Lucia, Anthony, Roy Spencer and David Stockwell (my, there’s a growing list of analytical blogs) have already posted on March 2009 temperatures and trends. While I’m not first off the mark on this, I’ll be the first to post tropical trends – something that I do from time to time. (This requires a little bit of heavy lifting since I need to download the entire gridded data sets from NOAA, CRU and GISS to determine their tropical averages.) Maybe I’ll keep this script private from the others for a while 🙂
The plot below shows the following series with their trend (deg C/decade in brackets): UAH T2LT (0.014), RSS TLT (0.09), CRU (0.103), NOAA (0.117), GISS (0.132), HadAT 850 hPa radiosonde (-0.001) and the following two series not shown UAH T2 (-0.013) and RSS TMT (0.066). The satellite and radiosonde series are the tropical troposphere series where Santer says that there is no statistically significant difference between observations and models for either the UAH or RSS data sets (this being the subject of a submission by Ross and I to IJC – on which still no response.)
For comparison, the multimodel mean trends (deg C/decade) reported by Santer et al were surface (0.13), T2LT (0.215) and T2 (0.199).
All recentered on 1979-1997. Troposphere anomalies divided by 1.2 (per Christy).
![Reblog this post [with Zemanta]](https://i0.wp.com/img.zemanta.com/reblog_e.png)
Climate Audit 
89 Comments
If the image represented the inflation-adjusted prices of a bunch of banking stocks one would have to say the value of the stocks is no higher now than it was during 1980. Further, one would be wary of betting on consistent price rises any time in the near future given the recent downward trend. In the finacial markets the simplest and best measure of value is the current price. I suspect a few climatologists would be well advised to be mindful of this when trying to derive meaning from trends in surface temperature. Would I be correct in asuming that these temperature series are not inflation (urban heat island) adjusted?
.
The kinds of conclusions one could draw from Steve McIntyre’s plot of recent temperature trends versus the kinds of conclusions one could draw from the US Government’s plot of temperature versus CO2 content, as illustrated in one of NOAA’s 2008 public reports, might seem at first glance to be at odds with each other. Who is right, the world wonders?
.
.
The US Government (NOAA):
.
.
.
.
Steve McIntyre citing UAH T2LT, RSS TLT, CRU, NOAA, GISS, HadAT 850 hPa radiosonde:
.
Re: Scott Brim (#2), Scott – NOAA chart is for the global record (presumably based on GISS data? – a reference to the source would be useful) whilst Steve is posting on tropical troposphere temp. anomaly.
The NOAA chart appears on the cover of NOAA’s Unified Synthesis Product, Global Climate Change in the United States, 2008. Quite obviously, Steve’s graph and the NOAA graph are not looking at precisely the same temperature anomaly. It is the range of implications that each one carries, and the interplay of those implications, which are of interest from both an accounting and a statistical perspective.
Re: Scott Brim (#5),
Looking forward to your reply to #6:
This is consistent with my understanding and it is puzzling why observations do not match predictions. (Usually that means the model is wrong. But hey, …)
Steve, any chance you could get your private R code, which you should not be intimidated into releasing, by the way, to plot lines showing the multi-model mean trends reported by Santer, calibrated to start at 0 in 1979?
I believe Dr. Christy recommended 1.3 in the tropics.
Re: Ross McKitrick (#7),
Now that’s code that while convenient, I wouldn’t even bother to ask Dr. Steig for. It sounds like SteveM is feeling the heat.
Scott Brim, the problem with the NOAA graph is that since 1934 was the warmest year ever in the USA, and since the USA used to have the best ground weather station network, the peak 1998 and the trough in 1934 implies to me that NOAA are completely wrong.
Re: Bruce (#10), ever? I don’t take kindly to this sort of a-historical alarmist nonsense. 😉
Tropical temperature anomaly trend slope differences for surface to T2 and T2LT, the standard error for those trend differences and the AR1 correlations of the residuals are the calculated values of interest to me.
Doing it for the “Santer time period” and for the time period to present would be a bonus. Putting that all in context of what the models predict a further bonus.
Re: Kenneth Fritsch (#11), see my article for further information on the change in amplification depending on the time period over which it is measured.
w.
why are the tropo anomalies divided by 1.2?
Re: 007 (#12), That’s about the expected aplification ratio, to get the implied surface change
Thanks, Andrew. I thought expected amplification was higher than that.
To wit: “For comparison, the multimodel mean trends (deg C/decade) reported by Santer et al were surface (0.13), T2LT (0.215) and T2 (0.199).”
Re: 007 (#14), It’s the number John Christy gave Steve via email some years ago now, and based on some straight forward calculations of the short term amplification observed during El Nino’s, I’d say it is probably correct.
