2011 Global Temperatures

[nzucker]

And this:

It is obvious that retrieval errors can be avoided if the raw channel weighting functions, or the effective weighting functions representing a combination of the raw channels, are taken at face value—that is, a weighted vertical average of the atmosphere. This was the basis of the SC92 method used to measure a lower-tropospheric temperature (LT) from MSU channel 2. In this framework there is no error arising from a misinterpretation of the satellite-sensed layer as a thinner layer than the satellite measurements can actually be resolved.

[nzucker]

This paper is actually heavily in support of UAH/AMSU temperatures, which is what many skeptics have been basing their arguments on:

"They also demonstrated that the University of Alabama at Huntsville (UAH)-produced LT trends were more consistent with observations than were the FJWS retrievals."

The authors are actually saying that using a LT analysis across one channel is better than extrapolating inaccuracies using the overlap from two channels. In essence, they're vindicating the methods used by UAH...which is exactly the opposite of what you're claiming, if I'm reading this correctly.

[BethesdaWX]

Yep...not surprising to say the least. UAH definitely has a boost on this one.

We also find it somewhat contradictory that FJWS rejected tropospheric trend measurements from radiosondes, which show little or no warming during the satellite period of record (e.g., Christy et al. 2003), and yet depended upon radiosonde-measured stratospheric trends, where problems are considerably greater (Parker et al. 1997), for their method to work.

Then...

We have presented evidence of problems inherent in the statistical retrieval method advocated by FJWS for estimating tropospheric trends from a linear combination of MSU channel-2 and -4 data. Because of the small amount of overlap between the near-nadir MSU 2 and 4 weighting functions, these two channels cannot be effectively combined to remove stratospheric influence and provide a direct measurement of the troposphere as was done by SC92 for the lower troposphere (LT).

The dominant feature of the FJWS weighting profile’s stratospheric sensitivity is a negative stratospheric lobe of weight that can potentially lead to misinterpretation of stratospheric cooling as tropospheric warming. The FJWS interpretation of their effective weighting profile as a tropospheric measure depends upon an empirical cancellation of signals from the stratosphere. Specifically, it is necessary for the contributions from the positively and negatively weighted stratospheric portions of their weighting function to cancel in the presence of a specific type of trend profile through those layers. This makes the FJWS method dependent upon knowledge of the temperature trend profile, which is in general not well known on a global basis, and on the assumption of statistical stationarity of that trend profile. The FJWS rejection of the tropospheric trends from radiosonde data (which show little warming during the satellite period of record) seems to us to be inconsistent with their method’s dependence on the stratospheric trends from those same radiosondes. Even though the above considerations alone are sufficient to cast doubt upon any tropospheric trends inferred from the FJWS approach, we additionally show that if FJWS-style regressions use globally averaged satellite data, rather than the spatially restricted local relationships from radiosonde data, coefficients results that, on average, do not result in negative weighting function weight in the stratosphere. We demonstrate this with regressions between satellite-observed T₂ and T₄ global anomalies and NCEP–NCAR reanalyses (or LKS or HadAT2 radiosonde) of the 850–300-hPa-layer temperature anomalies. The average regression errors resulting from application of the FJWS coefficients are considerably larger that those obtained from site-by-site trend comparisons between individual radiosonde stations and the UAH MSU LT trends. Again, all of these errors arising from statistical estimation do not occur with the selective choice of different view angle weighting functions that directly remove stratospheric influence from channel 2 (SC92).

We conclude that there is substantial uncertainty in tropospheric temperature trends deduced from near-nadir MSU channels 2 and 4, due to the inability of those channels to physically remove stratospheric influence on channel 2, and the necessary dependence of any other (statistical) method on statistically stationary correlations between tropospheric and stratospheric temperature variations, which are not well known on a global basis.

[nzucker]

Actually it seems that Spencer Christy and Norris' latest tack is to just criticize ALL satellite temperature reconstructions including their own. After initially publishing their data in an attempt to refute surface trends, and after being forced to revise their estimates up several times it seems that now they have discovered satellite data is not accurate.

