The Great Dying of Thermometers
That blue line on the top is the Northern Temperate zone, from 30N to 50N. In this posting we look at the impact of The Great Dying of Thermometers in America, with a specific focus on California. If a future posting, we will look to see if a similar thing happens in Europe. For now, The Great Dying, in America:
Who Needs Snowy Mountains When You Can Use The Beach
Are 4 near-ocean locations enough to measure California? Would they be warmer than the snowy mountains?
Lately we’ve been told that California and the west in general had set a 115 year record for high temperatures. All Time Heat! I called “BS”.
No way was this year even a regularly warm year. My tomatoes were not setting fruit (they can not set fruit below a 50F night temperature unless you use special varieties. I had 2 of the special varieties that were setting some fruit, but far less than in prior years. Further, my Runner Beans were setting seeds. Normally in mid summer they have lots of wonderful red flowers that attract lots of hummingbirds, but they can not set seed over about 93-95 F and in July and August I expect only decoration. I got beans… So I called a big load of BS but could not point to any reason why. That has changed.
I have been rooting around in the GHCN data set that feeds into GIStemp. My copy was downloaded some time ago (the 2009 data “cuts off” in June so that is about when I got my copy). I doubt they have changed the data much since then (though it is possible. It ought to be verified that the present conditions still hold in the present published data.)
UPDATE: I now have a completely independent confirmation (see comments below). GIStemp and ANY THING ELSE USING GHCN DATA produce no useful results after 2006. The temperature series are too broken by thermometer deletions. (Though I think I can fix GIStemp by merging in the USHCN.v2 data. Give it a day or two…)
And I have done a bit of “spot checking”. I went to :
http://data.giss.nasa.gov/gistemp/station_data/
where I’ve entered some places that ought to have records. I checked the “raw after combining GHCN and USHCN” for, for example, the State Capitol: Sacramento. It says the data cuts off in 2004. Chico in 2007. Redding 2007. Fresno 2007. Eureka, Ca. 2007. Yosemite Park Headquarters 2006. Death Valley 2007. Bishop Airport 2004. Cal Poly SLO 2007. You can even click on the map of the USA in California and get a list that sure looks to me like it matches:
The Sacramento graph:
What conditions am I seeing in the GHCN data set? The Thermometer Langoliers have eaten 9/10 of the thermometers in the USA; including all the cold ones in California.
California thermometer’s at the beach. Ski season is over.
I have made a little tool that shows me “thermometers by latitude” and another little tool that shows me thermometer records by year. At present they show an “issue” with too few thermometers; and they are are not located in representative places. When I investigated, I found California has all of 4 thermometers (assuming I read the LAT and LONG correctly and recognized the place names; someone check me on this.)
San Francisco
Santa Maria
Los Angeles
San Diego
42572494000 SAN FRANCISCO 37.62 -122.38 5 102U 6253FLxxCO15A 1COASTAL EDGES C3 66
42572394000 SANTA MARIA/P 34.90 -120.45 73 120U 62HIxxCO15A 2WARM CROPS C3 23
42572295000 LOS ANGELES CALIFORNIA 33.72 -118.27 -999 55U14531FLxxCO 1x-9WATER C3 32
42572290000 SAN DIEGO/LIN 32.73 -117.17 9 39U 2498FLxxCO 1A 1WATER C3 105
IIRC, Santa Maria is a nice little coastal town in Southern California. Everyone knows L.A. and San Diego from Beach Blanket Babylon, The Beach Boys, and Surfer Girl fame. That just leaves San Francisco to stand in for all the rest of Northern California.
SF is a nice beach town, with cool summers, but in winter the ocean keeps it from getting very cold at all. So all the frozen inland, all the Chico frost, Redding hail, Weed snow and all the Sierra Nevada under whatever frozen snow, well they all are represented by a nice 50F to 60F day in San Francisco. Forget Yosemite, Mount Shasta, Truckee / Tahoe, Mono Lake, Trinity and the Cascades, even the cold evenings of the Mojave Desert. It’s all just downtown LA…
Would You Believe a Little Over 2 Thermometers Per State?
