Ascension Decending

Ascension Island – Mid Atlantic

Well, here we have a drop in the middle of the Atlantic. But wait, there's more…

Ascension Island Monthly Anomaly and Running Total

Notice that after dropping hard into the 1970’s, we get all of one year of data in the 1980’s? ( 1980 ) It is then followed by a 1991 and 1992 with all zero’s (one year could be a change of “Duplicate Number” with a reset of the running total, but 2?) then in 1993 we get October and November. So that “middle time” is very flaky with a lot of missing data. We “take a rise” back to zero (still down 1/3 C from the start and 1/2 C from the line the ‘tops’ tend to run along). So I don’t have a lot of faith in that last decade…

Maybe we can look at other Islands in the Atlantic to get a clue what’s really happening?

Sao Tome and Principe

Closer to the equator and up close to Africa. Perhaps a bit more land influenced during times of stong offshore winds, but on the edge of the Benguela Current that surrounds Ascension.

Sao Tome And Principe  Monthly Anomalies and Running Total

Sao Tome And Principe Monthly Anomalies and Running Total

Ok, 1/2 C rise to the trend line, but… It is dropping until we splice on another thermometer. Then it rises all the way back to the zero line, we drop one of the two thermometers, then drop the whole record in 1980. (There are records for 1982, 1989 and 1991, but the anomalies are “all zero” so these are not very useful…)

So it confirms the early drop, but then we have a “splice” issue to deal with.

Cape Verde Islands?

While it is a different body of water (North Atlantic) it is somewhat the same type of climate. Mid Atlantic Island near the Equator (within 15 degrees or so.)

Cape Verde Islands Monthly Anomalies and Running Total

Cape Verde Islands Monthly Anomalies and Running Total

Well, gee, that’s a rising trend. Isn’t it?

But what if we look more closely at the splicing and the data? There is one minor bit of “tuning” in this graph. Since a Running Total of Anomalies lags the event, I’ve lagged the graph by one year from the actual events. So when the thermometers change, the graph picks up in the following year when the anomalies from that “hit the graph”.

Cape Verde Islands by Segments Monthly Anomalies and Running Total

Cape Verde Islands by Segments Monthly Anomalies and Running Total

Harumph. So we’ve got a downtrend. We take a “flat” step up on dropping to a single thermometer. Then a final “rise in the end”. But what’s in that rise? A large quantity of missing data:

Year     dT  dT/yr Count (12 monthly anomalies Jan - Dec)
1982  -0.64      0  1     0    0    0    0    0    0    0    0    0    0    0    0	
1983  -0.64   0.18  1  -0.5 -0.1    0  0.3 -0.5  0.1    0    0  0.3  0.7  1.2  0.7	
1984  -0.46  -0.38  1  -0.5 -0.4 -1.3 -0.4  0.7 -0.4  0.7    0    0   -1   -1 -0.9	
1985  -0.83   0.01  1  -0.1  0.2 -0.5   -1 -1.6  0.5    0  0.6  0.4  0.7  0.2  0.7	
1986  -0.83  -0.29  1     0 -1.2 -0.5  0.1  0.5   -1 -0.9 -0.2  0.6    0 -0.4 -0.5	
1987  -1.12   0.44  1     0    0    0  2.4    0  1.6	0    0    0    0  1.3    0	
1988  -0.67   0.16  1     0    0  1.8 -1.7  1.4    0  0.8  0.2 -0.2 -0.4    0    0	
1989  -0.52   0.09  1   0.1    1    0    0    0    0    0    0    0    0    0    0	
1990  -0.42   0.1   1   0.1    0    0    0    0    0    0    0  0.5    0    0  0.6	
1991  -0.32  -0.03  1     0    0    0    0    0    0    0    0 -0.4    0    0    0	
1992  -0.36      0  1     0    0    0    0    0    0    0    0    0    0    0    0	
1993  -0.36  -0.05  1  -0.7    0    0 -0.1  0.2    0    0    0    0    0    0    0	
1994  -0.41  -0.05  1  -0.3    0    0    0 -0.2    0    0 -0.1    0    0    0    0	
1995  -0.46   0.18  1   1.3    0    0    0    0    0  0.9    0    0    0    0    0	
1996  -0.28   0.13  1     1    0    0    0    0    0    0 -1.6  0.2    0    0    2	
1997  -0.14   0.47  1     0    0  2.7  1.3  1.6    0    0    0    0    0    0    0	
1998   0.32   0.05  1     0    0  0.6    0    0    0    0    0    0    0    0    0	
2003   0.38  -0.1   1     0 -1.2    0    0    0    0    0    0    0    0    0    0	
2004   0.28  -0.14  1     0    0 -1.7    0    0    0    0    0    0    0    0    0	
2005   0.13  -0.13  1  -1.6    0    0    0    0    0    0    0    0    0    0    0	

Aside from the rampant number of “zero anomalies” (that usually mean missing data as opposed to the months being identical) notice that we have a gap from 1998 to 2003. So what are we really measuring (and splicing on) after 1986 ?

