People send me stuff. Today, alert reader Clay Ablitt sends this:
I have been keeping a record of a lot of the different data sets that are put out by RSS and UAH because i believe they are a more reliable data set than NASA or NOAA data.
I noticed in the latest monthly update that added the September data, the historic temperatures were adjusted without any notes or version change.
I have attached the data from August and the newly adjusted data from September for your consideration. This will have an impact on all RSS data sets that include the ocean temperatures such as the global RSS TLT data which has continued to show a pause since February 1997.
He also attached an Excel spreadsheet with two pages, one for each month, a link to which is available here: rss-temperature-trend-sep-oct-2016 (.xlxs)
I checked out the worksheet, and he appears to be correct. There is an unannounced change to the Remote Sensing Systems data. The last change note I am aware of is this one: http://www.remss.com/node/5166
There seem to be no other mentions on the remss.com website that explain this change as observed in the flip chart below:
(click image if it doesn’t animate for you while reading this)
I asked UAH scientist Dr. Roy Spencer about it today, showing him the data and he replied:
We suspected they have a revised LT in the works, after they came up with a new MT.
“MT” refers to Middle Troposphere data, and “LT” refers to Lower Troposphere data. Last March, WUWT covered their adjustment to the MT data, making the trend warmer.
Of course, the unannounced LT adjustment discovered by Ablitt also makes the trend warmer, some thing that isn’t entirely unexpected given the remarks last year by RSS chief scientist Carl Mears:
I wrote last March in The ‘Karlization’ of global temperature continues – this time RSS makes a massive upwards adjustment:
All that is about to change. Readers may recall a video produced by the execrable “Climate Crock of the Week” activist Peter Sinclair that we covered here where the basic premise was that the “satellites are lying“. It seems to me based on his recent comments that Dr. Mears has gotten fed up with people using his RSS data set to suggest that the world isn’t warming as he expects it should. From the video Mears states:
They just wanted to know, you know, they wanted to fine-tune their statement about, you know, whether , you know, the surface temperatures are more accurate or the satellite temperatures are more accurate, and initially they wanted to say something like “But you really shouldn’t trust the satellite temperatures, you should go with these surface temperatures”, and I said, “Well, what I would like to emphasize, you’d really want to look at all the different datasets, so you don’t want to trust only the satellite temperatures, you want to look at the surface temperatures, and – and that sort of thing.
On his website, Mears makes this statement:
Recently, a number of articles in the mainstream press have pointed out that there appears to have been little or no change in globally averaged temperature over the last two decades. Because of this, we are getting a lot of questions along the lines of “I saw this plot on a denialist web site. Is this really your data?” While some of these reports have “cherry-picked” their end points to make their evidence seem even stronger, there is not much doubt that the rate of warming since the late 1990’s is less than that predicted by most of the IPCC AR5 simulations of historical climate. This can be seen in the RSS data, as well as most other temperature datasets. For example, the figure below is a plot of the temperature anomaly (departure from normal) of the lower troposphere over the past 35 years from the RSS “Temperature Lower Troposphere” (TLT) dataset. For this plot we have averaged over almost the entire globe, from 80S to 80N, and used the entire TLT dataset, starting from 1979. (The denialists really like to fit trends starting in 1997, so that the huge 1997-98 ENSO event is at the start of their time series, resulting in a linear fit with the smallest possible slope.)
Source: http://www.remss.com/blog/recent-slowing-rise-global-temperatures Archived here: http://www.webcitation.org/6fiS2rI7k
Mears uses the term “denialist” so there goes his objectivity when he feels the need to label people like that.
The last change note, as linked above, does say:
“The lower tropospheric (TLT) temperatures have not yet been updated at this time and remain V3.3. The V3.3 TLT data suffer from the same problems with the adjustment for drifting measurement times that led us to update the TMT dataset. V3.3 TLT data should be used with caution.”
Yes Nick.. it certainly seems that RSS should now be used with caution.
As we can all see….. IT CANNOT BE TRUSTED.
Who knows when the past data will get changed again.!
” IT CANNOT BE TRUSTED”
RSS aren’t asking you to trust V3.3. It’s deprecated.
