2015 Global Temperatures

[ORH_wxman]

Has UAH for March come out yet? I don't understand the discrepancy here, something is not right. March was massively warmer than January yet UAH shows a cooling trend. Is it really possible for the lower troposphere and surface to be so out of sync?

 

Is the CO2 forcing so strong that the heat just bypasses the LT? Trapped at the surface? How about tropical convection transporting the heat above the LT?

 

On a month to month basis...absolutely

[StudentOfClimatology]

Is the CO2 forcing so strong that the heat just bypasses the LT? Trapped at the surface? How about tropical convection transporting the heat above the LT?

Considering the fact that radiative forcing from CO^2 is actually thermalized in the troposphere, this is impossible. Short term temperature differences between the surface and LT are mostly driven by changes in surface wind speed/convection. The latent heat cycle is a powerful thing.

[bluewave]

A nearly 40 year trend line changed that much?  This revision is nearly 20%...  All things considered, that is a very large difference. It also means that UAH has essentially caved to RSS in terms of recent TLT temperature trends.  There is a reason the IPCC doesn't use these datasets.  The anomaly of a recent month of Feb 2015 was changed from 0.28C to 0.11C....  

 

Skier's point prior regarding the uncertainty of remote sensing is becoming even more relevant and applicable. (Sorry SoC- but how could anyone trust this dataset for empirical studies?).  This is a major revision of a very long term trend.

 

We also have to take clouds and  precipitation into consideration when evaluating satellite temperatures.

 

http://link.springer.com/article/10.1007%2Fs00382-013-1958-7

 

abstract

Microwave Sounding Unit (MSU) and Advanced Microwave Sounding Unit-A (AMSU-A) observations from a series of National Oceanic and Atmospheric Administration satellites have been extensively utilized for estimating the atmospheric temperature trend. For a given atmospheric temperature condition, the emission and scattering of clouds and precipitation modulate MSU and AMSU-A brightness temperatures. In this study, the effects of the radiation from clouds and precipitation on AMSU-A derived atmospheric temperature trend are assessed using the information from AMSU-A window channels. It is shown that the global mean temperature in the low and middle troposphere has a larger warming rate (about 20–30 % higher) when the cloud-affected radiances are removed from AMSU-A data. It is also shown that the inclusion of cloud-affected radiances in the trend analysis can significantly offset the stratospheric cooling represented by AMSU-A channel 9 over the middle and high latitudes of Northern Hemisphere.

[The_Global_Warmer]

I don't know if its nefarious.

But the circumstantial evidence doesnt mesh at all.

They already dramatically cooled 2010-2013 with their last update.

So now during a period of record ssts and surface temps we are way way cooler than 1998?

Just not feeling it.

[The_Global_Warmer]

So basically Inspite of verifiable independent sources 2014-15 in the LT is no where near 1998 2005 or 2010?

Sorry but that's bullshyt.

[AvantHiatus]

Like I said, something is keeping it out of the LT, such as massive ocean heat uptake. In hindsight, measuring LT temperature is probably a inferior way to monitor the Earth's energy balance.

[nflwxman]

So basically Inspite of verifiable independent sources 2014-15 in the LT is no where near 1998 2005 or 2010?

Sorry but that's bullshyt.

I'm not a conspiracy theorist, but the TLT datasets just generally don't match physical expectations of greenhouse gas warming. Whether it's diurnal drift or clouds or convective overturning, it just doesn't seem right.

[StudentOfClimatology]

I'm not a conspiracy theorist, but the TLT datasets just generally don't match physical expectations of greenhouse gas warming. Whether it's diurnal drift or clouds or convective overturning, it just doesn't seem right.

The TLT is more thermally unstable than the surface, as it's influenced heavily by a slew of latent heat-related processes that don't affect the surface to the same degree. The observed reduction in surface wind speeds since 2013 could theoretically explain both the abnormal surface warmth and the lack of thermal transport/latent heat release into the troposphere. If wind speeds/convective overturning increase, evaporative cooling at the surface increases, and latent heat release in the lower/middle troposphere increases. These variables are what drive the thermal response to ENSO.