I don't think the paper is about refuting all satellite measurements, Skier. It's a bit more nuanced than that....I actually think their point is pretty much the opposite of what you claimed.

[Clifford]

2) GRACE measures Gravity, and shows ice loss where it is too cold to melt Ice!......also has a HUGE error bar.

The thing about ice is that it doesn't stay in one place.

The best studies seem cover less than a decade of measurements, and show moderate gains of ice in the middle of the continents, and loss of ice around the edges.

This one, for example, seems to be covering 4 or 5 years, 2003 to 2007.

There seems to be thinning ice around most of the edges of Antarctica.

Stable, or slightly growing ice in the middle.

West Antarctica shows a prominent blue/yellow spot in the middle which appears to be a 200km or so shift in ice, presumably downhill, but I don't see an explanation of that rapid of movement. I'm seeing movement estimates of 1 to 100m per year glacier movements, not 200km. Or, it could be an indication of a calibration error.

The area around the Wilkins Ice Shelf on the Antarctic Peninsula actually seems to be increasing rapidly (or was increasing in 2007). I was actually wondering if rapid growth of the ice shelf could cause sheer forces at the land/ocean junction. And, thus, the partial collapse of the ice shelf may have had more to do with growth rather than loss of ice. It is probably a naturally occurring phenomenon with an instability between the weight of the ice and the water displacement causing collapse and growth cycles (and certainly could have multi-year periodicity).

Truthfully, we are seeing what appear to be rapid changes with our recent temperature, volume, mass records. And, it may be cause for alarm. But, what we don't have is good information about natural cycles, or how far we are outside of those cycles.

[Clifford]

Actually it seems that Spencer Christy and Norris' latest tack is to just criticize ALL satellite temperature reconstructions including their own. After initially publishing their data in an attempt to refute surface trends, and after being forced to revise their estimates up several times it seems that now they have discovered satellite data is not accurate.

We conclude that there is substantial uncertainty in tropospheric temperature trends deduced from near-nadir MSU channels 2 and 4, due to the inability of those channels to physically remove stratospheric influence on channel 2, and the necessary dependence of any other (statistical) method on statistically stationary correlations between tropospheric and stratospheric temperature variations, which are not well known on a global basis.

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

-Spencer, Christy, Norris and Braswell in 2010

That appears to be MSU data, and not AMSU data.

Thus, it may be a greater historical record problem than a future record problem.

However, perhaps the issue is that we have dozens of satellites spinning around the planet. And, every time we update one, we could loose some of the correlation between current and historical data.

Most of the shift in satellite temperature records over the last 2 decades occurred in the 1998 El Nino.

Which also corresponded to a shift in reliance on MSU to AMSU temperature data.

[nzucker]

That appears to be MSU data, and not AMSU data.

Yeah, I think it's more of a rejection of the Microwave Sounding Unit in favor of the Advanced Microwave Sounding Unit, which is what UAH uses. Interestingly, doesn't UAH show a slightly higher trend in global temps since 1998? So fundamentally, UAH may be the most accurate of all the sources.

[tacoman25]

Satellite data is not better than surface data.

For one they measure two physically different things.

For another, the published uncertainties for both UAH and RSS are quite high. UAH finds a trend over the oceans of .1C/decade +/-.11C/decade. UAH can't even decide if the oceans are cooling or warming rapidly at >.2C/decade. And these are error estimates published by the renowned skeptics Spencer and Christy.

So if you want to persist in saying UAH is accurate and precise you will need to renounce the published works of John Christy and Roy Spencer.

I think you are trying to use error bars to disprove trends a bit too often. Seems to be your catch-all response lately, and it's a bit simplistic and selective in its application.

[tacoman25]

Actually it seems that Spencer Christy and Norris' latest tack is to just criticize ALL satellite temperature reconstructions including their own. After initially publishing their data in an attempt to refute surface trends, and after being forced to revise their estimates up several times it seems that now they have discovered satellite data is not accurate.