Yep, the alarmist glitterati HATE it.
Why do you thing Carl is trying to “adjust” it.
They are saying it can’t be trusted because they are about to introduce a warming trend far greater than the little undocumented tweak seen here.
This is the gift that just keeps giving!
The logic:
All datasets are false until they are updated and then they are false until they are updated…
The absurdist formula:
Version + update = True version
True version + update = Truer version
Therefore:
Truest Version = version + update (To the n-th)
V-uck me!!! 😉
To clarify, datasets are false until they are have been updated for the n-th time!
“All datasets are false until they are updated”
RSS identified a specific issue with their index, and warned that it should be used with caution, pending a new version.
UAH also identified an issue with V5.6 and brought out a new version (which has had five adjustments already). But the new version had lower trend, so everyone thinks it is great.
From the thread on Whacky Whadams yesterday, I mused on the neglected state of a number of climate data sets and alarmist remarks that always precede a Karlization Event (sea ice, BOM stuff, sea level, argot buoy results dissing, questioning the definition of hurricane landfall) and said to watch for more human caused climate change by manipulation.
https://wattsupwiththat.com/2016/10/09/whacky-peter-wadhams-doubles-down-on-the-sea-ice-crisis/#comment-2316324
PS: I’m pleased two terms I coined seem to have served well: Karlization and The GangGreen.
It appears to be a step-change around the beginning of 2009, with all subsequent readings adjusted upwards the same amount. More on this tomorrow.
Yup, that is also what I see. I suspect this may be when a different satellite is used and they adjusted the data for that satellite. Of course, as with all climate manipulations the change always leads to an increasing trend.
do a difference plot. the changes go back to 1997.. in the 1/100ths and 1/1000ths.
not material
Yes Pat,
.
The differences start being most pronounced in 2009 but the differences first appear in 1998. In the following is a plot showing the differences between October and September RSS data sets and a similar plot showing the differences between UAH v6 beta 5 and beta 1 see –
To put all of this in perspective, here are the same plots, but this time including a plot showing the changes between UAH v6 and UAH v5.6 . see-
The trend values for the relevant periods are 0.14 degrees per decade for UAH v5.6, 0.114 for v 6 beta1, 0.1105 for UAH v6 beta5, 0.132 for RSS (September data) and 0.135 for RSS for the October data.
As you can see much ado about nothing.
If the annual bonuses for Meares and his fellow RSS workers are based on increasing the trend in the RSS data then they have not done much to deserve it.
I note the very low R^2 values: basically just noise….
I’m not, you know, so sure that, you know, we’re, you know, ever going to, to see the original, you know, data, you know, sets. You know?
Brian, when I had a problem with, you know, he said I should try marijuana, because, you know?
So, if you’re also having problems with, you know, ask your doctor about marijuana.
As the great philosopher said; there’s what you know and what you think you know but isn’t so.
You know?
The RSS TLT change from August to September appears to bring it into better alignment with the UAH TLT since 2001. Perhaps they adopted some of the UAH approach, but RSS should report this change and provide an explanation.
the changes is in 1/100ths and 1/1000ths of a degree. according to the posters own spreadsheet
Interesting that he did not do a difference plot…
The change isnt relevant
..then put the little red line back where it was
but when the “changes” always cause the trend to change in the same direction… that is relevant.
Its like GISS, BEST, Had….. multiple progressive changes that have created a totally false trend.
RSS will lose credibility for fraud.
However they will gain money from the NSF USA Government for verifiable fraud.
For RSS the truth is simple, get the money.
I highly doubt that RSS is trying to pass things buy unnoticed. They may have dumped the new dataset in or portions of it accidentally before they were planning on releasing it. The problem with all of this global warming stuff is that the differences are so tiny compared to the overall trend (coming out of little ice age) , it is all essentially noise. The only significant data are the El Ninos.
It is very important that someone keeps track of their past data, so thanks for that, since every version from the warmistas keeps changing history to fit the narrative.
“pass things buy” – a Freudian slip?
