Thursday, January 5, 2017

2016 was the hottest year on record for the troposphere

For the troposphere, 2016 was the hottest year on record!

The troposphere temperatures for December 2016 have been released. The lower troposphere is recorded in UAH v6 beta 5 and RSS TLT v3.3. This report also covers RSS TTT for the troposphere (without the "lower") and follows pretty much the same format as previous monthly updates.

For RSS TTT (troposphere), 2016 was the hottest year in the record. Last month was the second hottest December on record.

The lower troposphere (UAH beta v6.05) also showed 2016 as the hottest year in the record. However, December was only the sixth hottest December on record, with 2015 the hottest.

Troposphere temperature (RSS TTT v4) chart


First here is RSS TTT with the latest dataset, version 4. TTT seems to measure more of the troposphere than TLT (that is, it has a greater vertical profile) with less of the stratosphere than the mid-troposphere data (TMT). It shows a higher rate of warming than RSS v3.3 and higher than UAH. Hover the cursor (arrow) over the plots to see the data points, trend etc.

The chart below is the average of the 12 months to December each year from 1979 through to 2016. The annual averaged anomaly was 0.78 °C, which is 0.17 °C higher than the previous hottest year, 1998. The rate of warming is 0.18 °C/decade or 1.8 °C/century.
Figure 1 | Troposphere temperature for 12 months to December (TTT). Anomaly is from the 1979-1998 mean. Data source: RSS

From the RSS website, TTT is derived from TMT and TLS with the formula:

TTT = 1.1*TMT - 0.1*TLS. 

This combination reduces the influence of the lower stratosphere, which is cooling at most locations. TLT gives most weight to the temperatures closer to the surface. TTT gives more weight to the troposhere and less to the stratosphere than TMT does, but not as much to the lowest levels of the troposphere as TLT does. However TTT has version 4, while TLT is still only provided as version 3.3. For a fuller explanation see the RSS website or the July 16 report here.

Below is the TTT chart just for the month of December. The anomaly for December was 0.527 °C, which is 0.101 °C cooler than December 2015. The rate of warming just for December is 0.15 °C/decade.
Figure 2 | Troposphere temperature for the month of December only (TTT). Anomaly is from the 1979-1998 mean. Data source: RSS





Lower troposphere


The rest of the charts are from UAH beta v6.5. This is almost identical to the old version of RSS, which is v3.3, so is likely to be updated at some time. (Other RSS data sets, like TTT are now at version 4.)

The chart below is the annual average from 1979 to 2016. 2016 is the hottest on record by 0.03 C.
Figure 3 | Lower troposphere temperature for 12 months to December. Anomaly is from the 1981-2010 mean. Data source: UAH

Below is the UAH chart for the month of December only for each year going back to 1979. The anomaly was 0.24 °C above the 1981-2010 mean, which was only the sixth warmest December in the record, and 0.23 C cooler than the hottest December in 2015.

Figure 4 | Lower troposphere temperature for the month of December only. Anomaly is from the 1981-2010 mean. Data source: UAH


Comparing recent ENSO years


Below is a chart comparing the strongest El Niño years since 1979, which were followed by a La Niña, just for UAH v6 beta 5. I've included the 2015/16 period for comparison.

Figure 5 | Global mean lower troposphere temperature for strong or moderate/strong El Nino years that were followed by a La Nina. Data source: UAH

La Nina didn't end up happening. The odd thing is that even though it didn't emerge in 2016, the UAH December temperature has dropped back to the level of the December 1998, when there was a La Nina. I still think the UAH data record as a whole would be quite different if satellite drift were better accounted for by UAH - going by comparison with RSS data.

That's all for 2016 from the troposphere, at least as far as temperature articles go. From now on the monthly updates will be for 2017. I might even change the format, for a bit of variety:)


Update from RSS


This news release from Remote Sensing Systems (RSS) just arrived by email. Worth clicking the link because there are a lot of charts:
Using the latest version (Version 4.0) of the “Temperature Total Troposphere” (TTT) dataset, RSS scientists showed that 2016 was 0.31 degrees F warmer than the previous record, set in 1998. The third warmest year occurred in 2010. In addition, 9 out of 12 months for 2016 were the warmest of that month ever recorded in the satellite record. By this we mean that January 2016 was the warmest January, etc., for all months except for May, June and December (the December record was set in 2015). The record warmth was caused by long-term global warming combined with the strong El Niño event that occurred in the winter and spring of 2015-2016. …

