2011 Global Temperatures

[BethesdaWX]

And you didn't read, yet again.

[skierinvermont]

1) Don't twist things dude, of course there will always be something tiny...not the point. Total error potential on UAH is +/- 0.05C/decade at most, there is no way around it. UAH has better resolution, coverage, and in the areas where there is NO data for GISS/HADCRUT, data will NOT be +/- .04C.

2) You are the only one on this forum that believes GISS is better statistically than UAH, at least I believe so. Its laughable, because its not true.

So basically you believe Spencer and Christy when they say the statistical error on UAH is +/-.05C/decade.. but the dozens of scientists who say that surface trend error is +/-.04C/decade for a 30 year trend are just liars. Zou has a trend of .20C/decade +/-.04C/decade.. which falls outside UAH's error bars.

Neither of these error bars are complete. They include only the quantifiable error from things like satellite drift (UAH) or extrapolation (GISS). It does not include unknown methodological errors.

But for the error that is quantifiable, GISS and HadCRUT have smaller error bars than UAH.

I also don't believe I am the only one on here who believes this. I am sure Rusty believes this and I would hope after reading the summary I posted the other day and all the other studies I have posted indicating potential error in satellite temperature data that zucker, tacoman and ORH believe this too.

Moreover, this is what the large majority of scientists believe.

[BethesdaWX]

So basically you believe Spencer and Christy when they say the statistical error on UAH is +/-.05C/decade.. but the dozens of scientists who say that surface trend error is +/-.04C/decade for a 30 year trend are just liars.

Neither of these error bars are complete. They include only the quantifiable error from things like satellite drift (UAH) or extrapolation (GISS). It does not include unknown methodological errors.

What a Bunch of Bullsh*t! GISS is ignored by the majority of the climate science community, Phil Jones, IPCC, seriously??? The Majority of Scientists do not Believe GISS to be superior

Why else would you infill for the Arctic? There is No Data! Thus GISS is too warm. Where GISS measures/Wx station data, then yes. Extrapolations for thousands of miles? No way. Again, total error potential (that we are aware of), has been stated to be +/- 0.05C/decade at most by Christy and Spencer. They will likely come out with a 2011 Update this summer. Anyone can come out and say "Oh there will always be some additional error that we dont know of"...that goes for GISS/HAD too...but discussing semantics is silly.

Here........we're arguing over 0.01C/decade here! UAH is higher quality data, better resolution, more coverage, with no extrapolations...so I'm surprised you're not going with that. Not to mention Defending and using Hansen data Is that what you've come to?

Ask the Mets/Scientists on this Board what they think.

[tacoman25]

Totally stupid how you guys focus on the month to month differences between the two. I guess HadCRUT is biased because Nov to Dec GISS dropped .32 but HadCRUT dropped a mere .17. What nonsense HadCRUT is so full of **** biased warm. Oh and look how GISS cooled from Jan to Feb but HadCRUT warmed... what bull****... those ****ers.

Just get over it guys.. month to month sometimes one decreases while the other warms. It works both ways. In the long run, the divergence is almost entirely due to the polar regions, and when we infill HadCRUT with UAH poles, it is the same result as GISS.

Please. Like you don't follow the daily/monthly changes in AMSU, etc.

Months add up, and there's a reason GISS has diverged so much in recent years. There's nothing wrong with pointing out major differences that are apparent in individual months. GISS is running warmer in the Arctic than any other source now (and has been for at least a few years). The kind of ridiculous anomalies that pop up like in the March 2011 map are part of the reason why.

[tacoman25]

Yes but we find that when we take periods as short as 12 years HadCRUT 60-60 +UAH poles is closer to GISS than HadCRUT. Only by going to periods as short as 8 years do we find that not to be true. This is just the chance inherent in the extrapolation process.

In the long run we see that HadCRUT is the one diverging from HadCRUT+UAH poles (and from GISS, and from GISS+ UAH poles).

An assumption on your part. GISS continues to show increasing divergence, especially in the Arctic. That's what's going on right now.

[tacoman25]

I'm quite sure that 5, 10 years from now HadCRUT 60-60 + UAH poles will be closer to GISS than to HadCRUT even starting a trend in 2002. HadCRUT will still be lower because it misses the arctic.

And therein lies the problem. No one should be "quite sure" about such things on this topic. There is way more uncertainty than you or many scientists would like to admit. To be sure is arrogant, because we really don't understand the climate system that well...or the exact reasons we see the trends/divergences that we do.

