GISS vs CRU/RSS/UAH

[BethesdaWX]

Roy Spencer & John Christy on UAH verification, the Improvement in UAH, total error in the 95th Pecentile is +/- 0.05C either way in all regards.

Majority of Radiosonde evidence support UAH.

Finally, the much lower regression correlations for the LKS dataset regressions (0.73) versus the global NCEP–NCAR (0.93) or HadAT2 (0.92) fields suggests that the LKS radiosonde statistics, with relatively poor coverage of the earth, are not very consistent with the global satellite data. We note that, while the global temperature spatial patterns in the NCEP–NCAR fields are influenced by satellite (including MSU) data, the temperature changes over time are constrained by radiosondes (Christy et al. 2003). These results suggest that one cannot depend upon radiosonde trend profile statistics to constrain global satellite trend estimates without substantial uncertainty.

In contrast to the LKS sonde regression errors from Table 1 of about −0.07°C decade⁻¹, we find that direct site-by-site trend comparisons between the UAH satellite LT trends and sonde trends for the LT layer (Table 2) reveal median trend differences of less than 0.02°C decade⁻¹. Site comparisons eliminate spatial and temporal heterogeneities and provide the best method of independent comparison.

This suggests, at least for the lower-tropospheric layer represented by the LT profile, that much radiosonde evidence is supportive of the satellite-measured trends. Additionally, the over 150 unique radiosonde comparisons represented in Table 2 supports the result of Christy et al. (2003) in which the global UAH LT trend confidence interval (95%) was calculated as being ±0.05°C decade⁻¹

http://journals.amet...175/JTECH1840.1

[skierinvermont]

Roy Spencer & John Christy on UAH verification, the Improvement in UAH, total error in the 95th Pecentile is +/- 0.05C either way in all regards.

Majority of Radiosonde evidence support UAH.

Finally, the much lower regression correlations for the LKS dataset regressions (0.73) versus the global NCEP–NCAR (0.93) or HadAT2 (0.92) fields suggests that the LKS radiosonde statistics, with relatively poor coverage of the earth, are not very consistent with the global satellite data. We note that, while the global temperature spatial patterns in the NCEP–NCAR fields are influenced by satellite (including MSU) data, the temperature changes over time are constrained by radiosondes (Christy et al. 2003). These results suggest that one cannot depend upon radiosonde trend profile statistics to constrain global satellite trend estimates without substantial uncertainty.

In contrast to the LKS sonde regression errors from Table 1 of about −0.07°C decade⁻¹, we find that direct site-by-site trend comparisons between the UAH satellite LT trends and sonde trends for the LT layer (Table 2) reveal median trend differences of less than 0.02°C decade⁻¹. Site comparisons eliminate spatial and temporal heterogeneities and provide the best method of independent comparison.

This suggests, at least for the lower-tropospheric layer represented by the LT profile, that much radiosonde evidence is supportive of the satellite-measured trends. Additionally, the over 150 unique radiosonde comparisons represented in Table 2 supports the result of Christy et al. (2003) in which the global UAH LT trend confidence interval (95%) was calculated as being ±0.05°C decade⁻¹

http://journals.amet...175/JTECH1840.1

Old study which uses the old HadAT Radiosonde data, which has since been replaced by RICH and RAOBCORE which show somewhat more warming globally, but also a significantly different spatial pattern.

Here is a comparison of the tropical TLT trends for HadAT, RICH, RAOBCORE vs RSS and UAH. As you can see, UAH was in decent agreement with HadAT (which is what Spencer and Christy used in the above old study you posted)... not so at all with RICH and RAOBCORE. RAOBCORE and RICH show nearly 50% more tropical TLT warming than UAH. For TMT (last image) the differences are even more stark. Newer radisonde data is essentially a compromise between RSS and STAR... with UAH being way off on its own especially for TMT.

[BethesdaWX]

Old study which uses the old HadAT Radiosonde data, which has since been replaced by RICH and RAOBCORE which show somewhat more warming globally, but also a significantly different spatial pattern.

Here is a comparison of the tropical TLT trends for HadAT, RICH, RAOBCORE vs RSS and UAH. As you can see, UAH was in decent agreement with HadAT (which is what Spencer and Christy used in the above old study you posted)... not so at all with RICH and RAOBCORE. RAOBCORE and RICH show nearly 50% more tropical TLT warming than UAH. For TMT (last image) the differences are even more stark. Newer radisonde data is essentially a compromise between RSS and STAR... with UAH being way off on its own especially for TMT.

??

That based on assuption of warm bias in Radiosondes in the 80's/90's, and is unrelated to UAH satellites potential error of +/- 0.05C......and nothing has been replaced.