…For this work, we used the “Temperature Total Troposphere” (TTT) dataset instead of the more commonly cited “Temperature Lower Troposphere” (TLT) dataset. TTT measures the temperature of a thick atmospheric layer, extending from the surface to about 8 miles (about 13 km) high. We use TTT because TTT has been updated to version 4.0, while version 4.0 TLT is not yet available. (We use TTT instead of TMT because the latter has not been corrected for the influence of stratospheric cooling.) In the upgrade to version 4.0, we improved the method we use to correct for drifting satellite measurement time, leading to more reliable measurements, particularly since 1998. The new version shows more warming than the older version, particularly since 1998.

For more details see: Mears, C. A. and F. J. Wentz (2016). "Sensitivity of Satellite-Derived Tropospheric Temperature Trends to the Diurnal Cycle Adjustment." Journal of Climate 29: 3629-3646.

This paper is available online (open access) http://journals.ametsoc.org/doi/10.1175/JCLI-D-15-0744.1


RSS TLT version 3.3 contains a known cooling bias. We are working to eliminate the bias in the new version of TLT. Even with these known cooling biases, 2016 was a record warm year in TLT v3.3. In fact, 2016 was a record warm year in all RSS tropospheric temperature products (TLT v3.3, TMT v3.3, TTT v3.3, TMT v4.0 and TTT v4.0) 

Here is one of the charts, showing how new records keep being set over time, particularly in 2016. A red dot means a hot record and a blue dot means a cold record. I've added small black arrows down the bottom so you can more clearly see which years are 1980, 1990, 2000 and 2010.



As you'd expect, there were more records of both types in the early years of the data set. There have not been any cold records set since 1993. In 1998 there were a ten record hot months, all but two of which were broken again in subsequent years. In 2016 there were nine record hot months.

Added by Sou at 10:45 am AEDT 6 Jan 2017 (11:45 pm GST 5 Jan)





19 comments:

  1. Thanks Sue. Kevin Cowtan has a useful page on his University of York website that allows you to check latest trends and error margins. You may already know of it, but some of your readers may find it useful: http://www.ysbl.york.ac.uk/~cowtan/applets/trend/trend.html

    The full trend in RSS TLT v3.3 from Jan 1979 - Dec 2016 is now +0.14 C/dec with 95% confidence interval of +/- 0.6 C.

    This places it well within the range of the global land and ocean surface data sets, which show around +0.17 (+/- 0.4) C/dec since 1979.

    Kevin's calculator still uses the official UAH v5.6, which I guess they will update after v6 is formally published. UAH v5.6 is updated to Nov 2016 and shows +0.15 (+/- 0.6) C/dec warming. I calculate that UAH 6.5 (beta) shows about +0.12 (+/- 0.6) C/dec from 1979, which would still place it within the surface temperature range.

    Interesting that most 'sceptic' sites are only showing the satellite data in relation to the 1998 el Nino peak, rather than to the 1979 start date.

    ReplyDelete
    Replies
    1. DavidR, that RSS v3.3 is also old, but they have not yet made a TLT product of their new v4.0 (which is published). The latter is likely to give a larger warming rate. In other words, the deniosphere will likely switch back from RSS to UAH - the one with the lowest slope.

      P.S. Are you sure you meant .6 confidence interval? I think you left out a zero several places :-)

      Delete
    2. Marco,

      Thanks, yes you're right. I'm mixing up decadal and century trends. The confidence intervals should be +/- 0.04 and 0.06 respectively, according to Kevin Cowtan's site.

      Delete
  2. Count down to the Deniers calling the satellite temp data set fraudulent 5, 4, 3, 2, 1 :-)

    ReplyDelete
  3. This is really nice work, as always. It would be interesting to see how this compares to the CMIP3 and CMIP5 projections for lower troposphere anomalies.

    ReplyDelete
  4. I've just added an update about a news release that just arrived from RSS. Notice the caution about the cooling bias in RSS TLT v3. This would probably apply to UAH v6 beta as well, since it's almost identical to RSS v3.

    ReplyDelete
    Replies
    1. Thanks for that Sou.