[ORH_wxman]

An assumption on your part. GISS continues to show increasing divergence, especially in the Arctic. That's what's going on right now.

I think its going to continue to diverge over the next several years...especially since oceans are in a cooling cycle and it ignores a lot of the ocean data near the polar regions so it will extrapolate those much warmer than other sources. I think that is probably one of the reasons it was closer to other guidance in the 1980s/1990s...the oceans were in a warm phase.

[tacoman25]

Also, the 8 year divergence between GISS and HadCRUT is mostly due to between 60S-60N... in the long run those differences have always balanced out. HadCRUT warmed much faster 60S-60N from 1980-2002. GISS is just catching up.

When we look at 15-20 year trends.. there is no divergence between the two between 60S-60N. But there is a divergence due to the arctic, which is eliminated by replacing HadCRUT arctic with UAH arctic.

Only in certain time frames. Are you forgetting our discussions from a couple weeks ago? It's clear that GISS is diverging in the Arctic...no matter how much you'd like us to ignore glaring months like March 2011.

[tacoman25]

I think its going to continue to diverge over the next several years...especially since oceans are in a cooling cycle and it ignores a lot of the ocean data near the polar regions so it will extrapolate those much warmer than other sources. I think that is probably one of the reasons it was closer to other guidance in the 1980s/1990s...the oceans were in a warm phase.

Exactly. And I think that's part of the reason we saw different trends for UAH/RSS/HadCRU during that period as well. Different methodologies end up with different biases during different phases. But GISS is more off on its own now than any other one source has been since the satellite record began.

[skierinvermont]

Please. Like you don't follow the daily/monthly changes in AMSU, etc.

Months add up, and there's a reason GISS has diverged so much in recent years. There's nothing wrong with pointing out major differences that are apparent in individual months. GISS is running warmer in the Arctic than any other source now (and has been for at least a few years). The kind of ridiculous anomalies that pop up like in the March 2011 map are part of the reason why.

No it's actually not why... from 2002-present the divergence between HadCRUT and GISS is actually between 60-60 (I think mostly due to their different SST sources) Of course 1990-2002 HadCRUT warmed faster 60-60 than GISS.

From 1990-present the divergence is because HadCRUT doesn't include the arctic and GISS does. These arctic extrapolations are largely corroborated by UAH over the 20 year period (subbing in one for the other yields little change).

[tacoman25]

No it's actually not why... from 2002-present the divergence between HadCRUT and GISS is actually between 60-60 (I think mostly due to their different SST sources) Of course 1990-2002 HadCRUT warmed faster 60-60 than GISS.

From 1990-present the divergence is because HadCRUT doesn't include the arctic and GISS does. These arctic extrapolations are largely corroborated by UAH over the 20 year period (subbing in one for the other yields little change).

GISS is diverging from everyone since the early 2000s. And it can't all be explained by the Arctic warming, as we proved a couple weeks ago. Do you need me to bump a thread or two?

[skierinvermont]

An assumption on your part. GISS continues to show increasing divergence, especially in the Arctic. That's what's going on right now.

No.. the divergence the last 8 years is actually mostly due to 60-60.. not the arctic.

For the 20 year trend, the GISS arctic extrapolations are corroborated by UAH. Subbing in UAH instead of the extrapolations yields a very similar result.

[BethesdaWX]

No.. the divergence the last 8 years is actually mostly due to 60-60.. not the arctic.

For the 20 year trend, the GISS arctic extrapolations are corroborated by UAH. Subbing in UAH instead of the extrapolations yields a very similar result.

GISS has been trending away from UAH in the Arctic this past decade.

[skierinvermont]

GISS is diverging from everyone since the early 2000s. And it can't all be explained by the Arctic warming, as we proved a couple weeks ago. Do you need me to bump a thread or two?

Since 2002 it's diverging from HadCRUT between 60-60.. but we see this occur frequently throughout the record over short periods. So I am quite confident that GISS will not continue to diverge between 60-60.

It will continue to diverge globally because hadCRUT leave the arctic blank, but it won't diverge from HadCRUT+UAH poles.

[ORH_wxman]

No.. the divergence the last 8 years is actually mostly due to 60-60.. not the arctic.

For the 20 year trend, the GISS arctic extrapolations are corroborated by UAH. Subbing in UAH instead of the extrapolations yields a very similar result.

