May 09 2008
Antarctic Temperature Trends
There has been much discussion lately about the recent “cooling trend” in the blogosphere. RealClimate has offered up a bet that there won’t be any long-term cooling. Roger Pielke Sr. offered his commentary. James Annan covered it not once, but twice. And William Connolley had a go at it as well.
This week offers a different cooling article in the Journal of Geophysical Research by Monaghan, et al. They looked at the Recent variability and trends of Antarctic near-surface temperature. Anthony Watts has posted on it, mostly quoting from a press release. Interestingly, the image he shows is from a NASA satellite study (which looks wrong) while the paper in question uses surface stations.
Monaghan et al. use 15 stations to compute the temperature trends over Antarctica. I have copied the specific stations, as well as their latitude, longitude and elevation from the original paper to the table below.
| Station Number | Station Name | Latitude | Longitude | Elevation (m) | Country |
| 1 | Faraday/Vernadsky | 65.3 | 64.3 | 11 | UK/UKR |
| 2 | Bellingshausen | 62.2 | 8.9 | 15 | RUS |
| 3 | Orcadas | 60.8 | 44.7 | 6 | AG |
| 4 | Halley | 75.5 | 26.7 | 30 | UK |
| 5 | Novolarevskaja | 70.8 | 11.8 | 119 | RUS |
| 6 | Syowa | 69.0 | 39.6 | 21 | JAP |
| 7 | Mawson | 67.6 | 62.9 | 16 | AUS |
| 8 | Davis | 68.6 | 78.0 | 13 | AUS |
| 9 | Mirny | 66.6 | 93.0 | 40 | RUS |
| 10 | Casey | 66.3 | 110.5 | 42 | AUS |
| 11 | Dumont D’Urville | 66.7 | 140.0 | 43 | FRA |
| 12 | Vostok | 78.5 | 106.9 | 3490 | RUS |
| 13 | Scott Base | 77.9 | 166.8 | 24 | NZ |
| 14 | Amundsen Scott | 90.0 | 0.0 | 2835 | US |
| 15 | Byrd | 80.0 | 119.5 | 1515 | US |
Antarctica has a total area of about 14 million square kilometers. Or just about 1 surface station for 1 million square kilometers. For comparison, there are 1221 surface stations in the United States, which has an area of about 10 million square kilometers. This equates to one surface station for every 8000 square kilometers. If the United States had the same station density as Antarctica, there would be only 21 surface stations for the entire continental US.
I wanted to have a look-see at what the temperature data from these stations looked like. So I went to the website maintained by the British Antarctic Survey and downloaded the data for all the stations. Not surprisingly, there were many gaps in most of the temperature data, so doing a straight comparison between stations was out of the question. Therefore, I decided to write a quick routine to fill in each of the temperature time series based upon its surrounding stations.
The procedure was simple, and each of the points that needed filling fall into one of four classes:
- It was part of the time series where there was a short break in the record (< 7 months), and there were no data from surrounding stations
- It was part of the time series where there was a short break in the record, and there were data from surrounding stations
- It was part of the time series where there was a long break in the record (> 7 months), and there were data from surrounding stations
- It was part of the time series where there was a long break in the record, and there were no data from surrounding stations
For case 1, the missing data points were linearly interpolated and filled in. This method is only used because of the need to fill in the data. For extremely short periods of missing data (1 month), this method provides a quick way to estimate the temperature based upon only persistence.
Case 2 uses both the persistence method about, plus an average of the anomaly data from surrounding stations. For each bad point, I looked at only the surrounding stations that had valid temperature data. I then found the monthly anomaly at each of the stations. Then a weighted average was computed, where the weights were proportional to the inverse of the distance between the stations. This means that the closer stations will have a larger influence that stations further away.
The method used for Case 3 is exactly the same as case 2, except it does not use the persistence because there is too much time between good data points. Unfortunately, this is the method that was used most of the time.
Case 4 fills the values with NaNs (not a number). This is the fallback, and was only used at the extreme beginning of the time series.
I’ve included 4 of the reconstructed temperature anomalies below. The first 3 are typical of those used in the study. The last one is used in the study because it is isolated. The problem is that the station stopped taking data a long time ago. I filled in the data using the methods described above. Monaghan et al. used model results to help fill in the data.
