[QUOTE=jshore]
[QUOTE=Ají de Gallina]
- Is it on average colder that it was 10 years ago? YES
[/QUOTE]
By what measure? It is true that one can get negative trendlines over some periods of time, but with huge errorbars that make it essentially meaningless. (And, in fact, at least for the HADCRUT3 record, the period for the trend computation has to be either just about exactly 11.5 years [so it includes the 1998 super El Nino but not the cooler years before it] or less than 9 years.)
[/QUOTE]
I have to go with jshore on this one. The record is too short to get statistically significant figures for the cooling. What we can say is that per HadCRUT3, there is no statistically significant warming since 1996.
The UAH Satellite record shows a bit longer period of no significant warming, in their case since 1993.
However, this does not mean cooling. All that we can say is that there is no significant warming since the early to mid 1990s.
So yes, in response to the OP’s question of
Yes, me. There has been no statistically significant warming in the last 15 years or so.
[QUOTE=jshore]
[QUOTE=Aji de Gallina]
- Have scientists been able to predict short term trend in temperature without fiddling endelessly with models? NO
[/QUOTE]
I am not sure exactly what you are saying here. In fact, short term temperature fluctuations depend sensitively on initial conditions and thus the models aren’t good at predicting them. Ironically, the longer term trend, which is dependent on more basic issues of forcing, is easier to compute than the shorter term fluctuations.
In fact, they do quite a good job hindcasting the instrumental temperature record, when both natural and anthropogenic forcings are included. See, for example, here. (Admittedly, there is considerable uncertainty in the magnitude of the anthropogenic aerosol forcing, which limits the degree to which this good fit can constrain the climate sensitivity to CO2 and hence the prediction of future warming.)
[/QUOTE]
A couple of problems with this one. First, it is absolutely no surprise that the models can hindcast the temperature trend. They are tuned to do that (what A de G calls “fiddling endlessly” above). It would be surprising if they were unable to hindcast temperature. On the other hand, they do a very poor job with say hindcasting rainfall, because they’re not tuned to do that.
Second, there is absolutely no evidence that the longer term trend is “easier to compute than the shorter term fluctuations”. This claim has been made by the modelers many times, but it flies in the face of common sense. As near as anyone can tell, weather/climate are chaotic on all timescales. There is no reason to assume that long term fluctuations are any easier to predict than short term fluctuations.
[QUOTE=jshore]
… At the same time, the warming will tend to lead to more rapid drying and hence the possibility of worse droughts. (I also believe that generally less rainfall in the U.S. Southwest is one of the more robust regional projections found in the climate models.)
[/QUOTE]
Historically, droughts have been longer and more common during cold times rather than warm. Less heat = less evaporation = less rain. The models agree with each other, it’s true … but they don’t agree with historical data on rainfall. This is not surprising, they are known to be poor at hindcasting rainfall amounts and patterns.