Africa - Kilimanjaro
Africa – Different Elevations. I think this matters…
Hot but Variable
We have already seen that Africa is hot pretty much wherever you look. There is not a lot to be gained from heading to the beach vs. a very hot African Plain. Yet there are some tall mountains.
When we look at Africa, we must remember that coverage is sporadic. Revolutions happen. Countries drop out for years at a time. There are some large chunks left un-covered from time to time, even now. Yet we can still see some consistent changes in thermometer locations with altitude. Not as dramatic as the “4 on the Beach” of California, but still there are patterns to see.
All of this has made the Africa “by Altitude” analysis a bit more complicated than the Pacific or the Americas. With each analysis we find a new trick, and with Africa it is no exception. We will start with a straight contenent wide “by Altitude” chart, but where we end up took a while to figure out. While there is still more to do in Africa, I think I now have enough to make a posting with some value to it.
Africa – By Altitude
Notice that this chart has asymmetrical altitude bands. The gradations are smaller at the low altitudes than they are at the top. In the middle, I go by 100 m, then a 500 m jump to 1000 m, 2000m, then everything above 2000 m. We have the usual start with one thermometer in one place (in this case somewhere near, but not at, sea level), but by 1880 when GIStemp begins time, we are well into the mountains. Almost 12% are above 1000 meters and up to 2000 meters. By 1899 decade ending we have some above 2000 meters, though that “over 2000 meter” traunch tends to stay about 1% until today. The 500 m to 1000 m band also is fairly stable, hanging about 11% from 1899 onward. Of most interest is the 1000 m to 2000 m band. This starts at 11.9% and jumps to 17.5%. For most of history, it hangs in that 17% – 20% range, rising to 27% for 2 decades. But about 1959 it drops and, with a bit of a wobble, keeps on dropping to 13.1% in 2009 Decade Ending.
At the same time, we can see the “below 500 meter” bands gaining. The 100 m to 500 m bands add the most station percentages. Those hot plains of Africa, not cooled by an ocean breeze nor mountain snows. Oddly, that “below 100 m headed to the beach band” that grows in other countries, shrinks in Africa. I can only guess that being near the beach in Africa is not as hot as being in the central Sahara, the Congo Jungle, or the Kenyan Plains.
It would be very interesting to do a map of “distance to ocean over time” for Africa (and for Europe, too…) since I suspect that is the key driver for thermometer changes in both those continents. Europe moving toward the water to gain warmth from the Gulf Stream and inland seas, while Africa moves away from the oceans to avoid moderating winds. But that will have to wait for another day (and some more complex programming along with a “distance to ocean” metric for each station…) A nice “dig here” for anyone in need of a project or thesis.
[chiefio@tubularbells Alts]$ more Therm.by.Alt1.Dec.ALT
Year -MSL 20 50 100 200 300 400 500 1000 2000 Space
DAltPct: 1849 0.0100.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
DAltPct: 1859 29.0 19.4 25.8 25.8 0.0 0.0 0.0 0.0 0.0 0.0 0.0
DAltPct: 1869 51.3 20.5 20.5 7.7 0.0 0.0 0.0 0.0 0.0 0.0 0.0
DAltPct: 1879 36.9 36.9 4.8 1.2 2.4 0.0 0.0 6.0 11.9 0.0 0.0
DAltPct: 1889 24.7 19.9 17.8 4.1 2.1 0.0 1.4 12.7 17.5 0.0 0.0
DAltPct: 1899 24.2 22.0 15.2 4.4 0.9 1.4 3.0 10.9 17.8 0.2 0.0
DAltPct: 1909 21.4 18.3 14.0 6.2 1.3 2.6 2.9 10.8 21.5 0.9 0.0
DAltPct: 1919 21.0 15.7 12.7 7.2 2.3 6.2 4.8 9.5 20.0 0.6 0.0
DAltPct: 1929 16.8 15.6 11.6 6.8 2.9 6.8 5.0 10.6 23.0 0.8 0.0
DAltPct: 1939 15.5 13.5 10.4 7.1 2.7 6.9 4.5 11.0 27.8 0.6 0.0
DAltPct: 1949 14.8 11.0 8.5 6.5 4.9 9.2 5.2 11.9 27.3 0.7 0.0
DAltPct: 1959 16.5 10.3 8.7 5.0 7.8 13.3 9.0 14.4 14.6 0.7 0.0
DAltPct: 1969 17.3 9.2 8.4 5.9 7.8 12.5 8.8 12.8 16.6 0.8 0.0
DAltPct: 1979 17.9 9.9 8.7 6.4 8.5 12.0 8.6 11.3 15.7 1.0 0.0
DAltPct: 1989 15.8 9.5 8.8 6.8 9.2 12.5 9.0 11.8 15.7 1.0 0.0
DAltPct: 1999 14.9 8.9 9.1 7.8 11.3 14.1 9.6 10.2 13.2 1.0 0.0
DAltPct: 2009 13.7 8.0 9.1 8.2 12.3 13.9 9.6 11.1 13.1 1.1 0.0
For COUNTRY CODE: 1
How to approach Africa
I do not yet have any “by country” analysis for Africa. I may come back to that a bit later. For now, though, we can see “Altitude Bias” in the broad numbers and we can see a hint of what I expect to find is the movement away from cooling oceans (that is a big “dig here!”).
While most other places have had their change happen during The Great Dying of Thermometers, in Africa we have the reduction of mountain coverage happening earlier and it is the move from the beach that happens recently. Since GIStemp anomaly maps are sensitive to thermometer count and location changes (as shown in the early benchmark of STEP3) this will bias the African anomaly maps. The major questions are “How Much?” and “Where?”. I would not expect much change on a continent wide basis (there is not much change in the thermometer percentages, and much of that happens early on and inside the baseline period). It is most likley that a “by country” analysis will be needed to understand the African Anomaly maps today. In particular, looking for things like, oh, movement away from the coast and into the Sahara for North African countries. So we’ll do a couple of small “country” studies as examples.
Africa – By Latitude: Does it help?