Mears’ gubment funding was probably put on ransom by the WH Science Advisor. “Resistance to the Climate Borg is futile. You will assimilate or perish. Choose one,” is the message.
All data of interest to the Climate Alarmists is subject to, and paid for, having periodic adjustments that gradually move the data in the direction of the Alarmist predictions. It is only the efforts of those who have preserved the original raw data that will ensure that ,after the madness passes, real data can again be accessed for scientific purposes. The adjustments cover temperature, rain, sea level, ice area and volume, and even CO2 estimates. Very little is untouched..
polar bear counts
Individual tide-station measurements of sea-level still seem to be trustworthy. That’s why we can use NOAA & PSMSL data and graphs for things like this:
http://www.sealevel.info/680-140_Sydney_2016-04_anthro_vs_natural.png
That’s an illustration which for some reason was not included in the Slangen, Church, et al, (2016) paper in Nature Climate Change, which reported the discover that most sea-level rise since 1970 is caused by man, but most sea-level rise before 1950 was not.
🙂
The linear trend line in that graph shows the pre-1950 rise rate is less than the overall rise rate. The middle third has the linear trend line majority above the blue curve, and the ends have the trend line majority below the blue curve. I would like to see a 2nd-order best-fit curve – it would be concave-upwards – accelerating.
Not much, Donald. The linear trend for the full 1886-2010 period is a near-perfect fit to the post-1970 “Anthropogenic” period.
The “Mixed” and “Anthropogenic” periods appear to be almost perfectly linear, at 0.65 mm/yr. The pre-1950 “Natural” period is a little more irregular, but not much. That might be due to slightly different data collection methodology for 1886-1914; see the note on NOAA’s page:
http://tidesandcurrents.noaa.gov/sltrends/sltrends_global_station.shtml?stnid=680-140
“[Metadata: Data for 1886- May 1914 are based on high and low waters and on monthly mean tide levels plus a 1.7 mm correction]”
Interesting that an Australian would send you that stuff. In looking over it and checking web references given I found some problems that both I and others have overlooked. Let’s take the most recent slowdown/cooling of the twenty-first century. The info I had in 2008 for the twenty-first century then suggested an essentially flat temperature path. for the beginning of the century. Looking at the same region now, with a data-set going as far as 2015, it is clear that what looked flat then is actually part of a downward slope associated with cooling. The first figure in your paper shows it and I attempted to determine the degree of cooling by measuring the slope in this figure. It turned out to be minus 0.23 degrees Celsius per decade. There could be some degree of uncertainty in this because in order to get both ends of the line tied down I had to jump over the 2008 La Nina and the 2010 El Nino. Beyond 2012 global temperature starts to increase again in preparation for the 2015/2016 El Nino. The interesting question is, what will temperature do when that El Nino is finished? Some thought it might continue at the level of the beginning hiatus of the century but this is out because of the existence of the cooling. If you extend the straight line defining the cooling period it will point to the bottom level of the temperature scale on the opposite side of the 2016 El Nino. That bottom line is pretty much the same bottom line as that of the super El Nino of 1998. And that bottom line in turn was taken over from the eighties and nineties. There was a hiatus there also until at least 1997 when NASA refers to it. Unfortunately you cannot see it now because NOAA and friends decided to invent a non-existent warming they call “late twentieth century warming” for that spot. The hiatus was wiped out and official temperature curves were changed to show warming, not a hiatus. It is important to understand how the warming/cooling aspect of the twenty-first century developed. As soon as the super El Nino of 1998 had left a step warming started in1999. In three years it raised global temperature by a third of a degree Celsius and then stopped. It had nothing to do with the greenhouse effect. It was made possible by the warm water supply the super El Nino had brought across the ocean and then left behind. As a result, all temperatures of the twenty-first century were initially elevated by a third of a degree Celsius. This led to numerous claims of “warmest ever” temperature peaks. Hansen quickly noticed that nine out of ten warmest temperatures were all located in the first decade of the twenty-first century. He quickly claimed that greenhouse warming had done that which is nonsense. In the beginning warming dominated and it was not obvious that a cooling was on the way. The observed cooling is caused by the fact that the initial warm temperature supply was slowly cooling and could not be replaced because the super El Nino had already left. It is likely that the cooling itself will have run its course by the time the El Nino of 2016 is finished. It is likely that the new base temperature that will then follow will be similar to the original temperature of the eighties and nineties.