And no, that is not a “Maxwell Smart” imitation.
My “by years” tool told me there were 136 active thermometer records in the U.S.A. in 2008. For the whole thing. Including Alaska and Hawaii. But in fairness, Hawaii got three thermometers, all at airports
42591165000 LIHUE, KAUAI, 21.98 -159.35 45 86R -9MVxxCO 1A-9WARM FOR./FIELD C 21
42591182000 HONOLULU, OAH 21.35 -157.93 5 240U 836HIxxCO 1A 1WARM FIELD WOODSC 70
42591285000 HILO/GEN. LYM 19.72 -155.07 11 46S 38HIxxCO 2A 1WARM FOR./FIELD C 19
Thinking something must be wrong with my program, I went back and by hand extracted the “2008” data from the GHCN main data file “v2.mean”, and counted up the lines. My program was working fine.
UPDATE: we have the confirmation. The “caveat” is no longer needed
Caveat: IMHO, this is such a big deal that normally I would not publish this page until I had it confirmed. But with the quote “in comments on another page” about ‘dozens’ of sites dropped out, with the “115 year record heat” in the cold, and with Copenhagen so soon; and with the GISS site seeming to confirm it: I’ve decided to risk a face full of eggs and post. My intent it not to use this to toss rocks (yet), but rather to get confirmation. There will be plenty of time for hoopla just a bit later.
BTW, a quick scan looks like maybe Washington and Oregon got the same treatment.
A little bit of LINUX / UNIX tool magic was applied and I had matched this file of StationIDs against the Station Inventory information (in the file v2.inv) and could put names and LAT LONG on each of these stations. Here they are. This is EVERYTHING fed into GIStemp for the USA from the GHCN file. ( I still need to check USHCN, but as I recall; it was depricated a year or two ago and no longer gets any updates… so I’m pretty sure “this is all there is”. Oh, and STEP2 might still toss out some of these records.) Read it and weep. Since it is so short, I’ll give it to you in two formats. Pretty, but truncated right; then ‘ragged right with wrap’; but you can read all the fields:
[chiefio@tubularbells analysis]$ cat USA.2008.used 42570026000 BARROW/W. POS 71.30 -156.78 4 0R -9FLMACO 1A-9TUNDRA C 40 42570133000 KOTZEBUE, RAL 66.87 -162.63 5 0R -9FLMACO 1A-9WATER A 10 42570200000 NOME 64.50 -165.43 7 48R -9HIxxCO 1A-9WATER C 23 42570231000 MCGRATH 62.97 -155.62 103 136R -9FLxxno-9A-9MAIN TAIGA B 8 42570261000 FAIRBANKS/INT 64.82 -147.87 138 249S 31HIxxno-9A 4MAIN TAIGA C 46 42570273000 ANCHORAGE/INT 61.17 -150.02 40 8U 173FLxxCO 1A 5WATER C 53 