The “by Latitude” chart does show movement into the Sahara latitudes with a significant increase in the 10N to 15N band on the southern edge of the Sahara and more growth in the 15N to 30N mid-Sahara. Together, they both account for 39.7% of African thermometers:
Look at ./Lats/Therm.by.lat1.Dec.LAT (Y/N)? y
Year SP -40 -30 -20 -10 10 15 30 35 40 -NP
DecPct: 1849 0.0 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 100.0
DecPct: 1859 0.0 9.7 0.0 0.0 12.9 12.9 0.0 0.0 64.5 0.0 100.0
DecPct: 1869 0.0 25.6 0.0 0.0 2.6 15.4 7.7 2.6 46.2 0.0 100.0
DecPct: 1879 0.0 11.9 0.0 0.0 19.0 0.0 9.5 40.5 19.0 0.0 100.0
DecPct: 1889 0.0 9.9 3.4 0.3 8.9 0.0 11.6 33.9 31.8 0.0 100.0
DecPct: 1899 0.0 7.1 5.1 4.9 18.9 1.1 9.7 35.3 18.0 0.0 100.0
DecPct: 1909 0.0 4.6 8.3 8.2 23.7 4.8 15.4 21.6 13.3 0.0 100.0
DecPct: 1919 0.0 3.8 9.9 8.3 28.0 10.3 17.7 13.8 8.3 0.0 100.0
DecPct: 1929 0.0 3.2 10.9 13.5 27.6 11.1 14.2 12.2 7.3 0.0 100.0
DecPct: 1939 0.0 2.7 12.1 13.4 26.2 9.7 13.3 14.9 7.6 0.0 100.0
DecPct: 1949 0.4 6.0 12.4 13.7 25.4 11.6 14.3 11.1 5.1 0.0 100.0
DecPct: 1959 1.4 3.3 7.9 11.9 35.0 17.4 13.9 7.2 2.0 0.0 100.0
DecPct: 1969 1.4 4.3 9.0 11.9 34.5 16.1 14.8 6.4 1.7 0.0 100.0
DecPct: 1979 1.6 4.7 9.9 10.9 31.3 18.9 14.5 5.5 2.6 0.0 100.0
DecPct: 1989 2.1 6.6 11.2 9.4 27.9 17.2 14.6 6.5 4.4 0.0 100.0
DecPct: 1999 1.6 3.6 7.3 8.5 24.1 19.9 17.1 11.0 6.8 0.0 100.0
DecPct: 2009 1.3 2.7 5.5 7.8 22.0 19.7 19.3 13.3 8.3 0.0 100.0
For COUNTRY CODE: 1
We also see the 35N to 40N band (that narrow stripe near the Mediterranean Sea) shrink from 31.8% when GIStemp begins time to 8.3% now (a nice recovery from the 1.7% in 1969 decade ending, but still relatively low. The recent percentage increase is likely a statistical artifact of survivor bias in The Great Dying of Thermometers circa 1990.)
The clear loser is everything south of the equatorial band from 10 S on down. Now accounting for only 17.3% of all thermometers in Africa down from 32.5% at the 1949 decade ending. And the band from 30N to 35N on the Mediterranean side of the Sahara; shrinking from 40.5% in 1879 to a low of 5.5% in 1979, then have a bit of a recovery to 13.3 % now.
We exit with 61% of African thermometers in the 10 degrees each side of the equator and up to the 30N line of the Sahara (which is where most of Africa is located), but far different than the original distribution near coasts at each end. Further, the drift in the middle is toward the Sahara.
Trying to find the “altitude bias” inside that overall trend will be a bit difficult, and ought to focus on individual country studies in those countres with significant mountains.
Algeria and Morocco
For example, looking at Algeria and Morocco we have two different behaviours. For Algeria:
Look at ./Alts/Therm.by.Alt101.Dec.ALT (Y/N)? y
Year -MSL 20 50 100 200 300 400 500 1000 2000 Space
DAltPct: 1859 0.0 20.0 40.0 40.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
DAltPct: 1879 4.3 23.9 26.1 8.7 4.3 0.0 0.0 10.9 21.7 0.0 0.0
DAltPct: 1889 0.7 14.4 20.5 8.2 4.1 0.0 2.7 21.2 28.1 0.0 0.0
DAltPct: 1899 0.0 16.0 11.0 12.5 1.0 4.0 8.5 18.5 28.5 0.0 0.0
DAltPct: 1909 0.0 20.6 11.8 11.2 0.6 0.6 4.7 20.0 30.6 0.0 0.0
DAltPct: 1919 0.0 22.9 14.3 9.5 4.8 0.0 3.8 17.1 27.6 0.0 0.0
DAltPct: 1929 0.0 32.2 21.1 16.7 10.0 0.0 10.0 10.0 0.0 0.0 0.0
DAltPct: 1939 4.3 14.5 14.0 12.6 9.2 9.2 6.3 7.7 22.2 0.0 0.0
DAltPct: 1949 0.9 12.4 13.7 7.3 7.3 9.0 7.7 17.2 24.5 0.0 0.0
DAltPct: 1959 0.0 10.4 19.4 0.5 9.5 14.2 5.2 28.0 12.8 0.0 0.0
DAltPct: 1969 14.0 9.3 15.8 3.3 7.0 12.1 3.7 19.1 15.8 0.0 0.0
DAltPct: 1979 17.0 5.1 16.0 2.4 6.8 9.5 7.1 19.0 17.0 0.0 0.0
DAltPct: 1989 10.5 4.2 15.0 3.8 6.9 9.9 8.7 20.8 20.4 0.0 0.0
DAltPct: 1999 8.0 2.7 14.7 8.8 9.2 8.0 9.4 19.7 19.5 0.0 0.0
DAltPct: 2009 7.3 2.4 14.6 9.8 9.8 7.3 9.8 19.5 19.5 0.0 0.0
For COUNTRY CODE: 101
The 1000 m to 2000 m band drops from a peak of 30.6% in 1909 decade ending to 19.5% now. During that same period, the 500 m to 1000 m band wobbles around 20% plus or minus a few (with what looks like a W.W.II dropout) and the “below 200 m” seems to lose the most, until the “Jet Age” 1960’s to 1990 or so grows a bunch of sea level locations. (I would guess airports near the coast). But at the end of the day, we have had a drop from 50+% over 500 m in 1909 decade ending to 39%.
