John Christy doesn’t like criticism of the satellite temperature record. He has multiple objections, but the one that struck me as bizarre is this one which was recently brought to my attention:
“Do they ever say that, unlike the surface data, the satellite datasets can be checked by a completely independent system – balloons?”
I have more than once.
Every time I check the satellite datasets against balloons, to see whether they’re telling us the right story about the trend in temperature, the satellite datasets look like they’re wrong.
Maybe I missed it: in your last comparison do you mask the satellite data to match the balloon coverage?
This makes a big difference in model-observation comparisons of surface temperature change – how much does it matter in the troposphere?
[Response: The folks who created RATPAC took pains to make it emulate a global estimate. That’s why it’s called “RATPAC” — Radiosonde Atmospheric Temperature Product for Assessing Climate. One also has to wonder, why do satellites and balloons match so well until 2000 but diverge after then? All signs point to satellite data drift.]
Some years ago Christy was given an OpEd in the NYTimes. I was struck at the time by his claim that he was excluded and has trouble getting funded. The whole thing sounded like a nice literate plausible sour grapes. As far as with my limited abilities I can discern, it is his failings in the science he calls his career, not some cabal, that have resulted in his lesser standing. He criticizes the atmospherics without acknowledging his part in making real scientists not feel like shaking his hand or support him in any way. Judith Curry comes across much the same way.
Sadly, there appear to be a good few people unable to sort the reasons for their failures to achieve top rank, and blame others for their failings. There is a ready audience for victim bullying there.
At the time, I knew the now retired science editor, and had the opportunity present some materials as to why this was wrongheaded. But they will persist in trying to create balance, as was shown in their recent awful by Paul Thacker supporting Rep. Smith and Sen. Cruz’s inquisition on NOAA.
The NYTimes is not the world, but it is unfortunate that their leadership has become more rather than less beholden to a business model that requires obeisance to the false business model that dominates our public domain.
Over time I’ve come to the conclusion that most ‘contrarian’ climate scientists (and since that’s a very small group, also those who claim to be, such as geologists who’ve worked in areas that could claim some peripheral connection to climate) are generally at best mediocre in their fields.
I don’t want to jump immediately to an ad hominem explanation, but my impression is that they aren’t mediocre because they are ‘contrarian’, but are ‘contrarian’ because they are mediocre and fall prey to poor analysis and reasoning or lazy biases. (It annoys me that perfectly good traits and adjectives such as skeptic and contrarian have been co-opted by groups carrying out disinformation campaigns, hence the quotation marks.)
So if, for example, you reasonably enough note that atmospheric CO2 increases appear to lag temperature increases at the start of most interglacials, that’s all well and good. There are hypotheses for this that can be examined and tested. But to then conclude that the converse (CO2 forcing leading to temperature lagging CO2) cannot occur is a first year undergraduate-level error. And yet tenured professors who are vocal AGW doubters fall prey to that elementary mistake.
If Spencer and Christy were more honest and humble in their work, they would try to get to the nuts and bolts of the various errors and basic assumptions underlying microwave-derived atmospheric temperatures. I think it could well be an interesting field with information that could be (re)mined from the large amount of data available. But they seem to have taken a much more stubborn position, and hinge their skepticism about AGW on it.
Christy put forward a graph purporting to compare the average of 4 balloon datasets with the average of 2 satellite datasets of TMT in his May 2015 Congressional testimony. Do those results make sense?
Well done Tamino. The nails are in the coffin of the satellite temperature record. The denialists will of course bluff and bluster, but all half reasonable people will desert them now. And that should include all the media except for that owned by Jerry Hall’s husband.
I wonder if anyone is up to the task of trying to construct a temperature record based on the satellite data, or if it is just not worth the effort, given the balloon and surface temperature records seem to be doing just fine.
[Response: I think it’s well worth the effort, and Carl Mears seems to have the expertise and objectivity required.]
Maybe he forgot that lot’s of people would love to check his work but that he did not yet publish it……..
Tamino, tell Christy that he could try a different intersatellite adjustment, specifically give more trust to NOAA-14 and less trust to NOAA-15, and see if this attenuates the trend-break from that of radiosondes at about year 2000.
According to Mears et al (2011) http://onlinelibrary.wiley.com/doi/10.1029/2010JD014954/full the single largest difference/uncertainty between any pair of satellites is right there:
“ Examination of the differences between TMT from MSU channel 2 on NOAA-14 and AMSU channel 5 on NOAA-15 shows a long-term trend difference, with NOAA-15 cooling at a rate of 0.2 K per decade relative to NOAA-14 over the July 1998 to December 2004 period of overlap. This trend difference is not present for the other channel pairs, including, to our surprise, TLT, and is more than 2.5 times larger than the trend difference for any pair of MSU satellites with more than 18 months of overlapping observations. This trend difference is too large to explain using the difference between the MSU and AMSU TMT weighting functions. It is about 100 times larger than the trend difference simulated using HadAT data over the overlap period. The cause is not known and could be a drift in calibration in one or both of the satellites that is not explained by our calibration error model (equation (3)). Since we do not know which satellite is closer to being correct, we treat this drift as an additional source of uncertainty. It is unlikely that the drift is caused by errors in the diurnal adjustment because the magnitude of the drift is similar for land-only and ocean-only averages, which is unlikely to be the case for errors in the diurnal cycle.”