How About the South Atlantic

As we saw under the South American series, the Falkland Islands are a bit cool, but South Georgia show a bit of rise at the end:

Falkland Islands

Falkland Islands Hair Graph monthly anomalies and cumulative

Falkland Islands Hair Graph monthly anomalies and cumulative

but it, too, cuts off in the early 1980s.

South Georgia

South Georgia Hair Graph monthly anomalies with cumulative

South Georgia Hair Graph monthly anomalies with cumulative

Where we have a long down trend that suddenly reverses at the end. But wait a moment… that end. It jumps from 1982 to 2006… What are we really measuring THERE?

Year     dT  dT/yr Count (12 monthly anomalies Jan - Dec)
1981  -0.98   0.86  1   0.6  1.5 -0.5  0.4 -0.4 -0.7    3  3.7  1.5   -2  1.9  1.3	
1982  -0.12   0.03  1   1.2 -0.8    0    0    0    0    0    0    0    0    0    0	
2006  -0.08  -0.14  1     0    0    0    0    0    0    0    0    0    0 -1.7    0	
2007  -0.23  -0.99  1  -1.7 -1.4    0    0 -2.2 -0.5 -3.1  0.2 -0.9  0.9 -1.9 -1.3	
2008  -1.22    0.6  1     0    0    0    0    0    0    0    0    1  1.4  4.8    0	
2009  -0.62   0.86  1   2.4  0.9  2.6 -1.9  1.9  1.1  3.5 -2.4  0.6    0    0  1.6	
2010   0.24  -0.24  1  -2.9    0    0    0    0    0    0    0    0    0    0    0	

A lot of holes and missing data…

Conclusions

So there is the giant part of the globe, the Atlantic, that’s got nicely positioned islands suited to measuring the temperature. That are showing a very interesting cooling trend. And they substantially are cut off mid-stream.

Now, decades later in some cases, we “take a splice” and start recording anomalies again. And we’ve seen that spices are bad things…

Where there is some “warming” it is strongly tied to instrument changes and happens as something of a “step function” with those changes.

There are global scale oscillations where one part warms and another cools by a degree or so at a time. With no idea what really happened in the Atlantic (and especially the larger South Atlantic) since the 1980’s we have no clue if it continued to cool as “warming” was found in another quarter of the globe.

IMHO, it is Very Important to get vetted and validated records for those islands, from ORIGINAL RAW sources (not “corrected” nor “adjusted”) to fill in this part of the data set. With little idea what’s happening in at least 1/4 of the world, we can have no assurance that “elsewhere” means anything other than one side of an offsetting oscillator. Worse, it looks like we’ve begun to add those islands back in, now that the major ocean currents have swapped again. (We may be selecting specific phases of that oscillation and not realize it. Double booking “trends” that are just “half cycles”. )

The World Ocean seen from above the south pole

The World Ocean seen from above the south pole

Original Image

Pick up a globe and look at it from the bottom. Notice that large empty place between South America, New Zealand and French Polynesia? Notice that “big empty” in the South Indian Ocean? Punctuated by Prince Edward Island, Crozet Island and Kerguelen Island. Where are THEIR data in the GHCN? If you know nothing about 1/3 of the planet, what do you know?

Ships at Sea Monthly Anomalies and Running Total

Ships at Sea Monthly Anomalies and Running Total

About E.M.Smith

A technical managerial sort interested in things from Stonehenge to computer science. My present "hot buttons' are the mythology of Climate Change and ancient metrology; but things change...
This entry was posted in dT/dt and tagged , . Bookmark the permalink.

11 Responses to Ascension Decending

  1. R. de Haan says:

    You are a dangerous man E. M. Smith!
    Wonderful job!

  2. Keith Hill says:

    EM. As Tasmanians, we are in that “big empty” with nothing but Ocean west of us till the East Coast of South America. East of us there is only the South Island of New Zealand and lots more ocean till the West Coast of South America. (Maybe a few assorted little islands as well).
    This is why Cape Grim, on the North West tip of Tassie and in the path of the prevailing Westerlies, is known for the cleanest air and purest rainwater in the world.
    Air pollution and CO2 levels are among the items monitored, but rather curiously, GHCN does not use temperature data recorded there.
    Marrawah is only 30km or so south of Cape Grim and interestingly BoM warns those researching records that even in that short distance, cloud cover is far more prevalent than at Gape Grim. Regional factors again.
    I know your love of islands and you have touched on us in “Aussie Fair Go” etc., but I don’t think you’ve done a hair graph just on Tassie. As I’ve posted before, despite nicely rising CO2, Tasmanian stations have refused to respond with rising “global warming”, even Launceston Airport (now dropped”) – as shown in the graph you posted in “Aussie Fair Go” . A look at all Tassie’s cooling graphs show that it is no wonder GHCN dropped most of our stations like “hot cakes” in 1992.
    Ironically, two of the leading AGW alarmists in the world, leader and deputy-leader of the Greens, Bob Brown and Christine Milne, are two of our representatives in federal parliament. Their favourite phrase is “catastrophic global warming” (man-made of course) ! It seems not to phase them that Tassie’s trend won’t back them up. Cheers.