2010 tropics

http://hurricane.atm...styetal2010.pdf (reflects on RICH, RAOBCORE, HadAT, RATPAC)

RAOBCORE error, this also plagues STAR, and is a reason for error. I also emailed Dr. Roy Spencer on the issue, and if he responds, will post it here.

A problematic issue impacts RAOBCORE and RICH and is related to a warming shift in 1991 of the upper troposphere in the ERA-40 Reanalyses on which the two datasets rely. This was shown in [19] to be likely spurious due to a mishandling of a change in an infrared channel, a diagnosis acknowledged by the ECMWF (see also ECMWF Newsletter No. 119, Spring 2009). The shift also led to a sudden and spurious increase in estimates of (a) tropical rainfall, ( 200 hPa divergence and © low-level humidity. Since this has direct influence on RAOBCORE and to a lesser extent RICH, we would expect these products to display warmer-than-actual trends especially for TMT (which is shown in [9].)

UAH also uses a Non-Drifting Spacecraft

[skierinvermont]

??

That based on assuption of warm bias in Radiosondes in the 80's/90's, and is unrelated to UAH satellites potential error of +/- 0.05C......and nothing has been replaced.

2010 tropics

http://hurricane.atm...styetal2010.pdf

I am sorry I am not allowed to read this

It uses ERA for verification purposes and Zucker tells me that anybody that does that is a hypocrite.

In all seriousness.. it looks like a good read... I will dig into it.. I wasn't aware that Christy had updated their verification (which was originally done with HadAT) or that responded to the criticisms being leveled based on RAOBCORE and RICH that had been made in some of the articles I have cited (such as Thorne 2010). I'm not surprised that Christy is still arguing his case... although I think he is still in the minority.

As I've said before.. this is an area of active disagreement and all of the methods should be considered (except of course those with demonstrated errors).

[BethesdaWX]

I am sorry I am not allowed to read this

It uses ERA for verification purposes and Zucker tells me that anybody that does that is a hypocrite.

In all seriousness.. it looks like a good read... I will dig into it.. I wasn't aware that Christy had updated their verification (which was originally done with HadAT) or that responded to the criticisms being leveled based on RAOBCORE and RICH that had been made in some of the articles I have cited (such as Thorne 2010). I'm not surprised that Christy is still arguing his case... although I think he is still in the minority.

The issue in RAOBCORE with the warm bias, has been corrected for, otherwise the Entire UAH data-record would suffer That would suck. This is why all 4 data collection methods are compared & based off eachother.

The Methods used on STAR are homogenization methods to the raw data (the raw data matches UAH). Not sure why STAR does it since they are literally making themselves an outlier when they originally patched perfectly, STAR is a relatively small scale project.....the problem is they factor the underlying trend is factored into the adjustements, since the assumed trend is dictated valid beforehand.

(same paper)

The difficulty that arises is that the recommended adjustments are typically of the same order of magnitude as the underlying trend and, in one case, larger than the underlying trend, such that the sign of the adjusted trend is different from the raw trend. First here is a figure showing the net adjustments for the tropics in deg C for the 4 levels (going high to low). In each case, the adjustments are implemented primarily in the 1985-2000 period, so one is not dealing with the far past. All records end in 2006 are not fully up-to-date.

Figure 1. RAOBCORE (tropics) adjustments for 4 levels 1957-2006. Black – midnight; blue- noon.

Next here is a figure showing the original and RAOBCORE 1.4 trends for the tropics for the 4 levels (version 1.2 is not shown). The sign in the MSU3 level is reversed by the adjustment process.

For completeness, here are plots showing the original and adjusted versions for the 4 levels.

It is evident from the above plots that the RAOBCORE adjustments are the same order of magnitude as the trend that people are seeking to determine.

[skierinvermont]

None of this addresses how GISS can be more correct by extrapolating temperatures, nor does it address the potential error where GISS is consistently warmer than Hadley in the 50-70 bands. Hadley is actually slightly warmer than GISS in the roughly 20S to 50S band and 25N to 40N band, but the only spots where GISS is warmer is the polar regions, even in areas outside of significant extrapolation. Its starting to get nitpicky and slightly off tangent argument to discuss just how much data Hadley has vs GISS in the further north bands. They are still cooler up there in regions where they measure. That can be discussed in a different argument, but the bottom line is GISS shows warming in the southern polar regions where the 3 other operational agencies show cooling...and their warming in the northern polar regions is more amplified than the others.

One thing I find interesting is that from all the data we have, the polar regions cooled below normal in the 1950s-1970s time range. Yet, we don't see GISS cooling more than Hadley in that time range. If GISS is extrapolating areas that should be warmer now, then why were they not cooler back when those said regions in question were colder? It seems the arctic cooled faster than the middle of the globe.

They consistently are warmer going back to that time range as well, and if the main crux of the argument is "extrapolation", it seems to lose its consistency before the most recent years. Nevermind the actual validity of the regions in question.