You can also ask Christy what happens if he replaces UAH TMT with NOAA STAR TMT in the multilayer UAH v6 TLT formula.
(Right answer: The trend 1979-2015 goes up from 0.114 to 0.187 C/decade)
Magma, I also think it is more tempting for mediocre researchers to get the extra attention you more or less automatically get as one of few “skeptics”. Not many outside her field had heard about Curry before she started echoing “skeptic” talking” points. Now she gets to testify in congress, be interviewed in media etc. It must be flattering.
Look, I don’t think we have to attribute ill motives to Christy. We need to keep in mind that:
1) What he is trying to do is very hard.
2) He does have a lot of vested interest in his research. It is his entire career, and he is bound to be reluctant to make big changes.
3)He does have a bias that warming does not pose a big threat. All scientists have biases. If they are bad enough they can really hurt your career. However, I think Christy is sincere in his belief and he doesn’t want to see a lot of resources wasted on what he thinks will not be an issue.
Christy’s position is definitely a minority position, and the evidence weighs strongly against it.
Where Christy does deserve criticism is in his presentations to nonscientists. He gives short shrift to the consensus and to the evidence that supports it. In normal science, that would just hurt his career. In post-normal science, the sycophantic denialati latch on like lampreys.
I asked more verbosely at Eli’s thread that references Tamino:
“… hoping someone knows — did the atmosphere change, particularly locally? I look at the DSCOVR imagery showing that dark gray cloud moving around north and eastward from India, day after day.
And I wonder — are balloon data being taken from that area?”
Wondering if the satellites are seeing areas with more crap in the air, particularly since 2000 as India and China have burned much more coal.
Off topic, but I have a question. Could anyone here suggest a quick and dirty way to determine what a July or August high temperature anomaly equal in magnitude to December’s anomaly would look like? Not just adding the number of degrees above meteorology that December showed (6 deg C-ish I think) to the mean August high, but what would August temps that that were as extreme as the December temps (as many SDs from the mean) look like. Would like to do this for Maryland, USA.
How about the June 2015 highs for Seattle? They averaged 9.0°F above the 1981-2010 average. I live in Seattle and I’ve kept a temp log of daily highs/lows,averages and records for 5 years. Come June 26, I was concerned I was getting heat stroke and called a consulting nurse at a hospital. Nothing,but it was memorable to say the least (lol).
Tamino, recall that S & C wrote a paper published in 2003 which compared the MSU TLT with balloon data:
“Intercomparing data records observed by separate coorbiting microwave instruments provides one means of error testing.”
Here are some quotes:
“Radiosonde observations provide pressure-level temperature readings from which a weighted average may be calculated to match the microwave vertical profiles. We use a static weighting function (Fig. 1) applied to
the radiosonde data.”
“Temperatures from each acceptable sounding were interpolated to a 5-hPa
vertical grid and the appropriate static weighting function was then applied.”
“A radiosonde-simulated temperature was generated if the sounding contained at least the mandatory levels. Anomalies were calculated for each sounding (0000 and 1200 UTC separately) based on the 22-yr mean, daily annual cycle for each layer and observed time. For each observing time 10 observations were required to calculate a monthly average from the daily anomalies. Linear interpolation was applied to fill temporal gaps not
longer than 20 days for some levels, though missing data periods of this length were rare. Several stations reported consistently only once per day, 0000 or 1200 UTC, so their anomalies were based on the appropriately
Christy and Spencer’s analysis used their TLT algorithm to create an artificial series of simulated TOA observations from the balloon data. Thus, their analysis never actually tested their algorithm, as the algorithm appeared in both the TLT time series and in the fabricated balloon series.
Then too, the requirement for co-located stations implies that only 4 orbits (+/- 1 orbit around 0000 and 1200 GMT) a day might be used out of 14 each day. Whether that limitation had any impact could also be important, IMHO.
Tamino says the divergence in temperatures since 2000 looks like satellite drift, Christy says it looks like ongoing adjustments to the GISS based temperature record.
Tamino says his selected balloon record supports his position, Christy says his selected balloon record supports his position.