For Morocco, we have the odd case of thermometers moving uphill and away from sea level:
Look at ./Alts/Therm.by.Alt130.Dec.ALT (Y/N)? y
Year -MSL 20 50 100 200 300 400 500 1000 2000 Space
DAltPct: 1879 100.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
DAltPct: 1899 30.0 0.0 70.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
DAltPct: 1909 37.5 0.0 62.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
DAltPct: 1919 7.7 0.0 92.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
DAltPct: 1929 24.0 0.0 52.0 0.0 0.0 0.0 24.0 0.0 0.0 0.0 0.0
DAltPct: 1939 25.6 0.0 48.7 0.0 0.0 0.0 25.6 0.0 0.0 0.0 0.0
DAltPct: 1949 37.0 0.0 39.1 0.0 0.0 0.0 23.9 0.0 0.0 0.0 0.0
DAltPct: 1959 17.4 0.0 41.7 0.0 7.8 0.0 33.0 0.0 0.0 0.0 0.0
DAltPct: 1969 8.8 7.9 33.3 0.0 3.5 0.0 22.8 15.8 7.9 0.0 0.0
DAltPct: 1979 1.1 10.9 31.5 0.0 0.0 0.0 31.5 14.1 10.9 0.0 0.0
DAltPct: 1989 0.0 14.8 28.4 0.0 0.0 0.0 27.3 14.8 14.8 0.0 0.0
DAltPct: 1999 0.0 18.0 21.3 0.0 0.0 0.0 21.3 19.7 19.7 0.0 0.0
DAltPct: 2009 0.0 0.0 25.0 0.0 0.0 0.0 25.0 25.0 25.0 0.0 0.0
For COUNTRY CODE: 130
Is Morocco warmer in the Atlas mountains than at the Atlantic Ocean shores? When we look at the stations that “survive” into 2009 we find:
[chiefio@tubularbells analysis]$ more Temps/130.stns2009
13060150000 MEKNES 33.88 -5.53 549 562U 248MVxxno-9A 1MED. GRAZING C 34
13060155000 CASABLANCA 33.57 -7.67 62 12U 1506HIxxCO 3A 1WATER C 60
13060230000 MARRAKECH 31.62 -8.03 466 488U 333HIxxno-9A 1MED. GRAZING C 31
13060265000 OUARZAZATE 30.93 -6.90 1140 1154R -9HIDEno-9A-9HOT DESERT C 28
[chiefio@tubularbells analysis]$
One station above 500 m, Meknes, and one over 1100 m, Quarzazate. And the high altitude station is labled “HOT DESERT”. OK, I think we have our answer… Marrakech is kept, being inland and hot too. I guess it would have been a bit blatant to have deleted Casablanca…
Historically, there were many more stations in Morocco:
[chiefio@tubularbells analysis]$ inin ^130
13060060000 SIDI IFNI 29.37 -10.18 66 44R -9HIDECO 1A-9WARM GRASS/SHRUBB 12
13060101001 CAPE SPARTEL MOROCCO 35.80 -5.90 59 31R -9HIxxCO 1x-9COASTAL EDGES B 0
13060120001 PORT LYAUTEY 34.30 -6.60 12 14U 139HIxxCO 3A 1WATER C 23
13060136001 SIDI SLIMANE 34.23 -6.07 55 59R -9FLxxno-9A-9WARM CROPS C 0
13060141001 FES/CITY 34.00 -5.00 414 401U 325HIxxno-9x-9MED. GRAZING C 18
13060150000 MEKNES 33.88 -5.53 549 562U 248MVxxno-9A 1MED. GRAZING C 34
13060155000 CASABLANCA 33.57 -7.67 62 12U 1506HIxxCO 3A 1WATER C 60
13060156000 NOUASSEUR 33.37 -7.58 206 200U 1506HIxxCO25A20WARM CROPS C 21
13060190000 KASBA-TADLA 32.53 -6.28 518 521R -9MVDEno-9x-9COOL FOR./FIELD A 0
13060220001 MOGADOR MOROCCO 31.50 -9.80 17 8S 30FLxxCO 1x-9WATER A 0
13060230000 MARRAKECH 31.62 -8.03 466 488U 333HIxxno-9A 1MED. GRAZING C 31
13060230001 BEN GUERIR 32.12 -7.88 448 448R -9HIDEno-9A-9WARM CROPS A 0
13060250000 AGADIR 30.38 -9.57 23 30U 61HIxxCO 5A 3WATER C 26
13060265000 OUARZAZATE 30.93 -6.90 1140 1154R -9HIDEno-9A-9HOT DESERT C 28
13060318001 TANGER MOROCCO 35.50 -5.60 60 393U 188HIxxCO 1x-9WARM CROPS A 0
Including Kasba-Tadla as “Cool Forest / Field” and a couple of more “WATER”. Even some “Grass / Shrubb” and “Coastal Edges”.
Isn’t it facinating how everywhere that the mountains are colder, thermometers leave; but here, where they are “HOT DESERT”, thermometers run from the cooler shores and beaches up into the hills…
Cook Like an Egyptian
How about Egypt? What can you do with a river valley?
Look at ./Alts/Therm.by.Alt115.Dec.ALT (Y/N)? y
Year -MSL 20 50 100 200 300 400 500 1000 2000 Space
DAltPct: 1869 0.0100.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
DAltPct: 1879 0.0100.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
DAltPct: 1889 60.0 40.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
DAltPct: 1899 61.5 38.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
DAltPct: 1909 44.1 27.9 7.2 20.7 0.0 0.0 0.0 0.0 0.0 0.0 0.0
DAltPct: 1919 36.2 29.2 7.7 26.9 0.0 0.0 0.0 0.0 0.0 0.0 0.0
DAltPct: 1929 35.5 29.1 9.1 26.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0
DAltPct: 1939 38.8 21.5 8.3 31.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0
DAltPct: 1949 36.2 21.3 14.2 28.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0
DAltPct: 1959 31.0 21.2 29.4 18.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0
DAltPct: 1969 35.0 14.7 25.8 24.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0
DAltPct: 1979 29.1 12.5 29.4 29.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0
DAltPct: 1989 7.9 21.1 24.6 46.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0
DAltPct: 1999 1.6 20.6 23.8 54.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
DAltPct: 2009 0.0 25.0 25.0 50.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
For COUNTRY CODE: 115
It looks like “Not Much”. But there are footprints in this riverbank mud…
We start with everything below 50 m elevation. By 1889 decade ending, just as GIStemp has begun time, 60% of the recordings are from below 20 meters. That’s got to be near the Mediterranean or other sea…
[chiefio@tubularbells analysis]$ more Temps/115.stns1890
11562315002 KOM EL NADURA EGYPT 31.20 29.90 32 10U 2319FLxxCO 1x-9WATER C 111
11562332000 PORT SAID/EL 31.28 32.23 6 2U 263FLxxCO 1x-9WATER B 14
11562371001 ABBAISSIA/CAIRO HQ EGYPT 30.10 31.30 33 63U 5084FLxxno-9x-9WARM IRRIGATED C 147
11562440000 ISMAILIA 30.60 32.25 13 9U 146FLxxno-9A 1WARM CROPS C 53
11562450001 SUEZ 29.93 32.55 3 6U 194FLxxCO 1x-9WATER C 18
[chiefio@tubularbells analysis]$
Gee… “Water Water everywhere”…
But what about now? We’ve abandoned the sea level band and headed up a shallow slope. But to where?
[chiefio@tubularbells analysis]$ more Temps/115.stns2009
11562306000 MERSA MATRUH 31.33 27.22 28 14S 28FLxxCO 5A 2WARM GRASS/SHRUBC 17
11562378000 HELWAN 29.87 31.33 141 30U 5084FLxxno-9A 2WARM IRRIGATED C 57
11562414000 ASWAN 24.03 32.88 100 150U 144FLxxLA-9x-9WARM IRRIGATED C 24
11562435000 KHARGA 25.45 30.53 73 85S 26HIxxno-9A 4HOT DESERT C 25
Oh… into the “HOT DESERT” and “WARM GRASS/SHRUBC”, but with some very “WARM IRRIGATED” out near the Aswan dam where it used to be desert, but is now still warm, but irrigated.