I guess the main question is (to avoid nitpicking little details and going on a tangent)....why should GISS be believed over other agencies when they are trying paint over void areas and the areas closest to those void areas are already in question to begin with comparing the different sources?

Well as I said .. I think much of the discrepancy in the 50N-70N latitude bands is because HadCRUT has blank cells there as well which are being included in the analysis.. basically the same thing that is going on 70-90N where HadCRUT is basically 100% blank cells.

Given how similar they are basically from 50S to 50N... and the fact that GISS using a HadCRUT mask agrees with HadCRUT.. it is logically impossible that GISS run warmer 50N-70N than HadCRUT for where there is coverage.

1) GISS and HadCRUT basically agree 50S to 50N

2) GISS w/ HadCRUT mask aagrees with HadCRUT globally

3) Ergo, where there is HadCRUT coverage 50N-70N and 50S to 90S ... GISS can't run warmer.

Agreed?

Anyways... this is really besides the point.. we know that HadCRUT and GISS on average, agree for the areas they both have coverage (see the figures I posted). Ergo, the reason the two differ is 100% attributable to what is going on in the cells that HadCRUT leaves blank.

We know from a variety of sources (UAH, NCEP, ERA) that the areas left blank by HadCRUT have warmed rapidly over the last 15 years. HadCRUT must therefore underestimate the warming of the last 15 years.

I would be more than happy to scrap the GISS extrapolations and just use HadCRUT+UAH infilling of the blank squares. This would yield a temperature trend over the last 15 years roughly similar to GISS's.

[tacoman25]

I would be more than happy to scrap the GISS extrapolations and just use HadCRUT+UAH infilling of the blank squares. This would yield a temperature trend over the last 15 years roughly similar to GISS's.

This sounds reasonable. Although, again, it seems odd that you are willing to use a data source (UAH) when it fits your point, but tear it to shreds when it does not. We all know the many errors you believe are inherent in UAH, and you have basically dismissed it at times when it does not fit your agenda. In this case, you seem perfectly satisfied with UAH. Just odd.

[skierinvermont]

This sounds reasonable. Although, again, it seems odd that you are willing to use a data source (UAH) when it fits your point, but tear it to shreds when it does not. We all know the many errors you believe are inherent in UAH, and you have basically dismissed it at times when it does not fit your agenda. In this case, you seem perfectly satisfied with UAH. Just odd.

The magnitude of the trend in the blank HadCRUT grids is so large on UAH, it doesn't require high accuracy or precision to detect.

If UAH is off globally it is off by a few hundredths C/decade. The magnitude of the trend in the arctic is nearly a full 1C/decade since 1995 on UAH. Nearly two orders of magnitude larger than the potential error.

[tacoman25]

The magnitude of the trend in the blank HadCRUT grids is so large on UAH, it doesn't require high accuracy or precision to detect.

If UAH is off globally it is off by a few hundredths C/decade. The magnitude of the trend in the arctic is nearly a full 1C/decade since 1995 on UAH. Nearly two orders of magnitude larger than the potential error.

But you have used the argument before that UAH or RSS cannot accurately be compared to the land sources, because they measure the atmosphere - not the same thing. But now you are saying it is fine to fill in blank spaces on land with satellite data? This is inconsistent logic.

[skierinvermont]

But you have used the argument before that UAH or RSS cannot accurately be compared to the land sources, because they measure the atmosphere - not the same thing. But now you are saying it is fine to fill in blank spaces on land with satellite data? This is inconsistent logic.

I said that in response to people who were saying that GISS is bad and should be scrapped and/or 'corrected' by the satellites. GISS shouldn't be 'corrected' by the satellites. It should continue to exist as an independent methodology which relies entirely on surface data. That, in and of itself, is important. That is the purpose of GISS. We want to be able to detect subtle differences between surface trends and tropospheric trends in the long run. By 'correcting' GISS using satellite data, they cease to be independent.

This situation is quite different... I am not advocating a 'correction' of GISS... I am not suggesting that GISS is wrong based on the satellites. Even if we decide that HadCRUT + UAH infilling is probably the best estimate of surface trends, GISS should continue to be used as well as an independent methodology.

So in short: You can't get rid of or correct GISS based on the satellites because, as you quoted me "they are not the same thing."

I stand by that statement. They are not the same thing and you can't correct one based on the other. But you can, as an alternative to GISS, infill HadCRUT with UAH in the areas data is missing in an attempt to form a reasonable estimate of the surface trend. Each method would have its advantages and disadvantages. GISS would have the advantage of being based 100% on the surface, HadCRUT+UAH infilling would have the advantage of not using large extrapolations and yet giving a reasonable estimate of temperature trends in the data blank regions while the rest of the data set was still comprised of surface data.

In the end, it doesn't matter whether we use GISS or HadCRUT+UAH infilling because they show essentially the same thing.