I am still hopeful that science can survive the idea that if the data doesn’t support your hypothesis then your data needs to be adjusted.
“you know”, I mean “you know”, a man with one clock always knows what time it is. “you know”?
But a man with two clocks is never sure…
Meanwhile …. wildfires TWIIIIIIIIIIIICE AS BIIIIIIIIIIIIIIIIgggggggggg! Due to AGW … but of course!
too funny.
The changes look like they track with theintroduction of AMSU ( a big problem)
The monthly differences are mouse nuts
0.001
0.001
0.001
0.002
0.001
0.001
0.001
0.001
0.001
0.001
0.002
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
-0.002
-0.002
-0.002
-0.001
-0.002
-0.003
-0.003
-0.002
-0.001
-0.002
-0.003
-0.004
-0.002
0.001
0.001
-0.003
-0.004
-0.005
-0.001
-0.003
-0.004
-0.005
-0.002
0.001
-0.001
-0.004
-0.004
-0.002
0.001
-0.001
-0.001
0.001
0.001
-0.002
-0.002
-0.001
0.001
-0.002
0.001
0.001
0.002
0.002
-0.001
-0.003
-0.001
0.001
0.001
0.002
0.002
0.001
0.001
-0.001
-0.001
0.001
0.001
0.001
0.001
0.001
0.002
0.001
0.001
-0.001
0.016
0.008
0.011
0.012
0.003
0.006
0.019
0.014
0.014
0.007
0.009
0.009
0.008
0.001
0.008
0.006
0.003
0.012
0.014
0.022
0.007
0.004
0.016
0.002
-0.003
-0.012
0.003
0.01
0.002
0.003
0.005
0.008
0.006
-0.002
-0.001
-0.005
-0.008
0.007
0.015
0.016
0.018
0.019
0.009
0.016
0.025
0.015
0.018
0.012
0.02
0.013
0.011
0.023
0.015
0.011
0.011
0.022
0.021
0.015
0.014
0.017
0.018
0.015
0.02
0.021
0.022
0.016
0.02
0.02
0.022
0.028
0.019
0.017
0.016
0.015
0.029
0.034
0.023
0.021
0.021
0.012
0.006
0.027
0.028
0.02
0.017
0.026
0.029
mouse nuts…..the little red line changed
Latitude, ever notice how much Mr. Mosher posts when he is on the clock, being paid? Could that affect his “latitude?”
[snip]
At some point in the future, they will have to reverse the current delta-T they’ve added in, and subtract it from the actual data in order to keep a perpetual upward trend going on at the end of a current set of data. Otherwise, if they keep increasing the delta-T to be added to the actual data, the discrepancy between what temperatures are graphed and what temperatures we actually feel will be too disparate to be believable – but then faith in their religion may be quite blinding.
nonsense
If you read Bob Tisdale’s critique of Karl, you will see that is exactly what Karl did: Lower some intervening temperatures and raise the end.
https://bobtisdale.wordpress.com/2016/01/27/on-the-monumental-differences-in-warming-rates-between-global-sea-surface-temperature-datasets-during-the-noaa-picked-global-warming-hiatus-period-of-2000-to-2014/
This little nugget from the blog post suggests a real bias when picking statistical approaches:
Charlie Skeptic aka Dave Fair
Overall. .Over the whole series…He lowered the long term trend and it implies therefore a lower ecs.
Aaaah, Mr. Mosher. Karl did not care about some unmeasurable change in ECS. He had a higher duty: Get rid of the hiatus!
what temperatures we actually feel…ridiculous, obviously, in another 5-10 years, we will know what the temp is right now 😉
Mears suffers an incoherence that is persuasive when it comes to adjudging his scientific exactitude & precision.
You got me with that one Ross. What is a persuasive incoherence?
Perhaps President Trump will instruct his investigators to look into Climate Fraud too?