In all cases, the red line is the original temperature anomaly, and the black parts are where the time series was filled in. (Click images for bigger, better graphs)
|
|
|
|
As can be seen, there is very little trend at any of the stations. The actual linear trend was not calculated, but I would doubt it’s statistically significantly different than zero. The variance of each of the time series is large compared to any systematic change in temperature.
The Byrd station is interesting because it stopped taking data in the 1970s, yet this station is still used in the reconstruction of Antarctic temperatures. As I said above, this is because there are only 15 surface stations used in this reconstruction and it is extremely isolated from the others.
Tomorrow I’ll hopefully combine these together to get one temperature time series for Antarctica. Then I can compare it to the results of Monaghan.
References:
Monaghan, A. J., D. H. Bromwich, W. Chapman, and J. C. Comiso (2008), Recent variability and trends of Antarctic near-surface temperature, J. Geophys. Res., 113, D04105, doi:10.1029/2007JD009094., A. J., D. H. Bromwich, W. Chapman, and J. C. Comiso (2008), Recent variability and trends of Antarctic near-surface temperature, J. Geophys. Res., 113, D04105, doi:10.1029/2007JD009094.
(No Ratings Yet)
Loading ...Related Posts:
26 Responses to “Antarctic Temperature Trends”
Leave a Reply
To reduce spam, comments are automatically closed 30 days after the last comment. If you would like to comment on any closed thread, please use the contact form at the top of this page.
The NASA pic Watts uses is (a) wrong (obviously, as you point out; it was probably generated by some PR droid) and (b) obsolete. Wiki has the up to date one: http://en.wikipedia.org/wiki/Image:Antarctic_Temperature_Trend_1981-2007.jpg
Is Watts just ignorant or deliberately misleading?
Errrm… and why are you saying odd things about the T trends, when the T trends of the stations have already been analysed for you?
http://www.nerc-bas.ac.uk/public/icd/gjma/trends2006.col.pdf
William, recently a couple of big holes have been blown in Watts’ efforts:
1) He paid a visit to Asheville and got a 2-day briefing and tour from the NCDC. I can only imagine how many times he was asked “Do you have any more questions or concerns, Anthony?”
2) Atmoz did a comparison of Watts’ favorite Surface Station from Hell (on the UAriz campus smack in the middle of a bunch of pavement and buildings) with the new nearby CRN station and found a near-perfect trend alignment.
So for now Watts would really rather blog about other things. FYI he’s pretty much a dyed-in-the-wool solar crazy. To answer your specific question: Both.
BTW, in case you didn’t read that far I wanted to point you to the end of the current post where he refers to volcanoes having a substantial role in warming the Antarctic.
Atmoz, you’re looking at the wrong Monaghan et al paper. You want this one.
“Is Watts just ignorant or deliberately misleading?”
Uhhhh, maybe because the recent picture doesn’t match ATMOZ lack of trends either??
Notice that the most recent NASA picture shows definite warming trends. Yet, most of the data show no trend or cooling in the interior.
As an exercise you should look for papers on geothermal and active volcanos in the Antarctica, including the new ocean floor one up current of the peninsula!!
Collapsing ice shelves and warming trends are around active geothermal areas. But, naaaaah, couldn’t have anything to do with it!!! Just a concidence.
I believe that Byrd is one of the longest records, thus they give your the Byrd.
Kuhnkat, Watts is both ignorant and misleading; we know this because of the history and content of his website.
We have an undersea volcano to the east of the Antarctic Peninsula, somewhere near Larsen B. How does this explain the surface temperature trend on the peninsula, let alone the retreat of ice shelves on its western side?
As an exercise you should find the evidence to support you conjecture.
Atmoz, could you comment on some of the comments here? They are a bit baffling.
First, the impression given by your account of Arizona was that the siting had made a material difference to measured temperatures. What is Bloom talking about?
Second, the impression given by the graphs above is a total lack of warming in the Antarctic. Is this right? If so, is it consistent with the models and with a forecast of overall warming?
Thanks
atmoz,
I’m not getting the part about the photo AW linked to the nasa
study. The caption under that indicates its purpose.