I may come back to Africa and continue to add individual countries to this list, but I think the pattern is clear just from this sample. For Africa, the “formula” is to move away from water and toward the inland deserts. Not just by latitude or by altitude, but “by hydrophobia” (though those two prior approaches showed the way to the “hydrophobia trick”…)
Any Traditional By Altitude Candidates?
Though before going, I must point out that there are places, those with no ocean or beach to avoid, who must resort to the old tried and true decent. The Congo:
Look at ./Alts/Therm.by.Alt154.Dec.ALT (Y/N)? y
Year -MSL 20 50 100 200 300 400 500 1000 2000 Space
DAltPct: 1959 3.1 0.0 0.0 0.0 0.0 31.2 24.1 20.7 21.0 0.0 0.0
DAltPct: 1969 0.4 0.0 0.0 0.0 0.0 36.7 22.4 24.9 15.6 0.0 0.0
DAltPct: 1978 0.0 0.0 0.0 0.0 0.0 39.7 19.0 30.6 10.7 0.0 0.0
DAltPct: 1989 0.0 0.0 0.0 0.0 0.0 58.8 13.7 23.5 3.9 0.0 0.0
DAltPct: 1999 0.0 0.0 0.0 0.0 0.0 84.2 5.3 10.5 0.0 0.0 0.0
DAltPct: 2008 0.0 0.0 0.0 0.0 0.0100.0 0.0 0.0 0.0 0.0 0.0
For COUNTRY CODE: 154
In the baseline early years, we have 21% of the records from 1000m to 2000m altitude and 41.7% from above 500 m of altitude. By 2008, all the records are from 300m to 400m (and likely one lone thermometer). That the decade ends in 2008 says that thermometer did not survide into 2009. The entire center of equatorial Africa must now be made up from whole cloth. One must look to southern Sudan, northern inland Angola, etc. to find the “trend” in Congo. But at least we don’t have those pesky 1000m to 2000m mountains to deal with.
The Poster Child
And finally, no look at Africa could pause without a mention of The Snows of Killimanjaro.
Tanzania
Look at ./Alts/Therm.by.Alt149.Dec.ALT (Y/N)? y
Year -MSL 20 50 100 200 300 400 500 1000 2000 Space
DAltPct: 1878 100.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
DAltPct: 1899 50.0 21.1 13.2 0.0 0.0 0.0 0.0 5.3 10.5 0.0 0.0
DAltPct: 1909 49.2 11.5 16.4 0.0 0.0 0.0 0.0 11.5 11.5 0.0 0.0
DAltPct: 1919 85.7 0.0 8.6 0.0 0.0 0.0 0.0 2.9 2.9 0.0 0.0
DAltPct: 1929 62.5 0.0 0.0 0.0 0.0 0.0 0.0 18.8 18.8 0.0 0.0
DAltPct: 1939 30.5 0.0 0.0 0.0 0.0 0.0 0.0 28.8 40.7 0.0 0.0
DAltPct: 1949 14.1 11.3 1.4 0.0 0.0 0.0 0.0 28.2 45.1 0.0 0.0
DAltPct: 1959 13.7 1.4 26.0 0.0 0.0 0.0 0.0 2.7 56.2 0.0 0.0
DAltPct: 1969 8.1 7.3 23.4 0.0 0.0 0.0 0.0 0.0 61.3 0.0 0.0
DAltPct: 1979 5.9 6.6 19.1 8.6 0.0 0.0 0.0 0.0 59.9 0.0 0.0
DAltPct: 1989 1.1 1.1 24.7 14.6 0.0 0.0 0.0 0.0 58.4 0.0 0.0
DAltPct: 1999 0.0 0.0 18.6 17.1 0.0 0.0 0.0 0.0 64.3 0.0 0.0
DAltPct: 2009 0.0 0.0 16.9 16.9 0.0 0.0 0.0 0.0 66.1 0.0 0.0
For COUNTRY CODE: 149
Wow. The 1000 m to 2000 m band grows from 10% to 2/3 of the coverage! We do lose the sea level up to 50 meters band (so once again we leave the ocean / shore and move to the “couple of hundred” meters inland locations. But what’s up with that “up to 2000 m band”?
[chiefio@tubularbells analysis]$ more Temps/149.stns2009
14963756000 MWANZA -2.47 32.92 1140 1213U 111HIxxLA-9x-9WARM CROPS B 10
14963832000 TABORA AIRPOR -5.08 32.83 1190 1206U 67FLxxno-9A 2TROPICAL DRY FORA 0
14963862000 DODOMA -6.17 35.77 1120 1131S 46HIxxno-9A 1SUCCULENT THORNSC 15
14963894000 DAR ES SALAAM -6.87 39.20 55 71U 757HIxxCO13A 5WARM CROPS C 13
14963962000 SONGEA -10.68 35.58 1067 1204S 18MVxxno-9A 7TROPICAL DRY FORA 0
14963971000 MTWARA -10.27 40.18 113 74S 49HIxxCO 3x-9TROP. SEASONAL B 8
[chiefio@tubularbells analysis]$
So we go inland and upslope some, but into “TROPICAL” of various kinds and to airports. (That “A” flag in the “-9A” just before the land type). No need to go anywhere near those tall cold snowy mountains, though. For the GHCN data set from NCDC and for GIStemp, the snows of Kilamanjaro already do not exist.
Oddly, this does not manage to overcome the cooling that happens in Tanzania. The temperature averages actually drop over time. The right most column is the total number of thermometers in that year while the column just to the left of it is the annual average. We start out at 26.7 C and slowly drop to the 23 C range where we substantially stabilize in the 1930s and where we stay to today. (Missing data is shown as -99.0 but does not participate in the averageing process).