42570308000 ST. PAUL 57.15 -170.22 9 0R -9FLxxCO 1A-9WATER A 0 42570316000 COLD BAY 55.20 -162.72 31 88R -9MVxxCO 1A-9HEATHS, MOORS A 0 42570326000 KING SALMON 58.68 -156.65 15 71R -9FLxxCO30A-9TUNDRA B 7 42570361000 YAKUTAT 59.52 -139.67 9 4R -9FLMACO 3A-9WATER A 0 42570398000 ANNETTE ISLAN 55.03 -131.57 34 24R -9HIxxCO 3A-9WATER A 0 42572202000 MIAMI, FL. 25.82 -80.28 4 3U 1814FLxxCO10A 1WARM CROPS C3 118 42572203000 WEST PALM BEA 26.68 -80.10 6 6U 818FLxxCO 5A 1WARM CROPS C3 80 42572205000 ORLANDO/JETPO 28.43 -81.32 32 23U 971FLxxno-9A 2WARM FOR./FIELD C3 52 42572206000 JACKSONVILLE U/A TO WAYCRO 30.40 -81.70 9 6U 898FLxxCO30x-9WARM FOR./FIELD C3 41 42572211000 TAMPA/INT.,FL 27.97 -82.53 3 7U 1995FLxxCO 2A 1WARM CROPS C3 88 42572217000 MACON/ 32.70 -83.65 110 112U 107HIxxno-9A 8WARM FOR./FIELD C3 17 42572218000 AUGUSTA/BUSH 33.37 -81.97 45 46U 347FLxxno-9A 3WARM FOR./FIELD C3 23 42572219000 ATLANTA/MUN., 33.65 -84.42 315 285U 2960FLxxno-9A 1WARM FOR./FIELD C3 112 42572223000 MOBILE/BATES 30.68 -88.25 67 46U 443FLxxCO21A10WARM FOR./FIELD C3 28 42572226000 MONTGOMERY/DA 32.30 -86.40 62 58U 188HIxxno-9A 5WARM FOR./FIELD C3 15 42572234000 MERIDIAN/KEY, 32.33 -88.75 94 111S 41HIxxno-9A 2WARM CONIFER C2 18 42572235000 JACKSON/ALLEN 32.32 -90.08 101 95U 197HIxxno-9A 3WARM FOR./FIELD C3 26 42572240000 LAKE CHARLES/ 30.12 -93.22 10 5U 70FLxxno-9A 3MARSH, SWAMP B2 13 42572242000 GALVESTON, TX 29.30 -94.80 16 0U 59FLxxCO 2A 1WATER C3 56 42572243000 HOUSTON UNITED 29.97 -95.35 33 22U 3731FLxxno-9A 5WARM CROPS C3 48 42572248000 SHREVEPORT/RE 32.47 -93.78 79 58U 199FLxxno-9A 5WARM DECIDUOUS C3 68 42572250000 BROWNSVILLE/I 25.92 -97.42 7 6U 99FLxxCO25A 4WARM GRASS/SHRUBC3 32 42572254000 AUSTIN/ROBERT 30.30 -97.70 189 166U 846HIxxno-9A 1WARM CROPS C3 73 42572255000 VICTORIA/VICT 28.85 -96.92 36 24U 55FLxxno-9A 4WARM CROPS C3 19 42572256000 WACO,MADISON- 31.62 -97.22 155 142U 103FLxxLA-9A 4WARM FIELD WOODSC2 17 42572259000 DALLAS-FORT W 32.90 -97.03 182 161U 4037FLxxno-9A 5WARM FIELD WOODSC3 61 42572263000 SAN ANGELO/MA 31.37 -100.50 582 585U 84FLxxno-9A 6WARM GRASS/SHRUBB2 14 42572265000 MIDLAND/MIDLA 31.95 -102.18 872 866U 89FLxxno-9A10WARM GRASS/SHRUBC3 18 42572266000 ABILENE/MUN., 32.42 -99.68 546 541U 107HIxxno-9A 4WARM GRASS/SHRUBC2 19 42572267000 LUBBOCK/LUBBO 33.65 -101.82 988 978U 186FLxxno-9A 6WARM GRASS/SHRUBC3 29 42572274000 TUCSON/INT., 32.12 -110.93 779 799U 667HIxxno-9A 1WARM GRASS/SHRUBC3 46 42572278000 PHOENIX/SKY H 33.43 -112.02 337 391U 2395FLxxno-9A 1WARM