The last 8-9 years, the GISS trends in both polar regions have been warmer than UAH. They showed the southenr polar region warming while UAH had it slightly cooling and its arctic warming was much more significant than UAH.

[skierinvermont]

GISS has been trending away from UAH in the Arctic this past decade.

No it hasn't. GISS vs GISS w/ UAH poles:

[nzucker]

GISS and HadCRUT have smaller statistical error than UAH. For a 30 year trend, they have statistical errors around +/-.04C/decade.. vs UAH's +/-.05C/decade. I have posted this before.

The +/-.05C is the statistical error. It doesn't include methodological error. Some error is not quantifiable. UAH likely has large methodological unquantifiable error because we find that when we use different methods, like Fu, VG and Zou, we get much more warming.

Oh my god, that .01C/decade error in UAH is really killing me. Must be the reason that GISS is .3C warmer than UAH this month!

No.. the divergence the last 8 years is actually mostly due to 60-60.. not the arctic.

For the 20 year trend, the GISS arctic extrapolations are corroborated by UAH. Subbing in UAH instead of the extrapolations yields a very similar result.

Forget about all the arguments...just look at the GISS map. Does that map look realistic to you in any way, shape or form? That there'd be just a stripe of all red at the top of the world with no subtleties in anomalies...Especially when you compare to RSS maps?? If I posted a map of March 2011 anomalies in the United States, using just a few limited stations, and everything was exactly the same color with no nuance, would you believe it? Come on...

One can clearly see from looking at these monthly maps that the satellites have more resolution, more nuance, and thus are favored. No two ways about it.

[skierinvermont]

The last 8-9 years, the GISS trends in both polar regions have been warmer than UAH. They showed the southenr polar region warming while UAH had it slightly cooling and its arctic warming was much more significant than UAH.

Yes slightly.. but the primary reason for the divergence the last 8-9 years is between 60S and 60N. In the long run GISS and HadCRUT have never shown a disparity there (actually HadCRUT has tended to warmer slightly more). So I have no reason to believe this will continue. Probably just catching up for its much slower warming prior to 2002.

[skierinvermont]

Oh my god, that .01C/decade error in UAH is really killing me. Must be the reason that GISS is .3C warmer than UAH this month!

Forget about all the arguments...just look at the GISS map. Does that map look realistic to you in any way, shape or form?

No of course it doesn't look realistic, nor would I ever expect it to for a single month. To expect it to, is to fail to understand the method. There are lots of other maps where GISS has extrapolated too cold.

As this graph shows, subbing in UAH into GISS in the arctic makes no difference:

[skierinvermont]

Here's an idea. I'll create a graph of HadCRUT, GISS, HadCRUT 60-60 w/ UAH poles, and GISS 60-60 w/ UAH poles using 1980-1989 as the base. Now we can take guesses at which of the four is the outlier the last 5-10 years.

I'll guess HadCRUT.

Everybody guess.

[tacoman25]

Yes slightly.. but the primary reason for the divergence the last 8-9 years is between 60S and 60N. In the long run GISS and HadCRUT have never shown a disparity there (actually HadCRUT has tended to warmer slightly more). So I have no reason to believe this will continue. Probably just catching up for its much slower warming prior to 2002.

I guess I really do need to bump some threads for you.

GISS is diverging from all sources over the past 8-9 years, both globally and in the Arctic. That is all we are saying. You can claim it's not a long enough time period, or that you are VERY SURE the divergence will go away, but for now it is very real and undeniable.

And big monthly divergences do make a difference.

[BethesdaWX]

Data Manipulation FTW! Skier, I said AT THE ARCTIC, not Infilling the Globe with HADCRUT. This is GISS/UAH, not Hadley/UAH with infilling.

Why are you defending GISS when we have better data??? Its your stupid Bias that clouds your objectivity.

[ORH_wxman]

No of course it doesn't look realistic, nor would I ever expect it to for a single month. To expect it to, is to fail to understand the method. There are lots of other maps where GISS has extrapolated too cold.

As this graph shows, subbing in UAH into GISS in the arctic makes no difference:

Yes, we know about the warm phase agreement. I think GISS will continue to diverge from UAH in the polar regions over the next decade. Its clearly doing it more noticeably since the famous 2005 warmest year ever. I think the likely cause of the continued divergence over the next decade will be its lack of use of SST data wherever any ice forms during the year, and since the oceans are in a cooling phase, it will make the difference more and more glaring.

