is a very significant misrepresentation of the science.
Climate Science: Roger Pielke Sr. 