Look at ./Temps/Temps.149.yrs.GAT (Y/N)? y Thermometer Records, Average of Monthly Data and Yearly Average by Year Across Month, with a count of thermometer records in that year -------------------------------------------------------------------------- YEAR JAN FEB MAR APR MAY JUN JULY AUG SEPT OCT NOV DEC YR COUNT -------------------------------------------------------------------------- 1850 28.6 28.6 28.5 27.5 25.9 25.8 25.1 25.2 25.3 26.2 26.6-99.0 26.7 1 1859-99.0 28.8 29.4 26.3-99.0 26.2 24.4 23.8 24.4 26.3 26.9 28.1 26.5 1 1864 27.6 28.1 26.7 26.5 26.3 25.2 24.0 24.1 25.3 25.1 27.3 27.5 26.1 1 1875-99.0-99.0-99.0-99.0 26.3 25.9 24.9 25.4 26.3 26.9 27.1 27.7 26.3 1 1876 27.9 27.9 27.8 26.7 26.0 25.2 25.0 25.0 25.2 26.3 27.3 27.6 26.5 1 1877 28.5 28.5 28.4 27.6 26.9 26.2 25.8 26.0 26.7 27.0 27.1 27.7 27.2 1 1878 28.4 28.6 28.8 27.7 27.1 26.1 25.6 25.9 26.1 26.7 27.3 27.4 27.1 1 1880-99.0 28.1 28.0 27.2 25.9 25.9 24.6 24.7 25.7 25.7 26.5 27.5 26.3 1 1881 27.0 26.9 28.5-99.0-99.0-99.0-99.0-99.0-99.0-99.0-99.0-99.0 27.5 1 1892 28.1 28.4 28.3 27.3 26.3 25.7 24.6 24.7 25.2 25.9 27.0 27.8 26.6 2 1893 27.4 28.2 27.6 26.8 24.2 23.8 23.4 23.5 24.6 26.0 25.8 27.1 25.7 3 1894 26.7 27.5 26.4 25.8 24.8 23.7 23.5 24.6 25.5 26.3 26.3 26.9 25.7 3 1895 27.3 27.1 27.1 26.1 25.3 24.0 22.9 23.4 23.7 25.8 26.4 27.9 25.6 5 1896 28.6 28.8 27.8 26.7 26.0 25.5 24.3 24.0 24.6 25.7 26.0 27.8 26.3 3 1897 28.5 28.5 28.2 26.6 25.9 24.7 24.3 24.7 25.0 25.8 27.5 28.7 26.5 3 1898 28.8 28.9 27.3 25.4 26.6 25.3 23.2 23.2 24.2 25.1 26.2 27.0 25.9 4 1899 27.3 27.8 27.1 26.3 24.3 24.0 23.6 24.0 25.0 26.2 27.6 27.8 25.9 5 1900 27.6 28.7 28.4 27.5 26.5 25.5 24.8 24.8 25.5 26.2 27.3 27.7 26.7 4 1901 28.6 27.6 27.4 26.0 25.0 23.7 23.8 23.9 24.6 25.7 26.2 26.6 25.8 6 1902 26.8 26.5 27.5 26.6 25.2 24.3 23.8 24.1 25.1 25.3 26.0 26.0 25.6 6 1903 26.8 26.6 27.0 25.8 24.7 24.4 23.6 23.9 25.1 25.4 26.2 26.7 25.5 6 1904 27.2 28.2 27.8 24.8 24.2 23.3 22.9 23.4 24.8 25.0 25.4 26.8 25.3 6 1905 27.0 27.3 26.2 24.8 24.4 23.9 23.1 23.5 24.4 25.7 26.7 26.6 25.3 6 1906 26.6 27.2 25.7 24.8 24.3 23.3 22.9 23.8 24.6 25.8 25.8 25.9 25.1 6 1907 26.7 27.1 27.2 25.9 24.9 24.0 23.4 23.9 24.4 26.1 27.1 28.1 25.7 5 1908 28.8 28.1 28.2 27.6 25.3 25.2 24.7 24.8 25.2 26.1 27.1 28.5 26.6 3 1909 28.3 28.4 28.4 26.0 26.0 25.1 24.5 24.2 25.3 25.8 26.6 26.9 26.3 3 1910 27.5 28.3 28.5 26.2 25.2 24.9 24.0 24.4 25.2 26.0 26.8 27.7 26.2 3 1911 27.1 27.0 26.7 25.0 24.3 22.7 22.0 22.8 23.5 24.7 25.4 26.5 24.8 5 1912 28.4 28.2 27.8 26.9 26.9 25.6 24.7 24.8 25.3 26.3 27.1 26.9 26.6 3 1913 28.5 29.4 27.5 26.3 26.3 25.6 25.1 25.1 25.6 26.4 27.8 28.5 26.8 2 1914 28.4 29.2 28.7 28.7 26.9 26.6 25.6 25.9 26.4 27.2 28.3 28.3 27.5 2 1915 28.9 29.4 29.2 28.0 26.8 25.9 25.5 25.5 26.1 26.7 27.4 28.7 27.3 2 1916 28.6 28.7 28.5 27.2 26.6 26.1 25.0 25.0 25.9 26.2 27.5 28.3 27.0 2 1917 28.5 28.4 28.4 26.5 26.3 25.9 25.6 25.4 26.1 26.3 27.4 28.3 26.9 2 1918 27.8 28.0 28.8 27.3 26.6 25.6 24.9 24.8 25.4 26.2 27.5 28.2 26.8 2 1919 28.8 29.3 29.0 28.3 26.8 25.7 25.3 25.5 26.2 27.0 27.6 28.6 27.3 2 1920 29.0 29.1 28.9 27.6 26.9 25.6 25.1 25.2 25.7 26.5 27.7 28.1 27.1 2 1921 27.9 28.8 28.2 26.8 26.3 25.4 24.8 25.1 25.7 26.3 27.5 28.5 26.8 1 1922 28.8 29.0 28.9 28.2 26.1 25.4 24.6 24.8 25.4 26.3 27.1 27.9 26.9 1 1923 28.1 27.1 27.5 25.7 26.0 25.2 24.6 24.4 25.2 26.0 26.3 26.9 26.1 2 1924 28.8 27.0 28.5 27.2 26.9 25.4 24.5 25.2 25.3 26.0 26.9 27.5 26.6 2 1925 27.2 27.6 28.3 28.0 26.9 26.4 25.4 25.4 25.8 26.7 27.5 28.1 26.9 1 1926 27.0 27.3 27.1 25.9 24.8 24.5 23.8 24.2 25.2 25.4 26.1 26.6 25.7 3 1927 26.6 27.3 25.2 25.2 24.4 24.6 23.6 24.1 25.0 25.5 24.9 26.0 25.2 3 1928 25.9 26.6 25.8 24.8 24.2 23.5 22.6 23.3 24.4 24.5 24.8 25.2 24.6 4 1929 26.1 26.6 26.0 24.7 24.7 23.5 22.9 23.2 24.6 24.3 24.8 24.8 24.7 4 1930 25.5 25.4 24.9 24.4 24.3 23.2 23.3 23.2 24.3 24.8 24.6 25.2 24.4 4 1931 26.1 26.1 25.6 24.8 24.2 23.3 23.4 24.0 24.4 24.7 25.2 24.7 24.7 4 1932 26.2 25.7 25.6 25.3 24.6 24.2 23.5 24.4 24.7 25.3 26.2 25.6 25.1 3 1933 24.7 24.6 25.2 25.2 24.2 23.3 22.7 23.0 23.7 24.7 25.5 24.4 24.3 4 1934 23.9 24.5 24.4 24.3 23.5 22.8 22.6 23.3 24.3 24.8 24.8 24.1 23.9 5 1935 24.9 24.1 24.6 24.1 23.3 22.8 22.0 22.7 24.2 24.9 25.5 24.3 23.9 5 1936 23.9 24.3 24.6 23.5 23.6 22.5 22.5 22.9 23.8 25.4 25.9 24.5 23.9 5 1937 24.5 25.0 25.2 23.6 23.1 22.5 22.2 22.7 24.2 24.7 24.6 24.4 23.9 4 1938 24.5 