Satellite temperature records

Our BobW has little use for these. While I find them to present several shortcomings, they seem a very appropriate use of satellites (that are up there anyway), and Our mojo prefers them to all surface measurements despite the matter of brevity. All others might find themselves somewhere along that continuum.

They do require extensive processing which undergoes revision from time to time. To stay current see:

http://journals.ametsoc.org/doi/abs/10.1175/JCLI-D-16-0768.1

RSS full-length T trend is revised to “0.174 C/decade, ~30% larger than our previous version”. This thus becomes much more similar to surface-T trends over same interval. All surface-T trends, developed by several different ‘shops’.

No comment on this has yet appeared on Spencer’s UAH site. I gave a nudge.

 

For both RSS and UAH, TLT is weighted from very near surface to several km above. They are most often expressed as anomalies with reference to some base period, which is reasonable. But the actual proxy Ts are something in the -20 to -30 oC range (I've forgotten). That would be a very challenging life environment. A bit warmer than Greenland summit station and more warmer than south pole. People 'live' in both places but it pretty much sucks.

For me, most important distinction between sat-T and surface-T is wide swings in the former. This may jut be something that troposphere does, distinct from surface where there is so much thermal inertia in solids and seawater. Radiosonde-T data are similarly jumpy through time. I have suspected before that influence of different altitudes varies, especially during El Nino. But people who work on or professionally assess sat-T have not talked about that.

This jumpiness (from whatever cause) coupled with the short record (since 1979) places severe limits on what one can do with data. Becomes more severe if one intentionally misrepresents the record by starting at an El Nino high like 1998.

Word is now that EPA will do a red team/blue team exercise on climate studies. That, if done honestly, would stamp out 'starting at 1998'. I'd appreciate that.

 

Spencer wasn't consulted to peer review@6. How would we know this? Journal of Climate does anonymous peer review.

We might bear in mind that UAH TLT in its current version has a (full-length) trend of 0.124 oC per decade. RSS is +0.174. Both have undergone revisions through time, because satellite proxy-T are not straightforward to calculate. I would be surprised if purveyors of either RSS or UAH asserted that they are in 'final form' with no future changes anticipated.

 

Please note that UAH v6 TLT is +0.124 oC per decade. I first wrote 0.148 ad then edited it.

+++

For flying machines to measure tropospheric T. one would look for heavy coverage. There are already a lot of commercial flights. Not clear we need to invoke a new set of birds for this. Instead, log inflight T with well-calibrated equipment.

It would be a different record from satellite and surface and radiosonde. I'd not object to adding another. But I should think the main benefit would be to improve small-scale data as input to weather modeling. That is the activity crying out for such data. We have supercomputers that can do a global weather model at sub-1-km resolution, but do not know 'state variables' at that scale.

Similar is already done with in-flight CO2 measurements, but I forgot how many airplanes do that.

 

About a dozen people review each paper before publication.@8. Not for any journal I know. Three would be luxurious; two typical. J Clim does not say on their website, but I am sure editors would answer that question if asked. J Clim published 593 articles in 2015, web site says. That is actually quite a lot. They have a long list of editors.

 

OK, seems that these new birds will be for internet relays? So they will stay over land and cover <22% of earth surface. One would want to have clear goals for 'mass spec' that is not just one thing. Often with big heavy magnet.

Time of flight mass spec makes the magnet go away though - they are spiffy.

 

BTW, most recent radiosonde compilation seems to be 2005. They ought to crank out another. Their lowest-height T trend is closer to RSS than UAH. For whatever that may be worth. Bear in mind there are ~90 launch points. If a paucity of surface thermometers bothers you, this should bother you more.

But they do neat things; an easy to justify use of helium, and have an added plus of being called "RATPAC".

 

Surface T HADCRUT4 is 0.176 oC/decade since 1970. Lack of large differences in other 'data streams' is...boring. This is +T, until/unless it substantially changes.

Only relief from boredom comes when some poster says, any year now, a much stronger -T pattern will happen. We have heard such murmurs about since PriusChat began, though, so even that is tinged with boredom.

Here's RATPAC:

 

OK, Bob, choose: if you want birds so light that helium in wings 'helps', you can forget about flying mass specs or any other serious (heavy) instruments. I, instead, would put HD video on your birds and sell that. HD video from 20-25 km could have good resolution.