[ORH_wxman]

I said that in response to people who were saying that GISS is bad and should be scrapped and/or 'corrected' by the satellites. GISS shouldn't be 'corrected' by the satellites. It should continue to exist as an independent methodology which relies entirely on surface data. That, in and of itself, is important. That is the purpose of GISS. We want to be able to detect subtle differences between surface trends and tropospheric trends in the long run. By 'correcting' GISS using satellite data, they cease to be independent.

This situation is quite different... I am not advocating a 'correction' of GISS... I am not suggesting that GISS is wrong based on the satellites. Even if we decide that HadCRUT + UAH infilling is probably the best estimate of surface trends, GISS should continue to be used as well as an independent methodology.

So in short: You can't get rid of or correct GISS based on the satellites because, as you quoted me "they are not the same thing."

I stand by that statement. They are not the same thing and you can't correct one based on the other. But you can, as an alternative to GISS, infill HadCRUT with UAH in the areas data is missing in an attempt to form a reasonable estimate of the surface trend. Each method would have its advantages and disadvantages. GISS would have the advantage of being based 100% on the surface, HadCRUT+UAH infilling would have the advantage of not using large extrapolations and yet giving a reasonable estimate of temperature trends in the data blank regions while the rest of the data set was still comprised of surface data.

In the end, it doesn't matter whether we use GISS or HadCRUT+UAH infilling because they show essentially the same thing.

What about the cooling near the southern polar regions (esp 70S onward)? When comparing divergence, one of the differences since the start of the decade is cooling near the southern polar regions. It hasn't been as great as the arctic rise since '01, but its not insignificant and would offset a chunk of the arctic trend in that time.

GISS has the southern polar regions warming, the other agencies have it cooling.

I suppose you'll come up with an explanation on how GISS is correct here too. None of this so far (whether you want to substitute UAH for blank CRU or not) has explained though on how GISS is more accurate with their extrapolation of temperatures over sea ice and ocean partially covered by sea ice by using land based obs hundreds of miles away.

[skierinvermont]

What about the cooling near the southern polar regions (esp 70S onward)? When comparing divergence, one of the differences since the start of the decade is cooling near the southern polar regions. It hasn't been as great as the arctic rise since '01, but its not insignificant and would offset a chunk of the arctic trend in that time.

GISS has the southern polar regions warming, the other agencies have it cooling.

I suppose you'll come up with an explanation on how GISS is correct here too. None of this so far (whether you want to substitute UAH for blank CRU or not) has explained though on how GISS is more accurate with their extrapolation of temperatures over sea ice and ocean partially covered by sea ice by using land based obs hundreds of miles away.

It appears to me from the maps that HadCRUT leaves most of Antarctica and parts of the surrounding waters blank. That would explain why the GISS trend is greater than HadCRUT for that region.

The only thing that I can't really explain based on HadCRUT leaving it blank is the sharp divergence right around 62S. One thing I would like to note is that whoever made that WUWT map (in addition to being very deceptive by comparing GISS coverage to HadCRUT non-coverage) appears to only have done a value for once every 6 degrees or something. It's a very coarse graph. So we have no way of knowing exactly how wide or thin that divergence might be. It appears to me that the 70S and 58S data points show agreement, but there is a sharp spike at the 64S data point. If the graph is coarse... then we don't really know what's going on at 66S or 60S.

This divergence right around 62S may be related to more lack of coverage on HadCRUT... HadCRUT does have a big area of non coverage nearby but I can't quite tell if it makes it up to 62S. It may also be related to GISS extrapolating the rapid warming on the Antarctic Peninsula over the nearby ocean, while HadCRUT doesn't.

Anyways.. I really don't like using that WUWT graphic because the lack of coverage on HadCRUT artificially creates a divergence. What we really need is something which compared GISS and HadCRUT in the areas that have coverage.

I provided such a comparison that for the global average, but you seem to want to conduct a comparison between GISS and HadCRUT regionally. I'll try and find something like that.

However, differences in spatial patterns of warming, while interesting, don't concern me greatly if the global trend for the areas which have coverage is the same. Which it is. They have the same trend 1995-present for the areas which have coverage. So again it comes down to the areas that don't have coverage and if the GISS extrapolations into those areas (primarily the arctic and antarctic) are correct. We can use UAH to infill this lack of coverage on HadCRUT as an independent check. The two methods yield the same result, corroborating GISS.

[ORH_wxman]

It appears to me from the maps that HadCRUT leaves most of Antarctica and parts of the surrounding waters blank. That would explain why the GISS trend is greater than HadCRUT for that region.