24.1 23.5 23.8 23.3 22.4 22.2 22.6 23.7 24.5 24.1 23.6 23.5 5 1939 24.1 24.1 24.0 23.6 23.0 22.3 22.4 22.9 23.8 24.9 24.4 24.2 23.6 5 1940 23.4 23.9 23.5 23.7 23.5 22.8 22.3 23.1 24.6 25.3 24.7 24.3 23.8 5 1941 23.9 25.1 24.8 24.4 23.8 23.2 22.7 23.4 24.7 25.6 24.7 23.8 24.2 6 1942 24.3 25.3 24.4 24.0 22.8 22.9 21.9 22.7 23.9 25.0 24.7 24.0 23.8 6 1943 24.8 24.2 24.7 24.5 23.7 22.7 21.8 22.5 23.4 24.9 24.9 24.9 23.9 6 1944 24.0 24.2 23.9 23.3 23.1 22.9 22.4 23.0 24.0 24.9 24.2 23.9 23.6 5 1945 23.7 24.0 24.4 24.2 22.7 22.1 22.1 23.0 24.1 24.6 25.1 24.3 23.7 6 1946 25.0 25.1 25.2 24.6 23.7 23.1 22.8 23.4 24.2 25.1 24.7 24.6 24.3 6 1947 24.7 24.8 24.7 23.9 23.0 22.6 22.4 22.6 23.7 24.6 24.8 23.3 23.8 6 1948 23.9 24.7 24.2 24.0 23.7 23.1 23.0 23.1 23.9 24.8 24.8 24.2 24.0 6 1949 24.8 24.3 25.5 24.9 24.0 23.0 22.7 23.1 24.0 25.0 25.7 24.8 24.3 8 1950 24.0 24.7 24.2 23.9 22.9 22.2 21.8 22.4 23.4 24.1 24.8 24.3 23.6 8 1951 24.3 24.1 24.7 23.8 22.7 21.6 21.2 22.1 23.8 25.2 25.0 24.4 23.6 4 1952 25.4 25.1 25.1 24.3 23.2 21.7 21.2 22.1 23.4 24.4 24.6 26.1 23.9 4 1953 25.0 25.7 24.8 24.1 23.2 21.9 21.3 22.2 22.9 24.6 25.4 25.2 23.9 4 1954 24.6 25.0 24.9 23.9 22.9 21.5 21.2 22.9 23.5 24.4 25.4 24.9 23.8 4 1955 25.0 24.5 24.6 23.6 22.5 21.2 21.5 21.8 23.1 24.7 25.2 24.3 23.5 4 1956 23.5 24.5 24.9 23.7 22.8 21.2 20.1 21.8 22.5 24.3 25.1 24.7 23.3 4 1957 24.5 24.0 24.3 23.9 22.7 21.5 21.2 22.2 23.4 24.6 25.4 25.0 23.6 4 1958 25.2 25.2 25.1 24.8 23.1 21.8 20.9 22.4 24.0 24.9 25.8 24.8 24.0 4 1959 25.3 25.1 25.0 24.3 23.4 22.2 21.3 22.2 23.0 24.4 25.0 24.8 23.8 4 1960 24.4 24.4 24.3 23.7 22.8 21.7 20.8 21.9 23.2 24.6 24.9 26.4 23.6 4 1961 25.3 24.1 24.5 24.1 24.0 22.3 21.3 21.6 23.1 24.4 23.9 23.8 23.5 7 1962 23.7 24.2 24.1 23.7 21.9 21.6 20.5 21.7 22.6 23.4 24.5 24.4 23.0 7 1963 23.5 24.2 24.0 23.1 21.9 20.7 20.7 20.9 22.5 24.3 24.0 24.0 22.8 7 1964 24.3 24.6 24.2 23.5 21.7 20.5 19.4 20.9 21.9 23.3 24.4 24.2 22.7 7 1965 23.5 24.2 24.0 23.2 21.8 19.9 19.9 20.7 22.6 23.8 24.7 24.2 22.7 7 1966 24.5 24.6 23.9 23.3 22.4 21.1 20.7 21.5 23.1 24.0 24.5 24.3 23.2 7 1967 24.6 24.5 24.7 23.6 22.5 21.2 20.7 21.4 22.5 24.0 24.4 23.6 23.1 7 1968 23.9 23.8 23.5 22.8 22.0 20.5 20.0 20.9 22.3 24.0 24.2 24.4 22.7 7 1969 24.2 24.0 24.3 24.4 22.8 20.9 20.9 21.3 22.8 24.4 24.7 24.8 23.3 7 1970 24.1 24.7 24.2 23.2 22.2 21.3 21.1 21.6 22.7 23.6 24.8 23.1 23.1 7 1971 23.9 23.8 24.0 23.5 21.9 20.4 20.7 20.8 22.4 23.5 24.3 24.1 22.8 9 1972 24.0 24.1 24.0 23.7 22.8 21.0 21.0 21.6 23.1 24.4 24.1 24.3 23.2 9 1973 24.6 25.0 25.1 23.9 22.7 21.3 20.5 21.4 22.8 24.0 24.3 23.9 23.3 9 1974 23.6 24.3 23.7 23.2 22.4 21.1 20.6 21.5 22.2 23.7 24.4 25.2 23.0 9 1975 24.2 24.6 24.1 23.6 22.7 21.9 21.0 21.7 22.3 23.2 24.7 24.1 23.2 9 1976 24.1 23.4 24.0 23.6 22.7 21.6 20.8 21.5 23.0 24.3 24.9 25.4 23.3 9 1977 24.3 24.2 24.1 23.9 23.3 22.1 22.2 22.8 23.7 24.9 24.3 24.2 23.7 8 1978 24.4 24.8 24.2 23.6 22.5 21.8 20.8 21.9 23.1 24.3 24.2 23.9 23.3 9 1979 24.5 24.2 24.5 24.1 23.5 21.8 21.5 22.5 23.0 24.6 25.0 25.1 23.7 9 1980 25.3 25.3 25.1 24.2 23.7 22.1 21.9 22.4 23.4 24.4 24.7 24.4 23.9 9 1981-99.0 25.8 25.4 24.0 22.8 23.1 23.3 22.2 23.6 25.4 25.5 25.1 24.2 5 1982 24.9 24.3 24.9 24.5 22.8 21.8 23.6-99.0 25.2-99.0-99.0-99.0 24.0 5 1983-99.0 26.0 24.6 24.6 24.1 22.9 22.2 22.4 23.0 23.9 25.4 24.9 24.0 5 1984 24.8 25.9 24.7 23.4 22.0 21.3 21.2 20.8 21.8 24.7 23.7-99.0 23.1 5 1985 26.4-99.0-99.0 23.9 22.8 21.6 21.7 22.3 23.9 24.5 24.9 24.7 23.7 5 1986 25.5 27.7 24.1 25.4 22.3 20.7 20.3 22.3 23.4 25.3 24.8 23.9 23.8 5 1987 24.2 24.8 25.1 25.6 23.4 21.1 21.9 23.7 23.3 24.9 25.3 26.0 24.1 6 1988 25.1 25.2 25.2 24.5 23.4 22.3 21.4 22.7 23.9 25.0 24.7 25.9 24.1 6 1989 24.2 24.8 24.3 23.2 22.6 21.2 22.8-99.0 22.5 24.3 25.0 24.9 23.6 5 1990 25.2 25.2 25.6 24.2 23.7 22.1-99.0 21.8-99.0 24.4 25.2 25.1 24.2 6 1991 25.1 25.6 25.8 23.7 23.0 22.1 21.4-99.0 22.9 23.6 24.1-99.0 23.7 6 1992 24.7 24.8 25.4 24.6 22.8-99.0 20.9 22.0-99.0-99.0-99.0-99.0 23.6 6 1993-99.0 24.4-99.0-99.0-99.0 21.3 21.0-99.0-99.0 24.1-99.0 26.3 23.4 5 1994-99.0-99.0-99.0 24.1 23.0-99.0 20.0 22.4 23.6 25.0 24.7 24.7 23.4 6 1995 25.2-99.0 24.6 24.5 24.1 21.5 22.0 22.9 23.4-99.0 24.9-99.0 23.7 6 1996 24.3 25.2-99.0 23.8 22.6 22.0 21.1 21.1 23.0-99.0 24.9 24.7 23.3 6 1997 25.3 24.9 25.1-99.0 22.7 21.9 21.1 22.0 23.5 24.6 24.6 24.0 23.6 6 1998 25.0 25.2 25.4 23.7 22.6-99.0 21.6 22.3 23.4 25.0 