Re: Andrew (#16), And, I might add, Santer’s numbers result in a slightly higher factor of ~1.6.
I discovered an interesting parallel between 1972-1983 and 1998-2009, only that the 2009 El Nino did not happen..
For the co2 thermometer see this script
http://home.casema.nl/errenwijlens/co2/co2_lt_noaa.R
For readers who haven’t been visiting all that long here at Climate Audit, last summer, it was bender and myself who pegged the graphic from the NOAA 2008 Unified Synthesis Product as being the one most likely to replace the Hockey Stick as the iconic symbol of AGW alarmism — if the stick ever became widely recognized by the general public as being thoroughly discredited.
.
My point in placing the two visuals next to each other was to spawn the following question: Is the basic conclusion conveyed by NOAA’s 2008 graphic — that rising CO2 concentrations are directly responsible for an increase in global mean temperature — actually supported by the data which is being illustrated visually through Steve McIntyre’s plot? (Why or why not?)
.
To answer that question, it is necessary to know in some detail how, and from what data sources, each visual was generated; and also the pedigree of the data which was used as the basis for each graphic. I am particularly interested in the validity and pedigree of the data plotted on the NOAA graphic for the period of 2000 through 2008.
Re: Scott Brim (#18), Hi Scott – it is also on the cover of the second public draft Jan09. I checked the credits but couldn’t find a reference – might have missed it though:
http://www.climatescience.gov/Library/sap/usp/default.php
I remain intrigued by the apparent patterns in monthly temperature trends. Here are the RSS and UAH tropical linear trends, by month:
The trends appear to make a step change down in March thru June then return to the higher range in July thru February. If that apparent pattern is real then what 1979-2008 atmospheric behavior drove it? Is that behavior consistent with model output?
Re: David Smith (#21), Its looking more and more like you are writing up the paper I was actually planning on doing…But I have no clue as to the answers to your questions, sad to say.
Re: David Smith (#21),
Because the sun is strongest in January? Occam’s razor…
Re: Gunnar (#29),
That might explain why it was warmer then, but why would that specifically have an effect on the trends?
Re: David Smith (#21),
I’m encouraged to see that others find that an annual temperature trend is not detailed enough to adequately describe climate change.
.
This is the latest version of the NOAA graphic taken from page 17 of the January 2009 draft of the USP, Global Climate Change in the United States:
.
.
.
.
This is the text of the NOAA graphic’s citation taken from page 166 of the January 2009 draft of the USP:
.
.
.
Re: Scott Brim (#24),
Global temps are a terrible metric for monitoring climate change. The data sets are corrupted by UHI effects, microsite issues (see surfacestations.org) and arbitrary adjustments by various keepers of temperature data, all of whom are rewarded by scaring the public and policy makers into a crisis mentality.
It would be far better to submit a graphic of ocean heat content overlayed with CO2. The Argo Network data set is freely accessible to all and therefore less subject to mischief. This means the ocean heat content calculations are far more reliable than the global temp data.
Re: Ron Cram (#36), Hi Ron – FWIW: From following along I’ve formed the impression that the OHC metric is also problematic due to poor quality and coverage of longterm sea temp. records. For instance the current NODC chart of OHC follows a similar shape to the Global temp. charts above:
http://www.nodc.noaa.gov/OC5/3M_HEAT_CONTENT/
Argo data looks like it should be good but this system is a recent (2000 on, with 3000 floats deployed by Nov2007) implementation:
http://www.nodc.noaa.gov/argo/overview.htm
http://www.argo.ucsd.edu/
Re: curious (#37),
You are correct that Argo data is recent, but we have enough of a record to know the AGW hypothesis needs to be modified. According to Hansen, a radiative imbalance would create warming oceans year over year. The oceans have not warmed since 2003. In other words, with atmospheric data you need 100-150 years to know much of anything about a trend in relation to natural climate variability. With the oceans, the radiative imbalance shows up year over year because the excess heat is stored in the oceans. So you do not need a long record to see the observations do not match the theory.
Re: Ron Cram (#38), For info re: recent data – 2005 to date CO2 graphs here:
http://www.esrl.noaa.gov/gmd/ccgg/trends/
Re: curious (#39),
Thank you for the graph on CO2. What I would really like to see is a graph that overlays CO2 on an ocean heat content graph from about 1988 on. I think it would show good correlation during the 1990s and a real disconnect after 2003.
Sorry my comment was short earlier. I was heading out the door. Regarding the chart you linked on NOAA, there is no question ocean heat rose in the 1990s. Prior to that, however, the sea surface data had its own problems as Steve has documented here. A warming bias was introduced by changes in instrumentation and other factors.