The only thing that I can't really explain based on HadCRUT leaving it blank is the sharp divergence right around 62S. One thing I would like to note is that whoever made that WUWT map (in addition to being very deceptive by comparing GISS coverage to HadCRUT non-coverage) appears to only have done a value for once every 6 degrees or something. It's a very coarse graph. So we have no way of knowing exactly how wide or thin that divergence might be. It appears to me that the 70S and 58S data points show agreement, but there is a sharp spike at the 64S data point. If the graph is coarse... then we don't really know what's going on at 66S or 60S.

This divergence right around 62S may be related to more lack of coverage on HadCRUT... HadCRUT does have a big area of non coverage nearby but I can't quite tell if it makes it up to 62S. It may also be related to GISS extrapolating the rapid warming on the Antarctic Peninsula over the nearby ocean, while HadCRUT doesn't.

Anyways.. I really don't like using that WUWT graphic because the lack of coverage on HadCRUT artificially creates a divergence. What we really need is something which compared GISS and HadCRUT in the areas that have coverage.

I provided such a comparison that for the global average, but you seem to want to conduct a comparison between GISS and HadCRUT regionally. I'll try and find something like that.

However, differences in spatial patterns of warming, while interesting, don't concern me greatly if the global trend for the areas which have coverage is the same. Which it is. They have the same trend 1995-present for the areas which have coverage. So again it comes down to the areas that don't have coverage and if the GISS extrapolations into those areas (primarily the arctic and antarctic) are correct. We can use UAH to infill this lack of coverage on HadCRUT as an independent check. The two methods yield the same result, corroborating GISS.

Looking at UAH arctic vs antarctic trends this past decade (where the CRU divergence starts significantly...2001), it does not appear that substituting UAH for missing CRU explains this divergence problem with GISS. You are using UAH's arctic trend since 1995...however, a majority of that increase occurred prior to 2001 before the divergence becomes significant. I'm trying to explain divergence between the two and doing that is not explained by simply replacing UAH data with CRU. It has a slight cooling trend in the south and a warming trend in the north (but not nearly the warming trend you state when you picked 1995 as the start point). Despite this, the divergence continues to get greater the further into the 2000s we get.

GISS is on its own since the beginning of this past decade.

[skierinvermont]

Looking at UAH arctic vs antarctic trends this past decade (where the CRU divergence starts significantly...2001), it does not appear that substituting UAH for missing CRU explains this divergence problem with GISS. You are using UAH's arctic trend since 1995...however, a majority of that increase occurred prior to 2001 before the divergence becomes significant. I'm trying to explain divergence between the two and doing that is not explained by simply replacing UAH data with CRU. It has a slight cooling trend in the south and a warming trend in the north (but not nearly the warming trend you state when you picked 1995 as the start point). Despite this, the divergence continues to get greater the further into the 2000s we get.

GISS is on its own since the beginning of this past decade.

It's arguable when the divergence begins. In 95, and 96 GISS ran a fair amount warmer than HadCRUT. Then in 97,98,99,00 it was basically the same.. then it comes back with a vengeance 2001-present. It's possible that the divergence started in 1995 but something else masked it in 97-00. The divergence would still have been small at that point and easily masked by other factors or just pure luck (they do have error bars for single years of +/-.05C I believe.. it might be .1C I forget).

It also appears to me that nearly half of the arctic warming took place after 2001. And that the Antarctic actually warmed 2001-present. (UAH).

I would not expect the divergence and the arctic warming to line up perfectly. There are of course other factors at play. But over the period 1995-present UAH infilling of HadCRUT removes most of the divergence with GISS.

[Clifford]

I would not expect the divergence and the arctic warming to line up perfectly. There are of course other factors at play. But over the period 1995-present UAH infilling of HadCRUT removes most of the divergence with GISS.

The problem is that when you start with a broken data set, and patch it up.

You end up with a patched broken data set.

Is 1934 the hottest year in the USA?

I don't know... it is at least close.

What is obvious is that 1930 to 1940 probably wasn't the hottest decade in the USA.

http://www.globalcha...ents/us-impacts

What is obvious about looking at the past temperature records is that the year to year variability was on the order of 1°F to 2°F (½°C to 1°C).

Then suddenly around 1998, the year to year variability just vanished, and we had a decade of the hottest temperatures, and least year to year temperature variability in the entire US Temperature record.

In many senses, that is a good sign.

However, if they have changed their temperature recording methods to something that reduced the year to year variability, then the new data is no longer comparable to the old data.

We don't know if the new recording methodology would have caused 1935 to look just as hot as 1934, and thus the 30's to look like the 2000's.

It is best to keep the two data sets completely independent.

The old "high/low" data for comparisons of long-term temperature trends.

The new satellite imagery for current day to day data and analysis.

Mixing the data sets, and one ends up with a horrendous mess!!!