24.9 25.5 24.1 6 1999 24.8-99.0-99.0-99.0 23.6 22.5 22.1 22.8-99.0 24.1-99.0 24.5 23.5 6 2000-99.0 25.1 24.5 24.1 24.2-99.0 22.5 22.5-99.0 24.5 25.1 24.2 24.1 6 2001-99.0 24.1 24.8-99.0 24.1 22.0 21.4 22.3 23.4 24.7-99.0-99.0 23.3 6 2002 24.5-99.0 25.0 24.4-99.0 22.0 22.3 22.3 23.2 24.6 25.1 25.0 23.8 6 2003 25.6 25.5 25.9 24.9 23.9 22.8-99.0 22.6 24.0 25.0 26.0-99.0 24.6 6 2004 25.5-99.0 25.1 24.2 23.3-99.0 21.8-99.0-99.0-99.0-99.0-99.0 24.0 6 2005-99.0-99.0-99.0-99.0-99.0-99.0-99.0 22.4-99.0 24.5 25.5-99.0 24.1 5 2006 26.0 25.7-99.0 23.7 23.5 22.6 21.6 22.8 23.3 25.6 24.5 23.9 23.9 6 2007 24.4 24.3 23.8 24.0 24.0 22.6 22.3 22.5 23.8 24.7 25.2 24.6 23.9 6 2008 24.9 24.1 23.8 23.1 22.7 20.5 21.8 22.4 23.5 24.9 25.1 24.8 23.5 6 AA 25.1 25.3 25.1 24.4 23.4 22.4 21.9 22.6 23.6 24.7 25.1 25.1 24.1 Ad 25.8 26.0 25.8 25.0 24.1 23.1 22.6 23.1 24.1 25.1 25.6 25.7 24.7 For Country Code 149
The AA line is the Average of Averages while the Ad line is an average of the detail that maks up the individual averages. You can see the impact of a programmers decision on “which way to average” in the 1/10 C place. It is about 1/2 C that is entirely dependent on how you chose to do the math…
So I guess all that hoopla about global warming melting the snows of Kilimanjaro is based on the mountain tops getting warmer while all around them the rest of the country is getting cooler? How does that work again? Global warming having teleconnection from the Morocco Sahara straight into the snows and bypassing the rest of the country?
To quote someone or other: “I don’t think so Tim.”
For a basic overview of GIStemp, see:
https://chiefio.wordpress.com/2009/11/09/gistemp-a-human-view/
Musings from the Chiefio 
Interesting.
EM, your difference between Average of the Detail and Average of the Averages seems very high. Could you post the link for the data, I couldn’t find it when I googled.
The reason that it seems high to me is that I couldn’t replicate the scale of the difference with a smaller dataset. Not being able to find the full record I used your monthly averages.
I took your Tanzania table of averages, and converted all it to one column of figures, there are 1551 records plus blanks. I split that into batches of 50 (including blanks) to simulate years and I get an average of all records of 24.66 and an average of the yearly averages of 24.71. The range is 19.4 to 29.4 for individual records.
I realise that there are many times more records in your Tanzania data set, but I can’t account for the massive divergence, 0.05 to 0.5.
Any thoughts?
REPLY: [ The data are the GHCN. Directions for ftp download are under the GIStemp tab up top. You can select a given country by the first 3 characters of the record (it’s a text record, so Excel is fine). I’ve posted the code that was the precursor to this particular code under one of the other postings. If you would like, I can post the present version here. Mostly I just cleaned up the headings (like putting the AA and Ad in). The ‘effect’ of such large size variation comes entirely from the sparse nature of the early data. A single thermometer for, oh, 40 years of early records would either be averaged in as “1 thermometer record” or “1 monthly average record”. If later eras had 10 thermometers, you get a 10:1 variation in impact from that one thermometer. For most countries, the “first thermometer” was not situated at the center of heat of the country… But as a programmer you MUST pick a method to do your average. So I put this in just to remind folks that there are many averages and many ways to make them, and thus many answers are possible…and for most of the “professional temperature products” I’ve seen, the reasoning behind the various selections is vacant. So folks ought to know that there are variations in the 1/10 C to whole degree C range based on “programmer choices” that the programmer MUST make, like it or not. -ems ]
One last thought. I realised that I was thinking weeks when I used fifty, to I re-subtotaled with 30 day intervals to simulate monthly data. Average of 2 place averages 24.67 compared to 24.66. Rounding to 1 decimal place gives 24.67. Rounding to 0 decimal places gives 24.76. So that gets 20% of your difference.
However, I am using the full 12 months of data from the year, so my variability within the month is much greater than your original.