As for the UAH adjustment, it was a long time coming, as I've pointed out here before. It was obviously running warm, and the shape of the curve itself did not resemble any of the other datasets or the OLR aggregate. I'm a bit surprised at the magnitude of the adjustment, but can sleep easier now that UAH and RSS appear to agree both spatially and absolutely. The majority of CO^2 radiative forcing is actually thermalized in the mid/upper troposphere, so that is the place to measure it, if you want to do it directly:

[ORH_wxman]

I'm not a conspiracy theorist, but the TLT datasets just generally don't match physical expectations of greenhouse gas warming. Whether it's diurnal drift or clouds or convective overturning, it just doesn't seem right.

It's possible GHCN data could be too warm too in addition to TLT data being too cold. People in here like to analyze the data sets to more precision than they are intended for.

If you read up thoroughly on the temp data sets, it becomes clearer that several hundreths of a degree Celsius don't actually mean much.

The satellites have larger margins of error but the surface is probably not correct either.

[The_Global_Warmer]

I understand that global ssts being at record warmth last summer didn't have the LT impact because tropical heat with 28-31C ssts being 1-2C above normal releases way more heat than higher lat ssts being 6-15C being 3-5C above normal.

And for surface there is no difference because it's all the same spatially.

But according to the new UAH data 2013 was barely warmer than 1999-2000.

Even though ssta were 0.25-0.30C warmer in 2013.

That's kind of ridiculous.

[The_Global_Warmer]

I understand that global ssts being at record warmth last summer didn't have the LT impact because tropical heat with 28-31C ssts being 1-2C above normal releases way more heat than higher lat ssts being 6-15C being 3-5C above normal.

And for surface there is no difference because it's all the same spatially.

But according to the new UAH data 2013 was barely warmer than 1999-2000.

Even though ssta were 0.25-0.30C warmer in 2013.

That's kind of ridiculous.

[StudentOfClimatology]

I understand that global ssts being at record warmth last summer didn't have the LT impact because tropical heat with 28-31C ssts being 1-2C above normal releases way more heat than higher lat ssts being 6-15C being 3-5C above normal.

And for surface there is no difference because it's all the same spatially.

But according to the new UAH data 2013 was barely warmer than 1999-2000.

Even though ssta were 0.25-0.30C warmer in 2013.

That's kind of ridiculous.

If the sensible and latent fluxes were equal within the ocean-lower atmospheric boundary layer, yes, that would be odd. But they're not..they're inherently unstable and subject to external influence. Years like 1999-2000 featured stronger global surface wind speeds, which accelerated both evaporative cooling at the sea surface and latent heat release in the troposphere (there was a ton of IO/Indonesian convection during those years).

Now that the two satellite datasets are in good agreement, I think we can consider the interpolative issues between the two to be solved.

[skierinvermont]

Imagine the outcry if GISS revised its trend up 20% for the last 30 years. Deniers would go ape****.. it would constitute definitive proof in their minds of intentional manipulation and fraud.

 

 

I won't say the UAH revision is wrong. But it clearly demonstrates the uncertainties in MSU satellite analysis. It's yet another large revision to their methodology. When are they going to get it right? Nobody knows.

[nflwxman]

Imagine the outcry if GISS revised its trend up 20% for the last 30 years. Deniers would go ape****.. it would constitute definitive proof in their minds of intentional manipulation and fraud.

 

 

I won't say the UAH revision is wrong. But it clearly demonstrates the uncertainties in MSU satellite analysis. It's yet another large revision to their methodology. When are they going to get it right? Nobody knows.

Bingo.  These datasets are jumping around on long term trends every few years.  They are clearly too unstable to use in ECS or even really any physical or empirical study (which is why they are often not).