[ORH_wxman]

It's arguable when the divergence begins. In 95, and 96 GISS ran a fair amount warmer than HadCRUT. Then in 97,98,99,00 it was basically the same.. then it comes back with a vengeance 2001-present. It's possible that the divergence started in 1995 but something else masked it in 97-00. The divergence would still have been small at that point and easily masked by other factors or just pure luck (they do have error bars for single years of +/-.05C I believe.. it might be .1C I forget).

It also appears to me that nearly half of the arctic warming took place after 2001. And that the Antarctic actually warmed 2001-present. (UAH).

I would not expect the divergence and the arctic warming to line up perfectly. There are of course other factors at play. But over the period 1995-present UAH infilling of HadCRUT removes most of the divergence with GISS.

You can sort of make it fit if you are using UAH's big arctic jump from 1995-2001, however, since then we've seen arctic temperatures become flatter on UAH and anartcica temperatures have decreased after a jump up in 2001. It would make more sense if the divergence was lessening, but it is doing the exact opposite. The divergence between GISS and CRU has gotten worse despite UAH's flattening of arctic temps (to only slightly increasing) and antarctic temps now decreasing if you are going to use them to plug in the missing data for CRU.

As I said before...GISS has pretty much been on its own the past decade...and its been getting worse in the more recent years.

[skierinvermont]

The divergence would only decrease if the arctic temperatures actually dropped back towards their 1995 values. It also looks to me like Antarctic temperatures have slightly increased, so there would be little effect there. If arctic temperatures continue to increase (even slowly) the divergence will continue to grow.

As long as HadCRUT continues to leave a part of the earth blank that has warmed rapidly, it will run colder than GISS.

[BethesdaWX]

Antarctic Temperatures have been cooling for the Past 30 years via UAH (higher quality), which dominates surface data and extrapolations. Data Quality from UAH on the Antarctic is almost certainly Better than Surface data.

Everyone shows cooling down there Except for GISS.

1) We'd need to put UAH Antarctic Data to fill in GISS's Antarctic Gaps

2) We'd need to put UAH Data into Africa, The Arctic, and wherever GISS has no data.

That would really lower GISS's anomaly.

[ORH_wxman]

The divergence would only decrease if the arctic temperatures actually dropped back towards their 1995 values. It also looks to me like Antarctic temperatures have slightly increased, so there would be little effect there. If arctic temperatures continue to increase (even slowly) the divergence will continue to grow.

As long as HadCRUT continues to leave a part of the earth blank that has warmed rapidly, it will run colder than GISS.

Antarctica has been cooling for the better part of this last decade...from about 2002 onward. The divergence should not keep getting worse as we progress through the decade. GISS is on its own and its because they extrapolate exceptionally warm in areas they mask.

[tacoman25]

I said that in response to people who were saying that GISS is bad and should be scrapped and/or 'corrected' by the satellites. GISS shouldn't be 'corrected' by the satellites. It should continue to exist as an independent methodology which relies entirely on surface data. That, in and of itself, is important. That is the purpose of GISS. We want to be able to detect subtle differences between surface trends and tropospheric trends in the long run. By 'correcting' GISS using satellite data, they cease to be independent.

This situation is quite different... I am not advocating a 'correction' of GISS... I am not suggesting that GISS is wrong based on the satellites. Even if we decide that HadCRUT + UAH infilling is probably the best estimate of surface trends, GISS should continue to be used as well as an independent methodology.

So in short: You can't get rid of or correct GISS based on the satellites because, as you quoted me "they are not the same thing."

I stand by that statement. They are not the same thing and you can't correct one based on the other. But you can, as an alternative to GISS, infill HadCRUT with UAH in the areas data is missing in an attempt to form a reasonable estimate of the surface trend. Each method would have its advantages and disadvantages. GISS would have the advantage of being based 100% on the surface, HadCRUT+UAH infilling would have the advantage of not using large extrapolations and yet giving a reasonable estimate of temperature trends in the data blank regions while the rest of the data set was still comprised of surface data.

In the end, it doesn't matter whether we use GISS or HadCRUT+UAH infilling because they show essentially the same thing.

Lawyer talk. You clearly have rejected comparing UAH to surface sources on a number of occasions, but now you're doing the same thing to make your point. As long as it shows the same thing you're looking for, you're perfectly ok with it....inconsistent logic.

[tacoman25]

Antarctica has been cooling for the better part of this last decade...from about 2002 onward. The divergence should not keep getting worse as we progress through the decade. GISS is on its own and its because they extrapolate exceptionally warm in areas they mask.

Exactly. I don't understand why skiier continues to deny the obvious divergence the last decade from GISS, and attribute it entirely to Arctic warming.

[ORH_wxman]

Exactly. I don't understand why skiier continues to deny the obvious divergence the last decade from GISS, and attribute it entirely to Arctic warming.

The arctic warming has gotten pretty flat since '03 and the Antarctic has clearly cooled. There's zero reason for the divergence increase other than GISS going way warm in their "masking" areas...it has gotten worse and worse as we go through the last 9-10 years.