We conclude that there is substantial uncertainty in tropospheric temperature trends deduced from near-nadir MSU channels 2 and 4, due to the inability of those channels to physically remove stratospheric influence on channel 2, and the necessary dependence of any other (statistical) method on statistically stationary correlations between tropospheric and stratospheric temperature variations, which are not well known on a global basis.

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

-Spencer, Christy, Norris and Braswell in 2010

Yes, that sounds like an objective summary. C'mon, man, posts like this just make you look like a skeptic hater, no better than the idiots who refuse to acknowledge any possibility of AGW.

[nzucker]

Yes, that sounds like an objective summary. C'mon, man, posts like this just make you look like a skeptic hater, no better than the idiots who refuse to acknowledge any possibility of AGW.

Yeah he just totally manipulated the purpose of the study; I read a few more paragraphs and realized the message was completely different from what Skier claimed.

[skierinvermont]

I see Zucker has been having a field day.

Yep it looks like I misunderstood the study in my initial reading and the intended purpose is an argument against the methodology of Fu et al and in favor of their methodology.

Regardless, my original point was that there are large published error bars on the UAH data set. In the same paper Christy and Spencer state that

the global UAH LT trend confidence interval (95%) was calculated as being ±0.05°C decade⁻¹.

This matches the graphic in the IPCC report. And indicates that there is substantial uncertainty. The calculated uncertainty is representative of quantifiable error like sampling error.. it does not include further error due simply to flaws in methodology. Regardless, the confidence interval published by Spencer and Christy themselves is fairly large at +/- .05C/decade. This supports the graphic I drew from the IPCC report which Bethesda scoffed at. Sorry Bethesda - this comes directly from Spencer and Christy.. are you going to disagree with them too?

Regarding Spencer and Christy's argument against Fu et al. there are a number of papers dealing with the same issue which contradict Spencer and Christy.

And we have had Spencer and Christy argue emphatically for a lower temperature trend several times only to subsequently be forced to admit they were wrong.

In 1992 they published their original UAH version which showed a cooling trend of .05C/decade. They wrote a number of papers arguing that their analysis was more consistent with radiosonde observations and on the accuracy of their method. After other researchers pointed out that orbital decay was unaccounted for the trend was increased from -.05C/decade to +.07C/decade in version 5.1.

Spencer and Christy again wrote several papers in support of their analysis, critiquing others, and arguing their's was more consistent with radiosonde data.

Then again several papers to Science in 2005 pointed out that the diurnal correction made in UAH version 5.1 was incorrect. Spencer and Christy again conceded and the trend again increased over 40% from version 5.1 to 5.2 to the current trend of .14C/decade. The error estimates on version 5.1 and versions 5.2 were/are stated as +/-.05C/decade.

They published the following OOPS! letter to Science acknowledging their error and thanking Mears and colleagues at RSS for pointing out the error:

WE AGREE WITH C.A. MEARS AND F.J. WENTZ ("The effect of diurnal correction on satellite-derived lower tropospheric temperature," Reports, 2 Sept., p. 1548; published online 11 Aug.) that our University of Alabama in Huntsville (UAH) method of calculating a diurnal correction to our lower tropospheric (LT) temperature data (v5.1) introduced a spurious component. We are grateful that they spotted the error and have made the necessary adjustments. The new UAH LT trend (v5.2, December 1978 to July 2005) is +0.123 K/decade, or +0.035 K/decade warmer than v5.1. This adjustment is within our previously published error margin of [+ or -] 0.05 K/decade (1).

-Christy and Spencer 2005 letter to Science

So we have seen Spencer and Christy insist that they are right multiple times in the past, publish journal articles arguing that their method is better and more consistent with other observations, only to concede that their trend was biased too cold several years later.

I see no reason why I should take an article by Spencer and Christy too seriously given their errors in the past and the half dozen articles written in the past 2 years that disagree with Spencer and Christy's method. Some of these papers are discussed in the review I posted by Thorne et al. 2010.

Thorne, by the way, published in defense of Christy and Spencer back in 2004 agreeing with them that "There still remain large differences between observed tropospheric temperature trends and those simulated by a climate model."