The data we have from Argo is the least problematic of all and it shows no recent warming. It did have a cooling bias due to a problem in the fall rate of the floats but that bias was adjusted out, as the NOAA page mentions. Even with the adjustment, the data shows no warming for the last several years and recent cooling (since late 2007, I believe).
People who say the Argo data is not high quality (if anyone is actually saying that) are trying to turn the world on its head. It is the most reliable data we have on climate change. In the law, there is a saying. If the facts are on your side, argue the facts. If the law is on your side, argue the law. If the facts and law are against you, attack the prosecution. People who attack the Argo data without any evidence have become advocates and not scientists.
I suggest adding a CO2 line to the Tropical Troposphere chart. Then truncating the NOAA graph for the same time period and comparing them.
.
Reid, I think it would be more useful if Steve could convert the Tropical Troposphere data to degrees Fahrenheit, then re-reference the anomaly datum to the 1901 – 2000 average mean temperature of ~ 57 degrees F as taken visually from NOAA’s chart, then replot the data to the same or similar scale as the 1980 onward section of the NOAA graphic.
.
There may be more steps involved here to go where I think Steve ought to be going, but I’m not the expert on the low-level details of this data. However, I do think that one picture could potentially be worth 1,000 words, at a value of $1,000,000,000 per word.
Smith and Reynolds 2004 shows that this data incorporates the Folland and Parker bucket correction, but one can see that by looking at the graph — the giveaway is always the WWII temperature anomaly starting before 1939.
Is it just me, but does S & M 2004 look like a promising way of adjusting Antarctic temps?
JF
Query on the NOAA chart submitted by Scott. Why does it show 2005 as the hottest year on record? I thought that was reserved to 1998? Also, why is 2008 missing, which would otherwise show a substantial drop on 2007??
Re: Stephen (#28), Yeah, and weren’t several years in the 30s also in the top 10? I love NOAA. I’m getting so much for my money from them. Pretty pictures, and I’m sure something else.
Mark
Ok, missed the y axis label. Sorry. However, there is probably still a simple answer. The graphs are basically showing an increase peak to trough. Jan is getting warmer, while June is cooling.
A quiet sun will result in less heating of the oceans’ surface, which would lead to less evaporation. The result of this would be a decrease in water vapor. Water vapor is helping to reduce swings in temperature. Less water vapor, more peak to trough.
The warmers claim there is a ~30 year delay between GHG increases and warming. They need this delay to explain why warming since 1880 is ~0.7 C when it is supposed to be ~2.2C, given that the radiative forcing of GHGs has increased by 75% of the effect of doubling CO2.
So what is NOAA even trying to prove by plotting CO2 and temperature year-by-year and then claiming that any discrepancies are just “due to natural processes, such as the effects of El Ninos and La Ninas”?
Re: Dave Brewer (#33), True, a much more relevant, but still highly misleading image would have been model output.
Steve,per your analysis of Tropical temperatures and of proxies, you might find the following sediment analysis of tropical Lake Bosumtwi, Ghana to be of interest (6°30.3′N 1°24.5′W):
Atlantic Forcing of Persistent Drought in West Africa
T. M. Shanahan, J. T. Overpeck, K. J. Anchukaitis, J. W. Beck, J. E. Cole, D. L. Dettman, J. A. Peck, C. A. Scholz, J. W. King; Science 17 April 2009: Vol. 324. no. 5925, pp. 377 – 380, DOI: 10.1126/science.1166352
Interesting their finding that the current drought is not “anomalous” relative to three millennia. By inference, neither are the Atlantic sea surface temperatures anomalous relative to the last three millennia.
Further to Shanahan et al on drought vs Atlantic temperatures,
See Africa trapped in mega-drought cycle, 16 April 2009 by Catherine Brahic, noting:
Hi Ron – I think we are in agreement over Argos. I wasn’t criticizing it, merely noting it is a very recent addition to the datasets. IMO CO2 mapped to any measure on these very short timescales (ie aligning a year’s CO2 change to the same year’s OHC or temp or whatever) is a bit of a red herring. I think there is a relevant paper by Conway from 1994 (I’ve seen a web version but I haven’t got a .pdf including figures):
http://www.agu.org/pubs/crossref/1994/94JD01951.shtml
Re: curious (#42),
Without access to the paper you linked to, I do not understand your point. Why do you think it is a red herring? CO2 is supposed to cause a radiative imbalance. Everyone agrees a radiative imbalance would be seen in OHC in year over year changes. You do not need 30 years of data to see it. Of course, every new year of data gives us more to work with but we have more than enough already to draw some pretty interesting conclusions. OHC calculations are not limited to the time period of Argo data, it is just that we have much more confidence in OHC after Argo is global in 2003.