GRASS/SHRUBC3 85 42572290000 SAN DIEGO/LIN 32.73 -117.17 9 39U 2498FLxxCO 1A 1WATER C3 105 42572295000 LOS ANGELES CALIFORNIA 33.72 -118.27 -999 55U14531FLxxCO 1x-9WATER C3 32 42572306000 RALEIGH/RALEI 35.87 -78.78 134 106U 856HIxxno-9A 5WARM FOR./FIELD C3 29 42572314000 CHARLOTTE/DOU 35.22 -80.93 234 207U 1162FLxxno-9A 2WARM FOR./FIELD C3 55 42572315000 ASHEVILLE/MUN 35.43 -82.55 661 675U 60MVxxno-9A10WARM DECIDUOUS C2 20 42572317000 GREENSBORO/G. 36.08 -79.95 270 257U 184HIxxno-9A 3WARM FOR./FIELD C3 55 42572324000 CHATTANOOGA/L 35.03 -85.20 210 221U 152HIxxno-9A 1WARM FOR./FIELD C3 60 42572326000 KNOXVILLE USA 35.80 -83.98 299 70U 165HIxxno-9A 8WARM CROPS C3 40 42572327000 NASHVILLE/ 36.12 -86.68 180 163U 985HIxxLA-9A 2WARM CROPS C3 83 42572344000 FORT SMITH/MU 35.33 -94.37 141 155U 73HIxxno-9A 2WARM FOR./FIELD C3 35 42572351000 WICHITA FALLS 33.97 -98.48 314 293U 96FLxxno-9A 2WARM CROPS C3 38 42572353000 OKLAHOMA CITY 35.38 -97.60 398 380U 959FLxxno-9A 3WARM CROPS C3 27 42572356000 TULSA/INT., O 36.20 -95.88 195 192U 709FLxxno-9A 1WARM CROPS C3 58 42572363000 AMARILLO/INTL 35.22 -101.72 1098 1086U 158FLxxno-9A 3WARM GRASS/SHRUBC3 31 42572365000 ALBUQUERQUE/I 35.05 -106.62 1620 1587U 589MVxxno-9A 2WARM GRASS/SHRUBC3 48 42572386000 LAS VEGAS/MCC 36.08 -115.17 664 684U 853FLxxno-9A 2HOT DESERT C3 109 42572394000 SANTA MARIA/P 34.90 -120.45 73 120U 62HIxxCO15A 2WARM CROPS C3 23 42572401000 RICHMOND/BYRD 37.50 -77.33 54 30U 866FLxxno-9A 2WARM DECIDUOUS C3 27 42572405000 WASHINGTON/NA 38.85 -77.03 20 37U 3734FLxxno-9A 1WARM FIELD WOODSC3 96 42572406000 BALTIMORE/BLT-WASHNGTN INT'L 39.18 -76.67 45 33U 2342HIxxno-9A 3WARM CROPS C3 62 42572407000 ATLANTIC CITY 39.45 -74.57 20 13U 194FLxxCO15A 4WARM CROPS C3 32 42572408000 PHILADELPHIA USA 40.00 -75.20 9 46U 5892FLxxno-9x-9WARM DECIDUOUS C3 105 42572411000 ROANOKE/MUN., 37.32 -79.97 358 353U 97MVxxno-9A 1WARM FOR./FIELD C3 64 42572412000 BECKLEY(RALEI 37.78 -81.12 766 715S 18HIxxno-9A 3COOL FOR./FIELD C3 23 42572414000 CHARLESTON/KA 38.37 -81.60 299 252U 57HIxxno-9A 2WARM MIXED C3 38 42572421000 CINCINNATI/GR 39.05 -84.67 267 253U 1818HIxxno-9A 4WARM FIELD WOODSC3 37 42572422000 LEXINGTON/BLU 38.03 -84.60 301 286U 406HIxxno-9A 3WARM FOR./FIELD C3 28 42572423000 LOUISVILLE/ 38.18 -85.73 149 148U 949HIxxno-9A 1WARM FIELD WOODSC3 124 42572425000 HUNTINGTON, W 38.37 -82.55 255 199U 55HIxxno-9A 1WARM CROPS C2 21 42572428000 COLUMBUS UNITED 40.00 -82.88 254 250U 1345FLxxno-9A 2WARM FIELD WOODSC3 73 42572429000 DAYTON/. COX, 