What’s a person interested in the truth to do? It is your blog and you can do what you want with it. You certainly don’t need to respond to every comment requesting additional work on your part, particularly by someone like me who is on your sh*t list. But I would like to see you show us why Christy’s balloon record is wrong, not why yours is correct. Show us why all of the GISS adjustments since 2000 do not result in the divergence. I will ask the same of Christy.
[Response: Christy’s balloon record? Which record? Which records? All I see from him is “average of 4 balloon data sets” with no clue which ones he’s talking about.
As for those “adjustments” to the GISS record, there are a lot more adjustments to Christy’s satellite record, and it turns out they’re a lot bigger. And did you notice that all those “GISS adjustments” do not result in a divergence from the RATPAC radiosonde (balloon) data? Why do you think that is?]
Christy said “the ballon record” supported his version 5.6 record which is different from his current, version 6.0, record.
Maybe Christy is saying that through an amazing co-incidence, “the ballon record” suddenly switched from supporting his version 5.6 record to supporting his 6.0 record.
Balloons always supports Christy’s “satellite record”. It just requires amazing coincidences.
We now have Christy’s reply to the video, courtesy of a post on Breitbart:
The post includes a “sanitized” version of the graph Christy included in his written testimony before Ted Cruz in his Senate hearing on 8 Dec. This particular graph doesn’t mention that it presents TMT data, not the TLT, though Christy is quoted as saying “… the main product we use now for greenhouse model validation is the temperature of the Mid-Troposphere (TMT)…”. Note the use of the royal “we”, implying that this choice is now standard practice, even though Spencer and Christy’s first claim to fame was their finding that the TMT includes some of the cooling trend found in the Stratosphere, which led to their decades of promotion of the TLT as the gold standard for climate detection.
The graph is clearly intended for maximum impact, as it is scaled to make the difference between their satellite/sonde data and their model results appear even larger than shown in the graph presented at the Cruze hearing. Recall also that they “adjusted” the curves by adding a fixed amount to each series in order to force the trend lines to cross zero in 1979, which moves the model curve further upward in the graph, compared with the satellite/sonde curves. Those curves represent anomalies, which are intimately tied to the zero line during the base period and such adjustments break this connection. Furthermore, the curves are said to be 5 year averages, a process which should remove 4 years from the satellite time series, yet the graph shows 36 data points for satellites out of 37 possible from 1979 thru 2015. Christy’s testimony also includes a graph for the tropics, which has 37 satellite data points. Those extra data points are an obvious fabrication, IMHO.
The graph is clearly intended to deceive the viewer, who can not be expected to understand these details or know of the history surrounding the satellite data. I think we know who is spreading disinformation and it’s not the other scientists in the video…
Thanks for the link. The funny thing is that the graph, to me, doesn’t show a very good fit between Christie’s satellite data and his balloon data, with clear periods of discrepancy. The balloon data show a continuing warming trend throughout the period when satellite data supposedly shows a “pause”. Yet they think it’s a good fit!
And now we have Ben Santer and Carl Mears replying to the silliness at the Senate hearing …
[Response: Well worth reading.]
Every time UAH releases a new dataset, they’re acknowledging that the previous set was in some way imperfect, and that satellite measurements aren’t the rock solid gold standard that Cruz (and Curry) claimed. Upon the release of 6.0 Spencer wrote:
“…One might ask, Why do the satellite data have to be adjusted at all? If we had satellite instruments that (1) had rock-stable calibration, (2) lasted for many decades without any channel failures, and (3) were carried on satellites whose orbits did not change over time, then the satellite data could be processed without adjustment. But none of these things are true. Since 1979 we have had 15 satellites that lasted various lengths of time, having slightly different calibration (requiring intercalibration between satellites), some of which drifted in their calibration, slightly different channel frequencies (and thus weighting functions), and generally on satellite platforms whose orbits drift and thus observe at somewhat different local times of day in different years. All data adjustments required to correct for these changes involve decisions regarding methodology, and different methodologies will lead to somewhat different results. This is the unavoidable situation when dealing with less than perfect data.”
I don’t believe I’ve ever seen denialists objecting to adjusted satellite data.
I’ve been playing with the latest UAH v6 data. I used the TMT and TLS time series to build a “TTT” series with the thought of comparing that with the RSS TTT, which is calculated as (1.10*TMT -0.10*TLS). UAH and RSS use different latitude ranges in their data sets, for example, UAH Extra Tropical latitude is 20-90 and RSS Mid Latitude is 25-82.5.
Here’s what I found:
Comparison of “satellite Temperature” trends, degrees/decade
19 Jan 2016
Spencer’s description of their algorithm for each grid point is:
TLTv6 = (1.538xMSU2 – 0.548xMSU3 + 0.01xMSU4)
He didn’t indicate what they do with the AMSU data, which exhibits slightly different theoretical emission weighting profiles…
(Cross posted on Eli’s blog)