Instead of trying to scale gross atmospheric CO2 to OHC, another approach might be to graph annual mean growth rate of CO2 to OHC. The years atmospheric CO2 goes up the most should show the most warming, correct? Certainly it holds true in 1998, but it did not in 2005. I think a graph showing this would be hit viewers right in the face.
http://gallery.surfacestations.org/main.php?g2_itemId=27408&g2_imageViewsIndex=2
http://gallery.surfacestations.org/main.php?g2_itemId=27418&g2_imageViewsIndex=3
.
Just out of curiosity, I rescaled Steve’s tropical troposphere temperature anomaly plot to degrees Fahrenheit, and then I overlayed it on top of the 1980-2007 segment of the NOAA global mean temperature anomaly plot.
.
This was done by graphically adjusting the two plot image files to the same horizontal and vertical scales for Year and for Temperature Anomaly (F), respectively, and then by merging the two image files together.
.
.
Re: Scott Brim (#46), You know, you could just dig the data up yourself at the various places and plot a graph where you perform whatever procedures you like (some might quibble that you need to recenter the NOAA anomalies) instead of all that effort to stick two images together. That being said, nice work.
Re: Scott Brim (#46), Hi Scott – do you think if you now overlay CO2 level, at an appropriately chosen scale, the new plot informs your point from #18 above:
(You provided the NOAA graphic’s data sources in #24 and Steve gave his sources in the lead post).
.
Re: Andrew (#47),
What I did in merging the two image files together at the same scale was admittedly a brute-force approach which only took a few hours to accomplish and which did not have the finesse of directly recentering the NOAA anomalies using their own source data.
.
However, unless one has at one’s disposal direct access to an end-to-end auditable trail of data, software code, and analysis process description, one cannot really know where NOAA’s visual graphic actually came from. Nor can one do an absolutely accurate and precise job of performing a time-frame recentering.
.
.
Re: curious (#48),
.
Let’s note that the vertical scale and the vertical plot position NOAA chooses to use for illustrating a rising trend in CO2 concentration is purely an arbitrary decision on their part. Both the scale and the position have been arbitrarily chosen so as to precisely match the plot of their global mean temperature anomaly trend, with the obvious intent of visually reinforcing the NOAA graphic’s basic message — CO2 is responsible for recent global warming.
.
I decided not to include NOAA’s plot of CO2 concentration because it might interfere visually with my rough comparison of the two temperature anomaly trends, and because including it tells us nothing directly about possible causal relationships among GHG concentrations and an apparent continuous rise in global mean temperature.
.
I say “apparent” because when I look at the rough visual comparison of the two trends as illustrated in the merged plot, two questions come into my mind: (1) From a science perspective, why is there an apparent significant divergence between tropical troposphere anomaly trend and global mean temperature anomaly trend, and (2) from an auditing perspective, is the source data being represented in the NOAA graphic accurate and defensible, i.e. what is its pedigree from an audit trail perspective — which of course one cannot determine unless one has direct and unfettered access to an end-to-end auditable trail of data, software code, and process description for how the NOAA graphic was produced.
.
Hi Ron – The reason I referenced the paper I found in 42 is because it highlights (small) spatial variations in CO2 concentrations whereas it is usually reported as a single global value. It also notes the variations cannot be explained by changes in fossil fuel emissions and proposes pretty significant changes in the size of NH and SH CO2 sinks. – Abstract:
I’ll try and find more recent info. soon.
My point re: red herrings being that if AGW is supposedly a direct relation between OHC (or other metric) and (global) atmospheric CO2 level, falsifiable by a single years exception, then it seems that there are no shortages of examples which would do this – including the NOAA graphic Scott posted. I’m still not sure how Scott expected the tropical temp. anomalies to track the global ones but acknowledge this could be me missing something. Apologies if this is the case.
Re: curious (#51),
Let’s assume for purposes of argument that current theories about CO2- induced positive feedbacks acting through water vapor mechanisms are indeed operable in the earth’s climate system, and that general and continuous warming of the earth’s atmosphere on a global scale is the eventual result.
.
If these theories are actually correct, and if I understand these theories correctly, the tropical troposphere is where the most conclusive evidence ought to be found that the hypothesized feedback mechanisms are indeed operative in the earth’s climate system, and work pretty much in accordance with theory.
.
If this is so, then I think it perfectly appropriate from a scientific perspective to expect that patterns of trends in tropical troposphere temperature anomaly should roughly match patterns of trends in global mean temperature anomaly, and to do so over decadel time spans.
.
We should not expect a precise match, but we ought to at least expect a rough match using some reasonable standard of comparison based upon the predictions of CO2-induced positive feedback theory, while also considering the possible influence of natural climate variation over these decadel time spans.