[chubbs]

Differences in regional trends between UAH v6 and v5.6 are striking. All of the cooling in v6 occurs in the northern portion of the NHemisphere. On the other hand the tropics and S Hemisphere show more warming in V6.

 

[nflwxman]

It's possible GHCN data could be too warm too in addition to TLT data being too cold. People in here like to analyze the data sets to more precision than they are intended for.

If you read up thoroughly on the temp data sets, it becomes clearer that several hundreths of a degree Celsius don't actually mean much.

The satellites have larger margins of error but the surface is probably not correct either.

 

The GISS dataset has a defined uncertainty from 1979-2015 of ~0.04C per decade.  UAH and RSS are approximately ~.1C/decade given the same time period.  Thus, the uncertainty range of GISS is 0.12-0.20C/decade versus 0.02C-0.22C/decade of UAH/RSS

[StudentOfClimatology]

The GISS dataset has a defined uncertainty from 1979-2015 of ~0.04C per decade. UAH and RSS are approximately ~.1C/decade given the same time period. Thus, the uncertainty range of GISS is 0.12-0.20C/decade versus 0.02C-0.22C/decade of UAH/RSS

GISS's estimated uncertainty potential could also be wrong, just as UAH's apparently was. The CRU 2001 - 2004 transition is a good example to use here. Published error potential of 0.07C on a 10yr resolution, only undergo a major revision.

[nflwxman]

GISS's estimated uncertainty potential could also be wrong, just as UAH's apparently was. The CRU 2001 - 2004 transition is a good example to use here. Published error potential of 0.07C on a 10yr resolution, only undergo a major revision.

Well, CRU actually upgraded using more measurement stations at the time, which is not quite the same thing.  The published uncertainty of the existing measurement stations (which are used in the analysis) was not incorrect or out of range.  Remember, the uncertainty is only defined within the parameters of what is being measured. For example, HadCrut4 does not define uncertainty to the entire globe, but rather just the raster pixels in which it's being interpolated (which misses a fair portion of the globe).  GISS has a higher published uncertainty for HadCrut4 simply because of the infill techniques it uses, even though it may be a more realistic depiction of global temperatures.

 

UAH has used the same input data for years, only to change that same input data into 3-4 pretty different solutions over time.  See the difference?

[StudentOfClimatology]

Actually, the globally integrated error potential was suggested to be 0.07C on a 10yr resolution in 2001. It was the 2008 CRU adjustment that blew away the previously published error potential, not the 2004 (transition to version 2 of the 2001 data). The 2008 adjustment was very large, over 0.2C on a few intervals, despite the stated error potential suggesting otherwise. It happens.

[The_Global_Warmer]

Differences in regional trends between UAH v6 and v5.6 are striking. All of the cooling in v6 occurs in the northern portion of the NHemisphere. On the other hand the tropics and S Hemisphere show more warming in V6.

uah6_regional.gif

So the area that has seen the most direct warming and major changes turns out to be dramatically top v warm?

It's also coi coincidently a huge data sparse region from the surface data set view point.

And happens to be the driver of most of the warming the last 15 yrs.

Hmm

[ORH_wxman]

You have to add in the GHCN homogenization process too which would account for another couple hundreths. All sfc datasets use GHCN with the exception of BEST...which do their own methods. Hadcrut4 also does some other sfc-based stations in addition to the most of the same GCHN ones that GISS and NCDC use.

 

In the end, this is mostly just fodder over datasets that largely agree in the long run...once you account for the uncertainties in them. If you treat it like that, then it becomes easier to take in the data.

 

 

As for the satellites, we'll have to wait until UAHv6 is published in a journal, but the prelim version matching closer with RSS when the two use different methods would lead you to be confident that the adjustments were in the right direction and not the wrong one. You just have to remember what they are measuring and try and keep it independent from the GHCN-based datasets.

[nflwxman]

Look at these changes in the lower 48 and the arctic going back to 1979 from V5.6 and 6.0 beta.  Unreal.