I think there is a reason that people like Phil Jones (who is a big AGW advocate) would say that we have been flat lining for awhile. He's not unaware of GISS. He isn't stupid. There's a lot of pretty smart people in the AGW camp that simply believe GISS is flawed and don't follow their trends or at least are cautious of using them as a primary source.

Skier isn't one of them obviously, and that's ok, but there comes a point where you can just say that a lot of people do not agree with GISS. It doesn't mean it's wrong, but there's obviously some very glaring questions about them...especially in the past decade which is what this has been all about. GISS is the only agency that shows statistically significant warming during that time.

[skierinvermont]

Lawyer talk. You clearly have rejected comparing UAH to surface sources on a number of occasions, but now you're doing the same thing to make your point. As long as it shows the same thing you're looking for, you're perfectly ok with it....inconsistent logic.

It's not inconsistent at all. Zucker wanted to 'correct' GISS across the entire globe using UAH which shouldn't be done because they are 'not the same thing' and defeats the purpose of GISS as an independent surface-only data source. It would turn GISS into UAH. The surface data would be entirely obsolete if it were to be corrected on the basis of satellite data.

What I have suggested is not at all comparable. What I suggested was forming an alternative measure which could be used in conjunction with GISS, namely HadCRUT + UAH infilling. It is not the same thing as 'correcting' GISS.

It's not lawyer talk. I am sorry if you do not see the very clear and important difference here.

[VAwxman]

The arctic warming has gotten pretty flat since '03 and the Antarctic has clearly cooled. There's zero reason for the divergence increase other than GISS going way warm in their "masking" areas...it has gotten worse and worse as we go through the last 9-10 years.

I think there is a reason that people like Phil Jones (who is a big AGW advocate) would say that we have been flat lining for awhile. He's not unaware of GISS. He isn't stupid. There's a lot of pretty smart people in the AGW camp that simply believe GISS is flawed and don't follow their trends or at least are cautious of using them as a primary source.

Skier isn't one of them obviously, and that's ok, but there comes a point where you can just say that a lot of people do not agree with GISS. It doesn't mean it's wrong, but there's obviously some very glaring questions about them...especially in the past decade which is what this has been all about. GISS is the only agency that shows statistically significant warming during that time.

Well I don't need to make a thought-out post here since your post hits the main points.

[skierinvermont]

Antarctic Temperatures have been cooling for the Past 30 years via UAH (higher quality), which dominates surface data and extrapolations. Data Quality from UAH on the Antarctic is almost certainly Better than Surface data.

Everyone shows cooling down there Except for GISS.

1) We'd need to put UAH Antarctic Data to fill in GISS's Antarctic Gaps

2) We'd need to put UAH Data into Africa, The Arctic, and wherever GISS has no data.

That would really lower GISS's anomaly.

Except HadCRUT already leaves those areas blank. Substituting in UAH data to fill those blanks on HadCRUT would not change the anomaly. On the other hand substituting in extremely rapid warming of nearly 1C/decade since 1995 in the arctic into the HadCRUT blanks would change the anomaly.

I also reject the premise that UAH is the best source of Antarctic temperature trends, but that is a separate debate. I would be willing to use UAH to infill in the Antarctic on HadCRUT for the sake of consistency and because it is generally correct (even if it is not exactly correct). However, it would make no difference because you would be substituting blank data with no-trend data.

Also, as I've pointed out, UAH actually shows Antarctic warming of at least .1C/decade since 1995. I can perform the trend analysis on this if anybody wants to challenge that...

[skierinvermont]

Antarctica has been cooling for the better part of this last decade...from about 2002 onward. The divergence should not keep getting worse as we progress through the decade. GISS is on its own and its because they extrapolate exceptionally warm in areas they mask.

Think about this ... the anomalies in any given year have a fair amount of luck involved in them. Each source has substantial error bars for a given year. In 2008 for example, the GISS anomaly was not that much larger than HadCRUT's (a mere .03C). In 2007, on the other hand, GISS was .1C warmer. There's a lot of luck involved in any given year. As arctic and antarctic temperatures warmed 1995-present, the chance that GISS would finish higher than HadCRUT for a given year has increased.

We should not expect a perfect correlation between arctic and antarctic temperatures and the GISS-HadCRUT divergence. There are other factors involved.

Nevertheless we notice a good general correlation. As arctic and antarctic temperatures have warmed, the divergence has generally grown. The warmer arctic and antarctic temperatures are, the bigger the divergence will be (ON AVERAGE). You keep repeating that arctic temperatures haven't really increased since 2001 (they have still been increasing just not as fast). Well neither has the divergence. 2 of the 3 largest divergences were observed in 2002 and 2003.

Also antarctic temperatures warmed nearly .5C from 2001-2003 which does coincide with a big jump in the divergence.

