And it trouble me beceause of this. Take a look at people like Dano on RabbetRun. He puts
the case. Engage, ridicule or ignore.

Well since debate is over, engage is out. So you have ridicule and ignore.
Well they dont ignore so that is out. That leaves ridicule.
Which they are currently engaged in.

First you enage in debate. If that doesnt work you Ignore. If that doesnt work you ridicule.

Then what? when words fail you, then what?

Lets draw an analogy. When Bush announced that he would not talk to Iran, what did opponents conclude?
huh? they concluded that if there were no words, they there would be action.

So, is the climate debate over? no sense in talking? two tribes yapping at each other..

dunno

Students in English class were subjected to one of the more horrific practices of the day– diagraming sentences. See how this practice mutilated a cherished sentence.

Still Waiting for Greenhouse

A Lukewarm View of
Global Warming
- from Tasmania

I used it myself already in may 2007

I believe that giss also have data excluding the polar area extrapolation. and both provide maps
and gridded anomalies. the differences are small, but if you want every last edge to bloster a warmist
case you use giss, to question it you use hadcru or UAH, to be fair I suppose one would avaerage
giss and hadcru for the surface and average RSS and UAH for the troposphere

Well Boris, Tamino and I have disagreed on the hadcru versus giss before, actually on the choice of extrapolating versus living with the added uncertainty. If you believe that extrapolation give more accurate results, then that is something that you would have to prove, with a test of sorts. so for example, I would have say 1000 stations and I would hide 500 from you and say “extrapolate” and then we would see how well your extrapolator worked.

It could also be proved if GISS were to post a list of the stations they use for the extraploation steps, so that peolpe can run their own “test” of the system. CA has already proved that weight of data can be effected by shape of the area used (square vs. long/skinny). No one yet has shown what weights would look like if an “arctic ring” were used.

Also, how much of the uncertainy between GISS and HadCRU is attributed to the extrapolation, and how much to stations covered. At least GISS shows a map of area covered before the extrapolation. Anyone know if HadCru shows such a map?

Now, the giss correlations done in 1987 where done on stations in non polar regions. and the correlation figure
is .6 for the northern hemisphere. and its .5 for the SH. what is it for the polar region? one doesnt know.
one doesnt know because the area is void of stations. so you essentially have an untestble extrapolation. Using a GCM
to support the extrapolation across the pole, is not something I would do in GSMT. why? because the global temp record
is PRSENTED as an observation record. not observation plus modeling.

Reasonable folks can disagree on this, but its the responsibility of the extrapolators to demonstrate, not merely
assert, that the extrapolation gives “better” or more accurate results. that assertion needs testing. Until then
I am happy with the global warming that CRU shows and the uncertainity due to coverage is not that great

My beef with Lucia is that she claims choosing GISS–which is the highest–is automatically cherry picking.

In which case, we have no beef. What happened is this:
1) Steve mosher told me Tamino “rebutted” my argument.
2) I speculated that the “rebuttal” involved several “interesting” things, one of which happened to be using GISS only for the specific time period and analysis in question. It happens that for that specific analysis and time period, picking GISS to rebutt– when the original analysis used four available data sets, results in the highest posible trend.
3) So, I note that, in that particular case GISS used alone, ignoring the data sets used in the article actually being rebutted is deceptive.

I would note there were other issues with the rebuttal. (Changing my argument into testing the TAR when I tested the AR4 yada, yada, yada.)

That there may be circumstances where picking GISS might make sense, I would consider possible. That said, if GISS and other data sets give different results, and someone does not mention this fact, I would look askance at that analysis.

As it happens: In february no matter which data set we picked, if we tested the AR4 all sets resulted in rebuttal. Tamino was only able to make it there was no falsification through a pileup of strawmen and cherry picking that defies normal blog-rhetoric.

Fine and fair enough. The point I am making here is that skeptics often criticize GISS for having higher temperatures, but they don’t usually ask way or engage in a debate about what rally is a better analysis and why. I think extrapolation–even given the flaws that Stephen notes–is better.I will believe this when GISS is cooler than HadCRU. I’m also open to this possibility that CRU is, in fact, the better analysis. It might take some convincing, but I’m open to the possibility.

My beef with Lucia is that she claims choosing GISS–which is the highest–is automatically cherry picking. But there are significant differences that are important.

Take a look at the link in my comment 1935. Look at Russia. There is a lot of missing data in a very warm area and it is perfectly reasonable to include extrapolated temps. It’s not like there is some magical radius and then all correlation falls apart.

As you can see GISS does not have coverage for northern canada, africa, or south america. All areas with significant cooling according to the satellite data.

The above referenced map shows our point. If all those available stations were listed, and a 1200 km circle drawn around each one, we would see the REAL extent of the estimation.

Does this mean that they are estimating ARCTIC temps using stations from the northern US?

I do attempt to read as much of the literature as I can. I was referring to the fact that you often exhort people to read the literature without providing any references.

I freely admit that I am not as knowlegeable in this area as many of the people who post on this blog, although I am learning. There will be others who will come to this site who also are not experts. To merely tell them to read the literature does not help them in any way.

“Our analysis differs from others by including estimated temperatures up to 1200 km from the nearest measurement station (7). The resulting spatial extrapolations and interpolations are accurate for temperature anomalies at seasonal and longer time scales at middle and high latitudes, where the spatial scale of anomalies is set by Rossby waves (7). Thus we believe that the remarkable Arctic warmth of 2005 is real, and the inclusion of estimated arctic temperatures is the primary reason for our rank of 2005 as the warmest year. Other characteristics of our analysis method are summarized in footnote (8).”

Footnote 8 discussing the difference from the HadCRUT data set:

” One large source of differences is the attempt in the GISS method to estimate the temperature anomaly for all areas that have at least one station located within 1200 km, using weights for these stations that decrease linearly with distance from the station. At any given point the temperature anomaly estimated in this way can be substantially in error, but the increased coverage usually allows an improved estimate of the global temperature anomaly, as judged from tests made with spatially and temporally complete data sets generated by a general circulation model. However, in some cases this method can increase error by giving undue weight to one isolated station with anomalous temperature.”

The warmth of 1998 was too large and pervasive to be fully accounted for by the recent El Nino. Despite cooling in the first half of 1999, we suggest that the mean global temperature, averaged over 2-3 years, has moved to a higher level, analogous to the increase that occurred in the late 1970s.

Lucia adds: http://pubs.giss.nasa.gov/abst....._etal.html

as for the artic region GISS use stations 1200 km apart and use a GCM to fill in the missing data. I’ll get the quote
its on giss somewhere. since the gcm predicts a warming artic region, then GISS will of course be warmer than hadley.

GISS uses a larger radius from each data point, and this larger radius is supported in the literature.

Extrapolation may be justified in some cases but it does not provide new information. The coverage (a.k.a. information content) of GISS is limited by the number and location of the actual measurements.

I found these links posted in another thread to be very interesting:

http://data.giss.nasa.gov/cgi-.....mp;pol=reg
http://climate.uah.edu/

As you can see GISS does not have coverage for northern canada, africa, or south america. All areas with significant cooling according to the satellite data.

The coverage for the satellite is much larger and not subject to microsite issues.

Or you can pay $9.

If $9 is too steep, you could email the authors and politely ask for a copy. Sometimes they’ll send you one. Be sure not to accuse them of fraud!