 

 

 

Changes from 60-90N

 

 

More data:

 

[The_Global_Warmer]

Look at these changes in the lower 48 and the arctic going back to 1979 from V5.6 and 6.0 beta. Unreal.

Changes from 60-90N

More data:

If they found a way to modify the data without looking directly fraudulent it will still be uncovered.

Why would the satellite data happen to be God awful over the NP and USA?

The 1980s get warmed up.

The middle years stay mostly the same.

The last 15 years get dramatically cooled?

Like uhhhhh all of these changes fit the exact agenda one would expect.

Crazy

[skierinvermont]

You have to add in the GHCN homogenization process too which would account for another couple hundreths. All sfc datasets use GHCN with the exception of BEST...which do their own methods. Hadcrut4 also does some other sfc-based stations in addition to the most of the same GCHN ones that GISS and NCDC use.

 

In the end, this is mostly just fodder over datasets that largely agree in the long run...once you account for the uncertainties in them. If you treat it like that, then it becomes easier to take in the data.

 

 

As for the satellites, we'll have to wait until UAHv6 is published in a journal, but the prelim version matching closer with RSS when the two use different methods would lead you to be confident that the adjustments were in the right direction and not the wrong one. You just have to remember what they are measuring and try and keep it independent from the GHCN-based datasets.

 

Haven't papers been published criticizing the spurious cooling in RSS?

[skierinvermont]

Actually, the globally integrated error potential was suggested to be 0.07C on a 10yr resolution in 2001. It was the 2008 CRU adjustment that blew away the previously published error potential, not the 2004 (transition to version 2 of the 2001 data). The 2008 adjustment was very large, over 0.2C on a few intervals, despite the stated error potential suggesting otherwise. It happens.

 

HadCRUT has larger published error bars the farther back you go. A moderate revision of the 1940s is not at odd with their previously estimated uncertainty.

 

A major revision of post-1980 data (again) is at odds with Spencer and Christy's reported error bars and more in-line with the error bars reported in the AR5 for MSU products.

[ORH_wxman]

Haven't papers been published criticizing the spurious cooling in RSS?

 

 

I am not aware of any papers...but maybe there. I don't remember any though. Most of the papers on the satellites seem to focus on TMT data for whatever reason. It was actually Spencer himself that stated he believed RSS was running too cold in recent years due to the diurnal drift model it uses. It is still running colder than the UAHv6 but they are much closer now.

[StudentOfClimatology]

This is why you don't do drugs and make models.

Dafuq?

[chubbs]

Below is UAH6 minus RSS. There is still a strong non-random signal indicating systematic differences in the two satellite series. UAH6 is offset about 0.1C cooler. In addition there are two periodic fluctuations: a short cycle with a period of roughly 2 years and a longer cycle of roughly 20 years. Tracing the longer cycle: starting in 1979 UAH6 warmed for roughly 10 years relative to RSS then cooled for the next 20 years and has started to warm again in the past 5 years.

[StudentOfClimatology]

Below is UAH6 minus RSS. There is still a strong non-random signal indicating systematic differences in the two satellite series. UAH6 is offset about 0.1C cooler. In addition there are two periodic fluctuations: a short cycle with a period of roughly 2 years and a longer cycle of roughly 20 years. Tracing the longer cycle: starting in 1979 UAH6 warmed for roughly 10 years relative to RSS then cooled for the next 20 years and has started to warm again in the past 5 years.

UAHRSS.png

UAH and RSS operate on different baselines, which is why UAH's anomalies run cooler than RSS's. Not exactly a "systematic difference" worth mentioning.

[nflwxman]

UAH and RSS operate on different baselines, which is why UAH's anomalies run cooler than RSS's. Not exactly a "systematic difference" worth mentioning.

Disagree.  There is clearly some sort of drift there throughout the dataset.  Chubbs is aware of the baseline difference.