We also KNOW that it must be the arctic and antarctic causing the divergence, because they both agree on the rest of the globe. GISS has been diverging because it shows rapid warming in the arctic. If you take that away from GISS, it agrees with HadCRUT. We know for a 100% undeniable fact that the divergence is due to the arctic and antarctic because if you take them away, it agrees with HadCRUT. The question then becomes has the trend of its extrapolations in those regions been approximately correct. And the answer is, YES.

Like I said, we should not expect a perfect correlation between arctic temperature and the divergence. What I would generally expect is when the arctic is hot, the divergence will often be large. And that's exactly what we observe. When the arctic got hot (2001-2010) the divergence was usually large. To expect a better correlation than that doesn't make any sense given all the other factors involved.

[tacoman25]

It's not inconsistent at all. Zucker wanted to 'correct' GISS across the entire globe using UAH which shouldn't be done because they are 'not the same thing' and defeats the purpose of GISS as an independent surface-only data source. It would turn GISS into UAH. The surface data would be entirely obsolete if it were to be corrected on the basis of satellite data.

What I have suggested is not at all comparable. What I suggested was forming an alternative measure which could be used in conjunction with GISS, namely HadCRUT + UAH infilling. It is not the same thing as 'correcting' GISS.

It's not lawyer talk. I am sorry if you do not see the very clear and important difference here.

Actually, I believe it was suggested that comparisons to RSS (not UAH) maps indicated GISS was too warm in some areas, but you rejected that at the time because you said one couldn't accurately compare surface maps to satellite maps. So I don't see how you think it is ok to fill in some surface areas with UAH. Your base premise is that satellite temps cannot be directly applied to the surface.

[tacoman25]

Think about this ... the anomalies in any given year have a fair amount of luck involved in them. Each source has substantial error bars for a given year. In 2008 for example, the GISS anomaly was not that much larger than HadCRUT's (a mere .03C). In 2007, on the other hand, GISS was .1C warmer. There's a lot of luck involved in any given year. As arctic and antarctic temperatures warmed 1995-present, the chance that GISS would finish higher than HadCRUT for a given year has increased.

We should not expect a perfect correlation between arctic and antarctic temperatures and the GISS-HadCRUT divergence. There are other factors involved.

Nevertheless we notice a good general correlation. As arctic and antarctic temperatures have warmed, the divergence has generally grown. The warmer arctic and antarctic temperatures are, the bigger the divergence will be (ON AVERAGE). You keep repeating that arctic temperatures haven't really increased since 2001 (they have still been increasing just not as fast). Well neither has the divergence. 2 of the 3 largest divergences were observed in 2002 and 2003.

Also antarctic temperatures warmed nearly .5C from 2001-2003 which does coincide with a big jump in the divergence.

We also KNOW that it must be the arctic and antarctic causing the divergence, because they both agree on the rest of the globe. GISS has been diverging because it shows rapid warming in the arctic. If you take that away from GISS, it agrees with HadCRUT. We know for a 100% undeniable fact that the divergence is due to the arctic and antarctic because if you take them away, it agrees with HadCRUT. The question then becomes has the trend of its extrapolations in those regions been approximately correct. And the answer is, YES.

Like I said, we should not expect a perfect correlation between arctic temperature and the divergence. What I would generally expect is when the arctic is hot, the divergence will often be large. And that's exactly what we observe. When the arctic got hot (2001-2010) the divergence was usually large. To expect a better correlation than that doesn't make any sense given all the other factors involved.

The main problem with this post is that the divergence between GISS and the other sources did NOT peak in 2002-03, as you are saying. 2006-09 was the peak period of divergence.

[skierinvermont]

The main problem with this post is that the divergence between GISS and the other sources did NOT peak in 2002-03, as you are saying. 2006-09 was the peak period of divergence.

Arctic temperatures were higher in 2006-2009, so that actually makes perfect sense.

Even if arctic temperatures were not higher 2006-2009, it still wouldn't refute the fact that the divergence is due to the arctic. One should not expect a perfect correlation between arctic temperatures and the divergence. The basic fact that as arctic temperatures have risen, the divergence has grown, holds. We know that the rapid warming of the arctic is the cause of the divergence because when you remove it, GISS agrees with HadCRUT.

[ORH_wxman]

Arctic temperatures were higher in 2006-2009, so that actually makes perfect sense.

Even if arctic temperatures were not higher 2006-2009, it still wouldn't refute the fact that the divergence is due to the arctic. One should not expect a perfect correlation between arctic temperatures and the divergence. The basic fact that as arctic temperatures have risen, the divergence has grown, holds. We know that the rapid warming of the arctic is the cause of the divergence because when you remove it, GISS agrees with HadCRUT.

Actually it looks like both the arctic and antarctic regions were cooling late this past decade