Changing scientific data to falsify it is wire fraud because it is intended to paint a picture (a lie ) in order to relive tax payers of their money .
Scientists and the other promoters of the global warming fraud haven’t clued into what they will be held accountable for . Wire fraud is good for 20 years no matter how much money you have .
Except they are not changing data. The input data they use is unchanged. They like uah have algorithms to estimate the average temperature of a thick layer of the atmosphere. Improving changing modifying the algorithm is not changing data. They changed their method.
..it just changes the result….and that’s the whole idea
Isn’t the Federal Government paying for this data?
Wouldn’t this be a legitimate investigation for Lamar Smith’s committee?
Even with his 2nd graph,
the trendline from 2000 looks flat to me.
So even if he is in denial about 21st century trend he still shows the pause.
And of course all adjusted measurements show this century as warmer so even if it’s flat for 80 years it’s still warmer than last Century.
So what? It should reduce the winter electricity bills! (sure…)
Well to his credit, he did NOT put any “trend lines” on that graph at all. He lets the data speak for itself. Anyone can see just how much the model runs are overheating.
He also did not try Santer’s trick of trying to disguise the discrepancy by aligning in the middle. He starts both at the same point so that the divergence is clear.
Fair play to him.
Unfortunately is stops at the end of 2014 so misses the recent El Nino.
Authors:
Carl Mears
Date Added:
Monday, September 22, 2014
That graph is taken from his site a few years back before he allowed himself to be blugeoned by the alarmists.
Why care about RSS TLT 3.3? It is no longer endorsed by its producers due to drifts:
http://www.remss.com/node/5166
“The lower tropospheric (TLT) temperatures have not yet been updated at this time and remain V3.3. The V3.3 TLT data suffer from the same problems with the adjustment for drifting measurement times that led us to update the TMT dataset. V3.3 TLT data should be used with caution.”
Use RSS v4 TTT instead. TTT trends are very similar to TLT trends…
If there is a change to TLT 3.3, I suspect that they have discarded NOAA-15 after 2011 and Aqua after 2009. During the work with the new v4 dataset they found that those two satellites are not reliable after those dates.
Since NOAA-15 (AMSU-5) is the prime “pause-maker” every little limitation of its use in a dataset, will increase the trend. Actually, the least “pausiest” troposphere satellite dataset is UAH 5.6, since it mainly relies on nondrifting AMSU satellites, and only use NOAA-15 for a short while in the beginning.
Almost all of difference between satellite and ground is due to amsu.
Yes, between satellites and radiosondes as well. Year 2000 is in the middle of the MSU/AMSU transition:
https://drive.google.com/open?id=0B_dL1shkWewaNDVmS0t1bjZjQXM
O R
That’s a fascinating graph you have posted.
Looking at Christy et al.’s 2003 paper where they validated their UAH satellite data using radiosonde data, the last data they compared the two for TLT was up to April 2002. See – http://journals.ametsoc.org/doi/abs/10.1175/1520-0426%282003%29202.0.CO%3B2.
According to your graph the difference between the two was not that pronounced at that time in 2002 as it is later.
It would be interesting to know whether a paper that validates the UAH TLT data by comparison to radiosonde data has been published more recently than 2003.
I am only aware that Spencer did some comparisons with the radiosonde data earlier this year for TMT when RSS TMT v4 came out.
A similar discrepancy between RSS TLT v3. 3 and Ratpac radiosonde data has been noted by Tamino ( https://tamino.wordpress.com/2015/09/24/exogenous-redux/) and maybe this why they will be revising the RSS TLT product to version 4.
Is anyone wondering what the trend of a new RSS v4 TLT might be? Let’s make some qualified guesses:
1. The trend of RSS TLT 3.3 is 1.48 times higher than that of TMT 3.3. Apply the factor 1.48 on the trend of TMT 4 (0.138 C/dec) and you get 0.204 C/decade
2 Use the UAH v6 TLT formula “LT = 1.538*MT -0.548*TP +0.01*LS” with channel data fom RSS. However, RSS doesn’t think that tropopause channel data (MSU 3) is reliable before 1987, so you have to splice on early data from UAH to make the tropopause channel of RSS complete..