[tacoman25]

No of course it doesn't look realistic, nor would I ever expect it to for a single month. To expect it to, is to fail to understand the method. There are lots of other maps where GISS has extrapolated too cold.

As this graph shows, subbing in UAH into GISS in the arctic makes no difference:

Please show them. Every time I see a glaring difference between GISS and satellite temps, it's almost always because GISS has huge swaths of stupendous warm anomalies in the upper NH...and they are almost always more widespread or greater than the satellite anomalies, despite the overall maps matching well most other places globally.

[BethesdaWX]

Oh my god, that .01C/decade error in UAH is really killing me. Must be the reason that GISS is .3C warmer than UAH this month!

Forget about all the arguments...just look at the GISS map. Does that map look realistic to you in any way, shape or form? That there'd be just a stripe of all red at the top of the world with no subtleties in anomalies...Especially when you compare to RSS maps?? If I posted a map of March 2011 anomalies in the United States, using just a few limited stations, and everything was exactly the same color with no nuance, would you believe it? Come on...

One can clearly see from looking at these monthly maps that the satellites have more resolution, more nuance, and thus are favored. No two ways about it.

[nzucker]

No of course it doesn't look realistic, nor would I ever expect it to for a single month. To expect it to, is to fail to understand the method. There are lots of other maps where GISS has extrapolated too cold.

As this graph shows, subbing in UAH into GISS in the arctic makes no difference:

Why should I ever believe a method where a large amount of data looks to be unrealistic? The satellite maps look much, much more convincing...so I'm more inclined to believe them. You can say all you want that the GISS extrapolations balance out warm vs. cold, but they clearly aren't lately because GISS is coming in so much warmer than UAH/RSS, far beyond natural variation or the difference in warming between the troposphere and the surface would dictate. Maybe most of the faulty extrapolations have been in Africa and Central Asia the past decade, not the Arctic, but somewhere is clearly way, way off. I know the dramatic loss of stations in Africa during the 1980s is a point of controversy, and also one area in which GISS has been making large, generalized extrapolations. Maybe that was the problem in the last decade.

Also, suppose I wanted to do a project about a specific region's temperature anomalies, or how climate change is affecting the Arctic. Which map would give me more information, GISS or RSS? The answer is clear: these extrapolations are too broad-brushed and don't simulate the reality of the Earth's complex weather patterns. GISS appears to be extremely incorrect around Greenland and in some parts of Northern Asia, and watching the GFS runs and the Ryan Maue GFS anomalies during March clearly proved that even models weren't seeing the warmth that GISS is. An amateur weather hobbyist shouldn't be able to create a better map of world anomalies than NASA, but sadly that seems to be the case.

[skierinvermont]

I guess I really do need to bump some threads for you.

GISS is diverging from all sources over the past 8-9 years, both globally and in the Arctic. That is all we are saying. You can claim it's not a long enough time period, or that you are VERY SURE the divergence will go away, but for now it is very real and undeniable.

And big monthly divergences do make a difference.

The last 8-9 years it's mostly between 60-60. In the long run, there's no divergence between 60-60. A little bit has been because of the arctic warming faster than UAH over that particular period.

The last 10-20 years the difference is the arctic, and subbing in UAH for the arctic corroborates GISS over HadCRUT.

[tacoman25]

Yes, we know about the warm phase agreement. I think GISS will continue to diverge from UAH in the polar regions over the next decade. Its clearly doing it more noticeably since the famous 2005 warmest year ever. I think the likely cause of the continued divergence over the next decade will be its lack of use of SST data wherever any ice forms during the year.

Exactly. Prior to 2005, there was no year where one source had the warmest year on record, but the rest disagreed (despite different sources having different biases related to SST, extrapolation, coverage, etc). Then it happened again in 2007, and now yet again in 2010.

[tacoman25]

The last 8-9 years it's mostly between 60-60. In the long run, there's no divergence between 60-60. A little bit has been because of the arctic warming faster than UAH over that particular period.

The last 10-20 years the difference is the arctic, and subbing in UAH for the arctic corroborates GISS over HadCRUT.

But if you look globally AND include the Arctic over the last 8-9 years, an increasing divergence is seen with GISS. It's not just one or the other. That is the issue...which you believe to be a non-issue.

[skierinvermont]

The differences are small compared to the overall trend. And those are selective start and end points. If GISS arctic extrapolations were the problem, then replacing them with UAH would make GISS significantly cooler. It doesn't. Case closed.