.
A rough, first-cut comparison of NOAA/NCDC’s plot of global mean temperature anomaly with Steve McIntyre’s plot of tropical troposphere temperature anomaly — both plots covering approximately the same three-decade time span — indicates an apparent significant divergence between the respective patterns of anomaly trend.
.
If one is faithful to the principles and ethics of the scientific method, this apparent divergence should cause one to start asking some serious questions and to start doing some serious investigation:
.
— What is the pedigree of the source information?
.
— Is there actually a significant divergence?
.
If the source information is determined to be reliable and defensible, and if the divergence in the respective patterns of temperature anomaly trend is determined to be significant, then what implications follow concerning the validity of the CO2-induced positive feedback theory?
.
I now ask a further question concerning the software-driven climate models now being used as the primary scientific basis for making CO2-induced global warming predictions: Are these climate models capable of directly predicting trends in tropical troposphere temperature anomaly; and if so, should these predicted anomaly trends not be compared with any reliable information now available concerning the actual trend of temperature anomaly within the tropical troposphere?
.
LET me endorse Scott Brin (and bender’s) perception that the above temp/CO2 graph is (or has) replaced the Hockey Stick as an alarmist icon. I have seen it in a number of warmist posts.
But like Scott, its provenance has eluded me. GOOD to have it sourced to NOAA/NCDC publications…but as for its ultimate source, I’m still waiting to know myself! THANKS
as per Orson (Brin and bender) here is wikipedia
The post cites “This figure, based on Meehl et al. (2004)”, which reads in full:
Meehl, G.A., W.M. Washington, C.A. Ammann, J.M. Arblaster, T.M.L. Wigleym and C. Tebaldi (2004). “Combinations of Natural and Anthropogenic Forcings in Twentieth-Century Climate”. Journal of Climate 17: 3721-3727.
Googling this full citation and “image” yields three instances of our suspect graph on the first page (among other illustrations.
http://www.globalwarmingart.com
zfacts.com
bobtisdale.blogspot.com
AND another four on the second page of results
wikimedia.com
Thus, I believe we have a source for the mystery chart: Meehl, et al, (2004).
To save some time for those who are stopping by my website to check the link that Orson provided, the posts in which I referenced Meehl et al (2004) are “AGW Proponents Are Two-Faced When It Comes To Solar Irradiance As A Climate Forcing”…
http://bobtisdale.blogspot.com/2009/01/agw-proponents-are-two-faced-when-it.html
and “Climate Modelers Reproduce Early 20th Century Warming With The Help Of Outdated Solar Forcings”
http://bobtisdale.blogspot.com/2009/02/climate-modelers-reproduce-early-20th.html
Orson: Which mystery graph above are you attributing to Meehl et al (2004)?
Bob asks “Orson: Which mystery graph above are you attributing to Meehl et al (2004)?”
AND
Re: Scott Brim (#24),
It now appears that my attribution is close in one or two respects – but not spot on.
The time frame is right; the temps are about right, but the first is smoothed and has too many other data streams.
Close but no cigar? I think….
Roy Spencer’s group offers estimates of daily global temperature at various levels of the atmosphere. These are satellite-derived and cover the globe (except for the Poles). The nature of the data is such that they do not precisely measure temperature at one altitude but rather they cover a range of altitudes with different weightings.
Among the data offerings are global channel 5, which is weighted towards an altitude of 4.4 km ASL, and channel 6, weighted towards 7.5 km. These correspond with 600 mb and 400 mb pressure levels and lie, on average, in the global troposphere (though they likely cover the lower stratosphere to some extent).
Here are a few plots and a few things I ponder. I hope that anyone with insight into these will post explanations.
First is a plot of the average daily temperatures at 4.4 km and 7.5 km for 1999-2008.
I have recentered them for display purposes. The plot shows the warmest global temperatures occur in the Northern Hemisphere’s summer (boreal summer). I attribute this to the greater amount of land in the Northern Hemisphere, which warms more easily than ocean. No surprise there.
One mild surprise to me is that the 7.5 km minimum seems to occur several months earlier than the 4.4 km minimum (see “B”). Why is that?
Another is that the difference (see “A”) seems to vary over the year, as shown in greater detail on this plot –
The difference is about 1K greater in boreal summer than in boreal winter. Why is that? If the altitudes of the measurements are unchanged then the variation in temperatures seems to indicate a change to a greater global average lapse rate in boreal summer than in winter. Yet it seems like the global troposphere would have a greater moisture content in summer than in winter, which I think would lead to a lower lapse rate, not a higher lapse rate.
Here’s a third plot, showing the annual average temperature differences between the two channels for 1999-2008
The temperature spread seems to be growing. Why is that?