Doing this results in a RSS v4 TLT trend of 0.207 C/decade
From these two methods we can assume that the trend of RSS v4 TLT will be about 0.205 C/decade
Starts off with the “little” adjustments.. you get away with that
It builds and builds until your scientific integrity, and your adjustments, are on par with those of Gavin Schmidt.
And you have ZERO credibility, and ZERO integrity, and ZERO self -respect.
Is that really where Carl Mears wants to go ??
is the money worth it, Carl ??
You do have to live with yourself. !!!
NOAA STAR has a new v4 dataset (still beta, hidden in ftp), using Po Chedley (2015) diurnal drift corrections, which corroborates RSS TMT v4. The trend of STAR TMT v4 is slightly larger than that of RSS TMT v4, 0.143 vs 0.138 C/dec
With the UAH v6 TLT formula “LT = 1.538*MT -0.548*TP +0.01*LS” it is also possible to assess the uncertainty of such multichannel datasets.
If we use all available channel data from the different providers and combine freely with the formula above we get the following spread:
TLT-trend (C/decade) = 1.538*(0.071-0.143) – 0.548*(0.001-0.012) +0.01*(-0.330 – -0.259)
Hence, the formula can produce TLT trends between 0.098 and 0.216 C/decade for 1979-now.
Global surface dataset trends are between 0.172 and 0.185 C/decade for the satellite era. (The lower limit is 0.161 if also the not fully global NOAA land/ocean dataset is included)
Thus, the uncertainty in the multilayer TLT-concept is about NINE times larger than that of global surface datasets..
I believe this is incorrect. If each of the channels is independent and has normally distributed uncertainty, I suspect that the combination of the three channels reduces the error, not increases it. It’s basically the same as averaging a bunch of noisy estimates together…the average is closer to the truth than most of the individual estimates.
” It’s basically the same as averaging a bunch of noisy estimates together…the average is closer to the truth than most of the individual estimates.”
No, that’s wrong. When you average like that, the multipliers are positive and less than one. But here they are not. If the std error of each is the same, then the std error of the combination is multiplied by the quadrature (sqrt(sum squares)) value of the coefficients. In this case, that is sqrt1.538²(1.538²+0.548²+0.01²)=1.633
When adding data values from samples with errors, If the errors are fully independent, you can add the error margins in quadrature, which reduces the combined error margin. But if the errors are systematic you must add them arithmetically, yielding a larger combined error margin. (If the errors are some mixture of independent and systematic, then the true error margin is somewhere in between.)
So, for example, if you have two measurements (n=2), each with an error margin of ±1, and the errors are systematic (i.e., it is expected that if one of them errs on the positive side them both do), then the error margin of the sum of the two measurements will be 2 × ±1 = ±2, and the error margin of the average of the two will be ±2/n = ±1. So averaging multiple measurements gets you no improvement in error margin, if the errors are systematic.
OTOH, if you have two measurements (n=2), each with an error margin of ±1, and the errors are independent, then the error margins add in quadrature, so the error margin of the sum of the two measurements will be ±sqrt( 1² + 1² ) = ±1.414, and the error margin of the average of the two measurements will be that divided by n=2, i.e., ±0.707. Thus, averaging multiple data values reduces the error margin, if the errors are independent.
On the gripping hand, if it is known that the errors are of opposite sign, that’s even better, and the true error margin will be even smaller than if the errors are independent.
I do not understand the satellite temperature data, and I have no idea what that weighted sum is, so I cannot even begin to judge how to calculate the error margins.
Dr Spencer,
There are likely some errors with my simplified assessment, but still, the variation due to possible choices is a magnitude higher with the satellite datasets.
We can do it simpler but possibly more correct. Most of the variation is in the TMT layer, and TMT is the main “engine” of troposphere products.
Global TMT trends differ from 0.071 to 0.143 C/dec for the satellite era, a difference of more than 100%.
The trends, for the satellite era, of globally infilled surface datasets differ by less than 8% (0.172-0.185 C/dec; Gistemp loti, BEST l/o, C&W)