What Am I missing here?
Anyway, always of fan of the way you look at the data. Keep
up the good work
If you visit the Earth Observatory website, you’ll notice the previous and updated image of Antarctica is respectively based on data provided by “Josefino Comiso, NASA-GSFC” and “Joey Comiso, GSFC” - likely one and the same person, and more than a “PR droid”.
At
http://earthobservatory.nasa.gov/Newsroom/NewImages/images.php3?img_id=17257,
NASA published:
“This image shows trends in skin temperatures—temperatures from roughly the top millimeter of the land or sea surface—not air temperatures. The data were collected by the Advanced Very High Resolution Radiometer (AVHRR) sensors that were flown on several National Oceanic and Atmospheric Administration (NOAA) satellites. The data come from the AVHRR’s thermal infrared channel—a portion of the light spectrum we can sense as heat but that human eyes cannot see. This image shows temperature trends for the icy continent from 1982 to 2004. Red indicates areas where temperatures generally increased during that period, and blue shows where temperatures predominantly decreased.”
At
http://earthobservatory.nasa.gov/Newsroom/NewImages/images.php3?img_id=17838,
NASA updated by saying:
“Scientists from NASA’s Goddard Space Flight Center have been working for several years to create and refine a satellite map of long-term temperature change in Antarctica. This image illustrates long-term changes in yearly surface temperature in and around Antarctica between 1981 and 2007. (An earlier version of this map is pictured in a previous posting on the Earth Observatory.) Places where it warmed over time are red, places where it cooled are blue, and places where there was no change are white.
Making a long-term record out of data from different sensors is challenging because each sensor has its own quirks and may measure temperatures a bit differently. None of the sensors were in orbit at the same time, so scientists could not compare simultaneous observations from different sensors to make sure each was recording temperatures exactly the same. Instead, the team checked the satellite records against ground-based weather station data to inter-calibrate them and make the 26-year satellite record. The scientists estimate the level of uncertainty in the measurements is between 2-3 degrees Celsius.”
Given the stark difference in satellite produced imagery (and likely the data upon which they are based), are we experiencing ARGO deja vù by changes in sensor technology? In view of the foregoing, I think a thorough review of ground station data, particularly for the years 2005-2007, would be most important.
Nick,
keep looking, that is one.
Fred,
the measurements available indicate that the Antarctica is increasing ice mass. Depending on whose numbers are used you will get a slight warming to a slight cooling. As the averages are mostly below freezing you just won’t see a lot of melting except around the edges where you have interaction with sea water. The winds are circular around the pole protecting it from warmer winds.
The peninsula that is mentioned so much sticks out into the ocean and has a bit of geothermal activity including volcanoes. What else Nick missed are studies showing geothermal RELATIVELY close to the surface under the range inland of the Ross ice shelf. He also missed at least 3 other active volcanoes, 2 of which are in the area of the Ross Ice shelf.
The polar currents rotate around the pole clockwise. The volcano Nick DID mention is then UPSTREAM of the famous Peninsula.
http://vulcan.wr.usgs.gov/Volcanoes/Antarctica/Maps/map_antarctica_volcanoes.html
http://www.volcano.si.edu/world/region.cfm?rnum=1900
http://www.sciencedaily.com/releases/2004/05/040527235943.htm
http://www.smithsonianmag.com/science-nature/antarctica.html
http://news.bbc.co.uk/2/hi/science/nature/7261171.stm
http://nora.nerc.ac.uk/2004/
Another source for temp data:
http://www.antarctica.ac.uk/met/gjma/
Then again, maybe Anthony used that picture because he ALREADY HAD IT!! Here is a post he did on ANOTHER new volcano that was found back in January!!
http://wattsupwiththat.wordpress.com/2008/01/22/surprise-theres-an-active-volcano-under-antarctic-ice/
Under the Pine Island Glacier, next to the Ross Ice Shelf and upstream of the Peninsula.