-Thorne 2004 in Nature

But he has been persuaded by the studies of the past 2 years and now concludes:

"Overall, there is now no longer reasonable evidence of a fundamental disagreement between models and observations with regard to the vertical structure of temperature change from the surface through the troposphere."

-Thorne 2010 in Climate Change

[tacoman25]

I see Zucker has been having a field day.

Yep it looks like I misunderstood the study in my initial reading and the intended purpose is an argument against the methodology of Fu et al and in favor of their methodology.

Regardless, my original point was that there are large published error bars on the UAH data set. In the same paper Christy and Spencer state that

the global UAH LT trend confidence interval (95%) was calculated as being ±0.05°C decade⁻¹.

This matches the graphic in the IPCC report. And indicates that there is substantial uncertainty. The calculated uncertainty is representative of quantifiable error like sampling error.. it does not include further error due simply to flaws in methodology. Regardless, the confidence interval published by Spencer and Christy themselves is fairly large at +/- .05C/decade. This supports the graphic I drew from the IPCC report which Bethesda scoffed at. Sorry Bethesda - this comes directly from Spencer and Christy.. are you going to disagree with them too?

What's the big deal about .05C/decade uncertainty bars? Isn't that similar to other confidence intervals for similar time frames?

[tacoman25]

Regarding Spencer and Christy's argument against Fu et al. there are a number of papers dealing with the same issue which contradict Spencer and Christy.

And we have had Spencer and Christy argue emphatically for a lower temperature trend several times only to subsequently be forced to admit they were wrong.

In 1992 they published their original UAH version which showed a cooling trend of .05C/decade. They wrote a number of papers arguing that their analysis was more consistent with radiosonde observations and on the accuracy of their method. After other researchers pointed out that orbital decay was unaccounted for the trend was increased from -.05C/decade to +.07C/decade in version 5.1.

Spencer and Christy again wrote several papers in support of their analysis, critiquing others, and arguing their's was more consistent with radiosonde data.

Then again several papers to Science in 2005 pointed out that the diurnal correction made in UAH version 5.1 was incorrect. Spencer and Christy again conceded and the trend again increased over 40% from version 5.1 to 5.2 to the current trend of .14C/decade. The error estimates on version 5.1 and versions 5.2 were/are stated as +/-.05C/decade.

They published the following OOPS! letter to Science acknowledging their error and thanking Mears and colleagues at RSS for pointing out the error:

WE AGREE WITH C.A. MEARS AND F.J. WENTZ ("The effect of diurnal correction on satellite-derived lower tropospheric temperature," Reports, 2 Sept., p. 1548; published online 11 Aug.) that our University of Alabama in Huntsville (UAH) method of calculating a diurnal correction to our lower tropospheric (LT) temperature data (v5.1) introduced a spurious component. We are grateful that they spotted the error and have made the necessary adjustments. The new UAH LT trend (v5.2, December 1978 to July 2005) is +0.123 K/decade, or +0.035 K/decade warmer than v5.1. This adjustment is within our previously published error margin of [+ or -] 0.05 K/decade (1).

-Christy and Spencer 2005 letter to Science

So we have seen Spencer and Christy insist that they are right multiple times in the past, publish journal articles arguing that their method is better and more consistent with other observations, only to concede that their trend was biased too cold several years later.

I see no reason why I should take an article by Spencer and Christy too seriously given their errors in the past and the half dozen articles written in the past 2 years that disagree with Spencer and Christy's method. Some of these papers are discussed in the review I posted by Thorne et al. 2010.

Thorne, by the way, published in defense of Christy and Spencer back in 2004 agreeing with them that "There still remain large differences between observed tropospheric temperature trends and those simulated by a climate model."

-Thorne 2004 in Nature

But he has been persuaded by the studies of the past 2 years and now concludes:

"Overall, there is now no longer reasonable evidence of a fundamental disagreement between models and observations with regard to the vertical structure of temperature change from the surface through the troposphere."

-Thorne 2010 in Climate Change

At least they can admit when they are off. I have never seen Hansen concede that his early temperature predictions were too high.