REPLY: [ See:
And the effect is not a ’rounding thing’ it’s a weighting thing… The average of averages overweights months with few thermometers.
-ems]
Hi:
Lots of amazing work on this website……
Quick question – how many thermometers worldwide might have a continuous available 50 yr (or so) history to the present? If you did no adjustments and just calculated a straight average of these what type of anomaly might be present? IE with a static set of thermometers? If you separate those into rural and urban or whatever other categories you may choose globally, how different is that?
Apologies if this is on the site somewhere….
JPC
REPLY: [ I think you are looking for this:
A generally good place to start is the “GIStemp” tab at the top. Yeah, there’s a lot. I wrote it all and it take me a while some times to find a particular bit… so ‘no worries’ on asking for a pointer… -ems ]
Having the brain of a small child, I cannot help but think of this quote from “Counting Cats in Zanzibar”
“Everybody knows that hot air rises. So why are the tops of mountains so cold?
It’s one of those sort of questions that only geniuses and small children ask.”
http://www.countingcats.com/?p=4745
REPLY : [ A marvelous and wonderfully clear read. Anyone who would understand “greenhouse effect” ought to read that article. Thank you! -ems ]
Pingback: Don’t Hold Your Breath. What AGW Proponents Have Yet to Explain. « Thoughtful Analysis
Thanks for the reference, I’ll have a go at reproducing your results tomorrow. (So obviously, we’re not climate scientists. Ha ha.)
Your comment:
“The average of averages overweights months with few thermometers”
raises an alarm bell, hopefully without foundation. The thought occurs; You have identified thermometers being dropped from the record in “the dying”, have you encountered any indication of data loss within the retained thermometers that might increase the average in this manner? It wouldn’t take dropping too many lows to add a 1/10 degree. (Or is that more properly the function of GISSTemp?)
REPLY: [ The USHCN to USHCN.v2 transition is accompanied by defining more of the ‘old records’ as missing data due to failing the “quality control” assessment. IMHO this is a slightly smoking gun on re-writing the past… again… There are other changes between USHCN and USHCN Version Two as well. I’m increasingly coming to cringe at the words “value added” … -ems ]
Temperature reading in Africa being the current topic, you may be interested in this excerpt from science journalist Ron Bailey:
http://reason.com/archives/2009/12/01/the-scientific-tragedy-of-clim
In an email to University of Alabama climatologist John Christy I asked, “Is there a possibility that the teams that compile temperature data could all be making the same set of errors which would result in them finding similar (and perhaps) spurious trends?” Christy replied that he believed this was possible and cited some recent work he had done on temperature trends in East Africa as evidence. In that article he found that using both the maximum and minimum temperature rather than the mean temperature (TMean) used by the three official data sets gives a better indication of actual temperature trends in the region.
Christy found that the maximum temperature (TMax) trend has been essentially zero since 1900 while the minimum temperature (TMin) trend has been increasing. In his email to me, Christy explained, “As it turns out, TMin warms significantly due to factors other than the greenhouse effect, so TMean, because it is affected by TMin, is a poor proxy for understanding the greenhouse effect of ‘global warming’.” Or as his journal article puts it, “There appears to be little change in East Africa’s TMax, and if TMax is a suitable proxy for climate changes affecting the deep atmosphere, there has been little impact in the past half-century.” So if Christy’s analysis is correct, much of the global warming in East Africa reported by the three official data sets is exaggerated. Christy has found similar effects on temperature trend reporting for other regions of the world.
Cheers!
REPLY: [ Thank you! A fascinating data point. I’d been pondering running MAX and MIN sets through GIStemp and comparing the anomalies with MEAN but had been a bit under motivated. This finding pushes that effort up the priority scale. It ought to be rather interesting… -E.M.Smith ]
Pingback: Proposition 01: Thermometer drift « Climate Science
Wow, it looks like my trackback has already made it here. Didn’t know WordPress did that…
So, yes, I’ve been reading your site for a while and when you suggested the idea of distance-to-ocean checks I figured that was something I’d like to take a crack at. I’ve several posts tagged here and I hope to get the graphs up tomorrow.
Cheers, Russ.
REPLY: [ The trackback feature can be turned on or off for each thread as desired. If you want a trackback taken down, just let me know, otherwise it looks like a feature to me. I’ll be waiting to see what you find on the thermometer distance to ocean over time effort. When you have something, let me know, I’ll put up a posting. And yes, the first digit of station ID is the continent. Then the next two complete the country code. The next 5 are the geographic station inside the country (with some order by location) So you can pick out “Hawaii” or Hilo, or the whole Pacific Basin by substring of the StationID. An nice touch. -E.M.Smith ]
Hi EMS!
Oh, the trackbacks do seem useful, I just had no idea the my blog was going to run off and start posting comments on yours :)
Had a slight mishap with my data handling – parsing trouble – so have had to redo all of the calculations. The graphs are now up here.
I can’t find any sign of Africa running-to-the-hills, but it certainly looks like Europe has been heading for the beach.
I asked this on WUWT but I think it got lost in the scatter and you would probably have a better answer anyway if what I suspect is the case is true so here goes
I was going through the “raw” from GISS for State College Pa. when I came across something I can’t see how they came up with.
So there I am looking at 1973 and 1974, both of which had no data for the month of November. 1973 had 17.9C for Sept and 12.9C for Oct and a S-O-N of 12.3 . So then I go and look at 1974 and for Sept they had 15.9C and Oct 9.5C and a S-O-N of 9.6C . Now can anyone explain how you come up with a seasonal average lower then all inputs and another only .1C above all inputs?
To my poor tired brain it should be an average of 15.4C for ‘73 and 12.7C for ‘74.
Then in 1975 they don’t have data for Oct and I used the Sept and Nov data provided and what I calculated matches the S-O-N listed of 11.9C . However when I go to 1984 they don’t have data for Sept and when I calculate the S-O-N it doesn’t match the data file. My Calc: 8.5C what they got is 11.6C.
To me it looks like the USHCN v2 has a bug in it that when the data is missing from the center of the Seasonal mean calculation it gives a funking reading but when it is on either side (ie the S or the N in S-O-N) it works fine. I also checked to see if it worked the same way in M-A-M and J-J-A seasonal averages and they both do. If the missing data is in either the 1st or 3rd spot the Seasonal calculation is off, if its the 2nd spot the calculation is correct. I also found it this way on a second “Raw” data set.
REPLY: [ GIStemp has a homogenization step (STEP1) that happens before it makes the seasonal averages. The “missing data” may be made up in several ways. I believe that you are experiencing an artifact of those behaviors. The code in question would be the STEP1 Python and perhaps a bit of the STEP2 UHI molestation. If you want a more fine grained explanation than that, I think you will need to actually read the code. It’s, er, rather complex and, IMHO, not completely rational. -E.M.Smith ]