Here is a VERY interesting study on glacier streaming and rift systems done in the range close to Ross Ice Shelf and Pine Island.
http://pubs.usgs.gov/of/2007/1047/ea/of2007-1047ea031.pdf
Atmoz your data above shows there ain’t no warming in antarctica
Atmoz, nice work. Antarctica is a very large, very neglected part of the world. It’s a shame we have so little data for it and that what we have is weighted so low in some graphs.
moshpit: I’d bet money that Watts added that caption in later. The stealth corrections are a favorite of the Axis of Weasel.
“The actual linear trend was not calculated, but I would doubt it’s statistically significantly different than zero.”
So … no warming?
The recent satellite record shows the earth back to the zero baseline too.
No warming globally. None. Zilch.
You warmenizers are good at insults when the data destroy syour cults theories.
Steve Bloom:
The paper you cite is a bit weird. How did Monaghan go from uncertainties of +/- 0.37, 0.57, 0.67, 0.59 K/century for the seasonal (DJF, MAM, JJA, SON) trends, to an uncertainty of +/- 0.34 K/century for annual trends? How did they handle uncertainty in the IPCC model outputs?
Frank, I didn’t cite anything. I was just pointing out to Atmoz that the paper he analyzed wasn’t the one the one referred to by Watts and the press release. Just to be clear, the most recent paper (the one I linked) did make use of the results Atmoz analyzed.
On the substance, I’m confused by a lot more than that, in particular the claims about the models.
TCO,
I dunno, I went from here to there and found the note beneath the picture. In either case, I think the self correcting nature of the blogosphere is a good thing. Anyway, I like this place. Even bloom and I get along here.
[Reply: The caption was there the first time I visited.]
Ok. This time, the pole-smokers did not smoke pole. But we know it is their general behavior. Not that there’s anything wrong with that, nowadays…
ha TCO,no food fights here. Atmoz, is student and has more important things to do than clean up after us.
Atmoz admires my manliness. He saw me take Steve McI and rip his equivocating tits, almost literally, when the pussy was lying about Tucson airport.
[Reply: Ook.
]
I’m fairly new to the climate sites and blogs, and I’m trying to develop my own opinion of the state of the science and how I personally want to react.
There are a number of comments on this site about how to not alienate the public such that their sensibilties or push them into the denialist camp. There’s even some talk about the attitude at Real Climate.
This entry started as a look at data from Antartica. Instead of answering the THREE different requests to confirm that the data showed no temperature rise and comment on that point, or replying to the large number of links and data posted by kuhnkat related to geothermal / volcanic activity in the region of the ice shelves, you guys chose to post:
“Watts is ignorant or delibrately misleading”
“Watts is a dyed in the wool solar crazy”
“Watts is ignorant and misleading.”
Watts is a “weasel”.
McIntyre is a pussy with tits.
Are you seriously wondering why the public is turned off by your message in person or when looking on the net????????
Forget answering the questions, don’t comment on the posted data links and lets just insult everybody en masse. Sad.
The first graph above clearly has a positive temperature trend. The next two graphs are not obvious. They have small trends, one positive and one negative. According to the link in the second comment, the trends are not significant at the +90% level. The trends in other Antarctic data show positive trends, especially in West Antarctica.
Also, I think you’re confusing this blog with others, but I mainly focus on small topics (e.g. surface station temperature interpretation) or meta-topic (e.g. the role of science in society). I’ve decided that my policy on blog comments would be to allow pretty much everything, and let readers evaluate their usefulness.
Sometimes it’s not worth it to respond to comments. My time is worth something. And if I don’t respond to certain questions or comments it’s because I hope someone else will have the time that I lack.
Folks, we have now placed the US AGO temperature measurements on the GTS. We have spotty measurements for the six sites listed on:
http://space.augsburg.edu/ago/status.html
for the past 15 years. A student has done some preliminary work on temperature trends and we presented at AMS in January. I don’t have the paper on the internet yet but will try to get it roughed out for your viewing.
Any one interested in these data can contact me from the web page.
Interesting post but the GISS June anomoly seems to show amajor cooling evnt in the interior of Antarctica http://data.giss.nasa.gov/cgi-bin/gistemp/do_nmap.py?year_last=2008&month_last=6&sat=4&sst=0&type=anoms&mean_gen=06&year1=2008&year2=2008&base1=1951&base2=1980&radius=1200&pol=reg
Seems to be major cooling there