[The_Global_Warmer]

If that is you bethesda in your avatar you lost a lot of muscle bro

[BethesdaWX]

If that is you bethesda in your avatar you lost a lot of muscle bro

Nah I'm the same 255lbs, I just edited my first photo to make myself look bigger, changed things around a bit, and I looked like an instant Hunk.

That said, my 255lbs....only 6% body fat

[skierinvermont]

At least they can admit when they are off. I have never seen Hansen concede that his early temperature predictions were too high.

I have quoted him saying that the climate sensitivity used in his original study was too high on this forum multiple times.

[BethesdaWX]

I have quoted him saying that the climate sensitivity used in his original study was too high on this forum multiple times.

And Spencers report on UAH, about +/-0.05C, is less than GISS, HADCRUT, basically everyone. Thus, the Antarctic Trend is Validated.

This is because UAH really has more data than anyone else.

[tacoman25]

I have quoted him saying that the climate sensitivity used in his original study was too high on this forum multiple times.

Ah, that's true forgot about that.

Well, if you are going to try to discredit Spencer/Christy by pointing out they were off in the past, it's only fair to question anything Hansen says, since he has been off in the past as well.

[BethesdaWX]

Ah, that's true forgot about that.

Well, if you are going to try to discredit Spencer/Christy by pointing out they were off in the past, it's only fair to question anything Hansen says, since he has been off in the past as well.

haha+++

"off"...to put it lightly? I would not show that level of respect.

[WeatherRusty]

I have quoted him saying that the climate sensitivity used in his original study was too high on this forum multiple times.

Hansen used a climate sensitivity of 4.2C per CO2 doubling back in 88. It is now thought to be closer to 3C (2.8) +-1.5C.

[Snow_Miser]

Meanwhile, back in reality from the alarmists' hypothetical green-energy funded predictions, we can clearly see what is going to happen in the near future. Should bring the Global Temperature in Fenruary way down to the dumps.

[LakeEffectKing]

Mr. Had Crut has chimed in on Jan.:

2011/01 0.194

[WeatherRusty]

Meanwhile, back in reality from the alarmists' hypothetical green-energy funded predictions, we can clearly see what is going to happen in the near future. Should bring the Global Temperature in Fenruary way down to the dumps.

Meanwhile, back in reality from the denier's hypothetical fossil fuel industry funded predictions, we can clearly see what is going to happen in the more distant future.

[Mencken_Fan]

Nah I'm the same 255lbs, I just edited my first photo to make myself look bigger, changed things around a bit, and I looked like an instant Hunk.

That said, my 255lbs....only 6% body fat

You are very tall?

[nzucker]

Meanwhile, back in reality from the denier's hypothetical fossil fuel industry funded predictions, we can clearly see what is going to happen in the more distant future.

Those are models, whereas the GFS is showing projected temperatures for the next week...I think there's a slight difference.

[Snow_Miser]

Meanwhile, back in reality from the denier's hypothetical fossil fuel industry funded predictions, we can clearly see what is going to happen in the more distant future.

We were not even talking about sceptical predictions...

[tacoman25]

Meanwhile, back in reality from the denier's hypothetical fossil fuel industry funded predictions, we can clearly see what is going to happen in the more distant future.

Um, the maps for next week are much more grounded in reality than maps for decades in the future. What an absurd post.

[WeatherRusty]

Those are models, whereas the GFS is showing projected temperatures for the next week...I think there's a slight difference.

Are you trying to say the GFS is not a model?

[WeatherRusty]

We were not even talking about sceptical predictions...

No, your jumping up and down in glee seeing that the models project a cooling over the next 192 hours. Wonder why that might be? Maybe you think it's another nail in AGW's coffin? Think again.

[WeatherRusty]

Um, the maps for next week are much more grounded in reality than maps for decades in the future. What an absurd post.

Who's reality? I have little doubt the GFS will generally verify. I also have little doubt the GCM's will generally verify, being that they are grounded in well established physics. In both cases the divergence will likely be found only in the details.