      Notice the caution about the cooling bias in RSS TLT v3. This would probably apply to UAH v6 beta as well, since it's almost identical to RSS v3.

      It would seem so. In fact here's a simple but unambiguous demonstration that UAH v6.x beta is biased cool. Just compare it directly with RSS TLT 3.3 which has a known cool bias.

      From RSS news release Jan 05 2017:

      RSS TLT version 3.3 contains a known cooling bias. We are working to eliminate the bias in the new version of TLT.

      Whoopsie, UAH.

      Delete
    2. Well, its not like other scientists have had to point out a cooling bias to Spencer before.

      Delete
    3. It seems Spencer and Christy agree that RSS V3.3 has a cooling bias, but suspect a different reason: "my UAH cohort and boss John Christy, who does the detailed matching between satellites, is pretty convinced that the RSS data is undergoing spurious cooling because RSS is still using the old NOAA-15 satellite which has a decaying orbit, to which they are then applying a diurnal cycle drift correction based upon a climate model, which does not quite match reality."

      Delete
    4. So why does the all-new, singing and dancing UAH v6x beta *match* the known-to-be-biased RSS TLT v3.3 product almost exactly?

      It's very strange. Perhaps S&C just forgot those old blog posts...

      Delete
    5. BBD, one reason for the low UAH trend is the infamous "Cadillac calibration choice". Spencer and Christys " treatment" of the largest intersatellite uncertainty around, is to cherrypick NOAA-15 with the lowest trend and discard NOAA-14 data "due to drifts".
      The RSS and STAR teams don't do this kind of arbitrary science, they keep both satellites, since they honestly don't know which one is right.

      I dont know why the satellite teams don't use radiosonde data for validation of significant choices. Maybe they wish to be satellite-only self-sufficient...?

      IMO radiosonde validation would suggest that NOAA 14 is the right choice:
      https://drive.google.com/open?id=0B_dL1shkWewaOEd5TUlTYWlMUW8

      Delete
    6. Exactly Olof, and something *sceptics* should be reminded at every turn when they refer the the Sat temp data.
      To my mind the sonde comparison clearly shows the AMSU sensor to have a cold bias.
      RSS are merely being pragmatic in splitting the disconnect between 14 and 15.

      Delete
  5. Thanks for the vital work you're doing.

    There's been an interesting piece just posted on the Congressional war on science here: http://www.ozy.com/pov/the-congressional-attack-on-science/74438 A few of the comments beggar belief, though: One guy blithely states that "if nothing else, the hockey stick has been completely discredited by the facts" . . . a statement so counterfactual that all I could do was listen to the sound of my one jaw dropping.

    ReplyDelete
    Replies
    1. It doesn't take much effort to tell lies about science. It does take a lot of work to do science.

      (I added my 2c worth.)

      Delete
    2. I knew a painter who did very very large walls in commercial buildings. Some passer by would always let him know with the epithet and the pointing of a finger.

      "You missed a bit!"

      Little did they know it was done on purpose so the quality inspector would have something to report.

      Otherwise he would go looking for orange peel or other subtle defects that were very difficult to fix.

      Bert

      Delete
    3. Reminds me of a friend in university. He used to write a report, save it, deliberately then add three points that the prof would object to and submit the second. Then when he got feedback he printed out the original.

      Said it saved lots of time.

      Delete
    4. @ sou

      the time and effort required to refute bullshit is an order of magnitude larger than it is to produce it

      Delete
    5. Bert, have you told us that story before? I'm sure that I've read it somewhere, because I was painting architraves last week and picking bits of bristle from the paint, and remembering this story and wondering if I should use a brush or roller for the walls, exactly because of the orange peel which is conspicuous in the paint that someone had previously applied. I also have a friend whose ex is a painter, and I was wondering if I should leave a little spot uncovered somewhere so that I didn't get the micro-scrutiny that might otherwise occur...

      Delete

Instead of commenting as "Anonymous", please comment using "Name/URL" and your name, initials or pseudonym or whatever. You can leave the "URL" box blank. This isn't mandatory. You can also sign in using your Google ID, Wordpress ID etc as indicated. NOTE: Some Wordpress users are having trouble signing in. If that's you, try signing in using Name/URL. Details here.

Click here to read the HotWhopper comment policy.