I realize that the answer may be that these apparent patterns are actually “illusions” created by the limitations of the measurement methods and by use of a the coarse (global) view of phenomena which are actually regional in nature. But, I thought I’d offer these for comments and hopefully learn the explanations from other readers.
Re: David Smith (#58),
Here’s a working link for daily global temperature , I hope.
Re: David Smith (#58), Careful when using the daily data-as I understand it there are processing steps it has not gone through which must be done to get it to climatological quality like the monthly products.
Re: David Smith (#58),
Don’t the models project the trend to be the reverse?
Could you be seeing the effect of land use changes?
Re: David Smith (#58),
if they are not illusions as for Andrew, the 7.5 minimum let me think of a stratospheric warming at the southern latitudines when the Sun starts to shine there. Maybe you should check the stratospheric channel.
A larger lapse rate during the boreal summer is what I expect.
Over most of the continent interiors, you have dry, well mixed air masses during summer, while moist convection occurs mainly in the southern part of the northern hemisphere (SE Asia, SE North America, Sahel).
During winter the lower troposphere is decoupled over the continents from the upper levels.
So I think that dry convection over boiling continents is what could make the difference with respect to the southern hemisphere.
Just a thought.
Re #60 Thanks, and understood. The answers may lie in the lack of subsequent processing. It’s far from clear to me, though, how subsequent processing might affect the particular types and magnitudes of the patterns in question.
Re #61 Barry, I think the models project moistening of the middle troposphere with warming, at least in the tropics. Regarding land, there are questions in my mind as to whether the troposphere above large land masses behaves differently in AGW than does the troposphere over tropical oceans and whether land use effects are large enough to have a meaningful impact. At this point, though, the key question for me is whether the apparent patterns in #58 are “real” or simply an illusion for reasons like Andrew mentions. They may be illusions. This is a learning exercise, I hope.
I need the annual rain fall data and average daily temperature for Papua New Guinea from 1999-2008
I’ve asked one of the Daily Earth Temperature authors for some basic information on the nature of their product, how it might be tied to the monthly UAH troposphere anomalies, limitations, etc. As the story comes together I’ll post my notes here.
What intrigues me is that daily satellite data may reveal higher-frequency or off-calendar patterns which get lost in the monthly averaging. Perhaps there are learning opporuntities for a layman like me, clues about how the troposphere handles its heat.
Any apparent patterns in the data may prove to be illusionary, an artifact of the technology, but learning nothing more than that would be worth the time.
Here is the daily tropospheric temperature anomaly in the reported 4.4 km (600 mb) region. The 1999-2008 average for each day was subtracted from the respective 2008 daily value, so this is a daily anomaly and should not show seasonality.
There appears to be some oscillatory behavior on about a 25 to 40 day cycle. What causes that? Is it real? MJO oscillations? It looks like a warming followed by a almost-as-rapid cooling and my guess is that the warming (heat) comes from lower in the troposphere, via precipitation. Is that right? Is it from tropical deep thunderstorms? Is the apparent cooling radiative, perhaps a spreading of the warmth across a large, radiating region? Why does the warming and cooling appear to stay within bands, almost like limits? Again, is any of this real or is it all artifacts?
Here’s the reported 7.5 km data for 2008:
It shows similar behavior which, if real, indicates vertical transport, which is expected, But, unlike the increasing anomalies at 4.4 km, the 7.5 km data shows flatness across the year. Why is that? Is that an indication of an increasing lapse rate which ultimatley self-corrects?
Here the difference between the levels, which illustrates the different trends:
Interestingly, a shift seems to have occurred during a particularly sustained warming around February. What was that? What caused the apparent sustained drift from April to October? Is this simply random behavior?
Again, I recognize that this data may be severely limited in its quality, or ultimately useless, but the exploration deserves a little time.
Re #66 I asked Dr Spencer for his view on the apparent oscillations in the first plot (600mb level) and whether it is real or some artifact of measurement. His view is that it is probably real, would be tropical in origin and would involve heat transfer from the tropical oceans to the atmosphere.
I assume the apparent oscillations involve thunderstorm clusters, or maybe some phenomena downstream of the storms, and perhaps the MJO, though I could not get the oscillations to correlate with any measure of MJO. It’s an interesting pattern to me, almost like the tropical warm pools pulse heat into the atmosphere.
The troposphere appears to be quickly cooling from its recent peak. Here’s the plot of daily data through August 5:
If past behavior is an indicator of future activity then I expect that the temperature will continue to rapidly fall, leveling in the zero to -0.2C anomaly range.
And, in the next 60 days or so, another spike will occur.