  3. ditmar says:

    Chiefio, I can’t understand the lack of data for ascension is during the eighties, it was a very busy place see operation black buck @ http://en.wikipedia.org/wiki/operation_Black_Buck. As. is. Grew cosiderably during that decade to house and feed all the groundstaff. I think 1990s ascension bears little semblence to 1980. Ck it out.

  4. ditmar says:

    There would be corresponding developement in the falklands and south georgia. The data for these islands is there somewhere.

  5. ditmar says:

    Finally, the uk are now developing oil fields there(risking a further territorial spat with argentina. I would assume even more developement to support that.

  6. ditmar says:

    If you reintroduce s.georgia, faklands and ascension you would be measuring uhi at the airports which will show a marked increase from pre 82.

  7. vjones says:

    I am sure you are right about the importance of islands. I’ve not looked at many outside the Pacific yet.

    That ‘Ships at Sea’ graph is very interesting [BTW did you see Peter O’Neill’s “Stations at Sea” posting ? – http://oneillp.wordpress.com/2010/03/13/ghcn-metadata/%5D
    It would be interesting to see the locations of the survivors after the dropouts at 1982 when the hair gets ‘wilder’.

    I can probably take a look at that easily and report back.

    REPLY: [ That Peter O’Neill article is just great. I especially loved the simple approach of looking at stations “out to sea” to see if there was an island present… and finding various cities had moved off shore… Thanks for the pointer to it. Folks really ought to read that article.

    The ‘ships at sea’ is an interesting set. It ‘cuts off’ after being thinned out. Yet it is in theory just a place in the ocean where, as a ship passes by, it takes the temperature. So how does it “go out of service”? Are the data stacking up somewhere waiting for the next decade to be updated? It ought not to be a CLIMAT, so it’s very unclear who decides what gets in when. Looking at the v2.inv file (v2.temerature.inv as downloaded from NCDC) they are mostly in the range of LAT from 30N to 66N and LONG from +2 to -145 with one at +164 (lat 34) and most of them -16 to -50. I think that makes it a lot of N.Atlantic and a bit of elsewhere.

    If he oceans are dumping heat and cooling, what the land is doing does not really matter. And if there is a land / ocean oscillator with one warming as the other cools, we’re just fooling ourselves to think we know what’s going on by measuring (largely) northern land. -E.M.Smith ]

  8. Rod Smith says:

    E.M. Saint Helena Island in the South Atlantic used to be a reliable and useful weather source in my day. You might want to check it out.

    REPLY: [ AARRRRGGGG! Just when you think it can’t get any worse….

    OK, thank you so much for pointing me at that. It’s not in the v2.country.codes list from the 28 December 2009 vintage GHCN that I use for a lot of this stuff. So it never made it to my “canonical list of countries” as that was just a print out of the v2.country.codes. Yet…

    Doing a search of the station metadata shows a station at 14761901000 named St. Helena Island.

    A “grep” of the v2.mean temperature file shows data from 1892 (for the “0” duplicate number or “modification history flag”) to 2009 (for the “2” flag).

    So now I have a ‘mini-project” of: Audit the v2.country.codes file… but at least I know now that it’s broken too and that there may well be more countries in the temperature data that I’ve not got “on the list”. Oh, and I get to add St. Helena Island to the list of graphs to do.

    You would think they would keep their data sets in sync. If they think. Which I’m now doubting.
    -E.M.Smith ]

  9. E.M.Smith says:

    OK, it looks like v2.country.codes is missing:

    143 Crozet, Port Aux France
    147 Saint Helena Island
    642 Slovenia: 4 stations. Ljubljana (2), Maribor/Slivn, Portoroz
    701 Base Orcadas

    And the v2.country.codes file has entries for these countries that lack any temperature data:

    164 Melilla (Spain)
    166 Rwanda
    167 Swaziland
    235 Maldives
    426 Antigua & Barbuda
    428 British Virgin Islands
    430 Dominica
    436 Saint Lucia
    437 Saint Vincent and the Grenadines
    439 Turkes and Caicos Islands
    522 Brunei

    But no surprises there. We’d flagged them as missing when we did those continents.

  10. Rod Smith says:

    E.M.: Somehow, I’m not surprised. I see NCDC calls St Helena Country “Malta” (code “MT”) – and the WMO Block # matches the Med. That makes it ‘goofy’ to begin with and splits a WMO block.

    It is really an out of the way place where, if I am not mistaken, Napoleon was first exiled. In my day it was a mainstay of the South Atlantic reports . And FYI, the ship traffic in the SA was (is?) almost non-existant. I remember vividly when NASA was pleasantly surprised that I furnished seven SA ship reports for one of their launches.

    And dredging my memory I seem to remember that the Ascension Is. weather was reported by the US Navy until well after WWII. I don’t know why.

  11. vjones says:

    E.M. Re St Helena be aware of:

    http://data.giss.nasa.gov/gistemp/sources/gistemp.html
    “Non-overlapping records are viewed as a single record, unless this would result introducing a discontinuity; in the documented case of St.Helena the discontinuity is eliminated by adding 1C to the early part.”

    REPLY: [ Interesting “catch”… Now I just need to figure out what to do with it ;-) -E.M.Smith ]

Comments are closed.