I’m still unable to relate these periodic rapid changes to any atmospheric phenomena. All ideas on what this might be are appreciated.
A wild guess: Tide. Spring tide and the peak in Your picture coincident with this:
http://www.27east.com/story_detail.cfm?id=224470&town=East%20Hampton&n=Experts%20explain%20why%20tide%20swings%20have%20been%20unusually%20high
And:
http://www.dailyexpress.com.my/news.cfm?NewsID=66398
And
http://news.outlookindia.com/item.aspx?663265
If the tide can affect the top meters of the oceans will it for sure have some impact of the atmosphere.
http://en.wikipedia.org/wiki/Tide
I’ve found the solution…Cloud Ships.
I’ve found the solution…Cloud Ships.
The temperature of the troposphere continues to drop quickly –
If past behavior repeats then I expect the drop to end at either red dot one or red dot two.
It’s almost like the troposphere behaves modally. Perhaps there are modal states of heat input (possibly tropical thunderstorms) or heat radiative output (states of cloud cover or water vapor?). Or maybe this reflects state changes in the movement of heat from the tropics to the extratropics. Or, maybe it’s my imagination.
I belive this is a atmospheric lunar tide effect.
http://www.scientificamerican.com/article.cfm?id=does-the-moon-have-a-tida
Click to access HaganResRev06.pdf
It fits perfect to the thermal anomaly.
The temperature differance of +/-0,3 K is equal to a change in height of only +/-50m, with a std lapse rate of 6 K/1000 m.
Re: Cold Lynx (#73), Very interesting – thanks for the links.
Presumably the temperature oscillations affect the rate of IR loss into outer space. If the frequency or amplitude of the oscillations change over time (for whatever reason) then tropospheric heat removal might be affected. Perhaps a solar/climate connection? Wild conjecture, of course.
Re: David Smith (#74), putting aside, just for a moment, solar/climate connection, if you wish to look for a causal link to the sharp rise of recent tropospheric T, you should look at daily maps of tropo-temperature. If heat springs up suddenly on a restricted region, you could have a resticted area of research.
I don’t know if satellite daily maps or data are available. In any case, one can try with common weather analyses.
Re: Paolo Mezzasalma (#76), Thanks, Paolo. A problem may be that the magnitude of the temperature difference (about 0.3C) is likely too small to detect from maps. It may be evident from the daily satellite data but that may be geographically coarse and is not publically available, to my knowledge.
This whole thing could be a data or measurement technique artifact, of course. But, I’d sure like to understand it.
Re: David Smith (#77), as a first step, I would not look for a few tenths of a degree of a sudden global warming but for a regional warming of a few plain degree. That would ba a much easier job and reanalyses could be the tool to accomplish it. I have no idea about the findings, I was just stimulated by your thoughts.
If the sudden warming is global or tropic-wide, then a burst of regional convection could be dismissed…perhaps…
I belive the diurnal solar induced variation is amplified by the lunar influence.
When sun and moon is in line will the atmosphere be “thicker” towards the sun and that will change the altitude = temp in the measurement.
The maximum temperature is during the day when the atmosphere have it´s tide and will therefore be able to be warmer due to larger thickness.
VERY intresting when reading this document: “Correcting the MSU Middle Tropospheric Temperature for Diurnal Drifts”
Click to access Correcting_MSU_Channel_2_Temperature_for_Diurnal_Drifts_IGARSS.pdf
A diyrnal modification is
A warm july trophosphere measurement might just be an unusual strong lunar influence.
Other oscillations with periods in the 30 to 50 day range include length-of-day, Madden-Julian and atmospheric angular momentum. These may be related and there is some conjecture in the literature that these may be traceable to solar processes.
It does seem plausible that a sudden temperature change in the troposphere, even 0.3K, would be noticeable in angular momentum and possibly Earth’s rotation period.
Our tropospheric temperature continues its rapid decline towards either red dot 1 or red dot 2, as forecast:
I think it will level at red dot 1 for a week to ten days.
On the same topic, here’s the plot of daily anomalies and the 60-day average since January, 2004:
It visually hints at cyclic behavior combined with longer-term variation over that period. The anomaly swings seem to be within a fairly well-defined band.
Here’s a plot of the difference between the 20-day average anomaly and the 60-day average:
Removing the 60-day average, I hope, acts to detrend the time series. Using the 20-day average, I hope, removes much of shorter-term variation. What’s left give the visual impression of cyclic behavior. It’s a crude technique but I think it captures the essence of the situation.
Finally, here’s a plot of the duration of those cycles for 2004-2009:
It seems to indicate a common cycle of 45 to 50 days, which I believe is similar to the MJO. If that is correct then perhaps the cyclic anomaly behavior is related to the active/inactive phases of the MJO in the Indo-Pacific warm pool, which I think has a typical period of 40 to 55 days. Alternately, perhaps the MJO does not drive the temperature swings but instead shares an unidentified common cause with the tropospheric temperature anomalies. Dunno, but I wish I did.
Re: David Smith (#80), The raw data is here , which comes from UAH .
The 0.2K spikes (or valleys, if one prefers) may play a significant role in how Earth removes its heat. In a way they appear like mini El Ninos, moving heat from the surface into the troposphere and then into outer space. If their duration, frequency or amplitude change, does global heat removal change? Does the global weather pattern, and possibly heat removal in remote regions, change?
Re #80 The tropospheric temperature leveled at +0.1 to +0.15 versus the forecasted 0 in #80:
The question is, when it breaks from this level, will it break sharply higher or sharply lower?
The troposphere has ended its rest at +0.15C anomaly and has begun a sharp rise upwards –
The spike is apparently related to MJO behavior in the tropics and, if history holds, should peak (green dot) in the next several weeks at +0.45 to +0.60C, then decline back to a base anomaly for several weeks. We’ll see.
So, the satellite-derived temperature anomaly for September may well exceed the anomaly for August. August had four “peak” days, four “transition” days and the remainder were “base” days around +0.15C. September may see the opposite, with relatively few base and transition days and the remainder in the peak region.
The rather repetitive base-peak-base behavior of the tropospheric temperature is interesting. The spikes represent the removal of a lot of heat from the planet’s surface. Do the period and amplitude of the spikes vary naturally, and why? How would increased GHG affect this pattern? Does the apparent rapid cooling during the downleg indicate that the tropical upper troposphere (the heat radiator, so to speak) indicate that the upper troposphere has ample ability to radiate away heat even with today’s higher CO2? Many questions to my mind on what seems to me to be an important topic.
Re: David Smith (#83),
Haven’t checked for sure, but at first glance, it would also appear that the overall general trend looks to me like it shows the transaition from La Nina to El Nino as well. Have you overlaid this with an ENSO chart to see if the general upward trends match?
RE #83 An alternate hypothesis is that the base-peak-base behavior reflects tropospheric modes of heat removal efficiency and is not directly related to the amount of heat entering the troposphere from the surface. Dunno.
Jonathan, I think you are right. In addition to ENSO variation one may see, with imagination, a hint of volcanic dust suppression of temperature in early to mid 2009. But, this is like reading tea leaves or tree rings or silt – caution is needed to avoid bringing ones preconceptions to the effort.
Re #85 Here’s the tropospheric plot plus an ENSO index (ONI, scaled):
There’s a pretty good hint of a troposphere / ENSO relationship, as expected.
Re: David Smith (#87),
Thanks for the graph David. There certainly appears to be a relationship, but of course, it could also be just that; an appearance only.
Re #87 I forgot to mention that the ONI plot is shifted to the right by four months (ONI activity preceeds the tropospheric response)
There certainly appears to be a relationship with both MJO and ENSO.
But what drives this waves?
I say tide.
The suns diurnal tide have a the strongest but still weak gravital force impact at sunrise and sunset.
This small but percistant solar gravitation acceleration will in the morning have a horisontal composant that will bring on a small force on everything on the planet. Fluids like air will accelerated horisonatlly. Yes that is wind.
In the morning east going. During the day just little bit less gravitation, and in the evening slow down the wind by a force westwards.
Wind start in the morning continue over the day and slow down in the evening.
Familiar to most of us but we are all thought that this is only by convection.
That is another source but tide is one important as well.
This solar tide accelerate the air up to of about 5 -6 m/s during the day.
And make the wind slow down with about the same figures wind during night time.
MJO travels with about this speed.
With a lunar component will the atmospheric tide change even further.
More wind and by that more evaporation will induce more vapor into the atmosphere. Most important is that the wind will lower static pressure and help tha cloud formation. Vapor that form clouds release a big amount of heat and that is the heating of the trophosphere. Visible in the graphs. High troposheric temperatures in August is a sign of strong tide.
Unfortunatly is this increased cloud formation cooling the earth significantly.
An even moore intresting aspect is the inclination of the horisontal gravital force.
Earth tilt of max 23.5 degrees make the solar tide induced wind will come from south west in the NH summer. And from north west in the NH winter.
The lunar inclination will modify these direction and that have a strong impact of local climate.
A period with less tilt and a lunar tilt not in favor will reduce the south west composant and cool the NH.
Forgot some of the links, about tide in general,
http://en.wikipedia.org/wiki/Tide
Especially Laplace´s tidal equation
http://en.wikipedia.org/wiki/Theory_of_tides#Laplace.27s_tidal_equations