39.90 -84.20 306 277U 951FLxxno-9A 5WARM CROPS C3 56 42572432000 EVANSVILLE/RE 38.05 -87.53 118 128U 126FLxxno-9A 4WARM FIELD WOODSC3 41 42572434000 ST.LOUIS/LAMB 38.75 -90.37 172 149U 2493FLxxno-9A 2WARM FIELD WOODSC3 104 42572438000 INDIANAPOLIS/ 39.73 -86.27 246 229U 1380FLxxno-9A 2WARM FIELD WOODSC3 88 42572440000 SPRINGFIELD/M 37.23 -93.38 387 381U 140HIxxno-9A 2WARM FIELD WOODSC3 21 42572445000 COLUMBIA/REGI 38.82 -92.22 274 239U 69HIxxno-9A15WARM CROPS B2 13 42572450000 WICHITA/MID- 37.65 -97.43 409 400U 485FLxxno-9A 1WARM CROPS C3 33 42572451000 DODGE CITY/MU 37.77 -99.97 790 776S 21FLxxno-9A 3WARM CROPS C3 28 42572456000 TOPEKA/MUN., 39.07 -95.62 270 286U 120FLxxno-9A 2WARM CROPS C2 26 42572458000 CONCORDIA/BLO 39.55 -97.65 452 439R -9FLxxno-9A-9WARM CROPS C2 10 42572465000 GOODLAND/RENN 39.37 -101.68 1124 1111R -9FLxxno-9A-9COOL GRASS/SHRUBC2 16 42572476000 GRAND JUNCTIO 39.12 -108.53 1475 1491S 29MVxxno-9A 3COOL GRASS/SHRUBC3 29 42572486000 ELY/YELLAND, 39.28 -114.85 1909 2054R -9MVDEno-9x-9COOL DESERT C3 22 42572494000 SAN FRANCISCO 37.62 -122.38 5 102U 6253FLxxCO15A 1COASTAL EDGES C3 66 42572508000 HARTFORD/BRAD 41.93 -72.68 55 43U 755HIxxno-9A 7COOL FOR./FIELD C3 38 42572509000 BOSTON/LOGAN 42.37 -71.03 9 15U 4110FLxxCO 1A 1WATER C3 98 42572513000 WILKES-BARRE- 41.33 -75.73 289 360U 737HIxxno-9A 2COOL FOR./FIELD C2 29 42572520000 PITTSBURGH/GR 40.50 -80.22 373 309U 2395HIxxno-9A 2WARM FIELD WOODSC3 43 42572521000 AKRON/AKRON-CANTON REG AP 40.92 -81.43 378 349U 653FLxxno-9A 6WARM FIELD WOODSC2 40 42572525000 YOUNGSTOWN/WSO AP 41.25 -80.67 365 324U 96FLxxno-9A 5WARM CROPS C2 23 42572530000 CHICAGO/O'HARE, ILLINOIS 42.00 -87.90 205 197U 6216FLxxno-9A 1COOL CROPS C3 125 42572532000 PEORIA/GREATE 40.67 -89.68 202 182U 114FLxxno-9A 2WARM CROPS C3 42 42572533000 FORT WAYNE/MU 41.00 -85.20 252 237U 173FLxxno-9A 3WARM CROPS C3 29 42572535000 SOUTH BEND/WSO AP 41.70 -86.32 238 223U 106FLxxno-9A 2COOL FIELD/WOODSC3 50 42572536000 TOLEDO/EXPRES 41.60 -83.80 211 183U 614FLxxno-9A 3WATER C2 32 42572537000 DETROIT MICHIGAN WBAS 42.40 -83.00 191 182U 4352FLxxno-9x-9COOL CROPS C3 93 42572544000 MOLINE/QUAD CITY ARPT 41.45 -90.50 177 195U 364FLxxno-9A 2WARM CROPS C3 36 42572546000 DES MOINES/MU 41.53 -93.65 294 267U 392FLxxno-9A 1WARM CROPS C3 45 42572552000 GRAND ISLAND/ 40.95 -98.32 566 558S 39FLxxno-9A 3WARM CROPS C3 35 42572556000 NORFOLK/KARL 41.98 -97.43 473 487S 21FLxxno-9A 3COOL GRASS/SHRUBB2 11 42572557000 SIOUX CITY/MU 42.40 -96.38 336 332U 81FLxxno-9A 4COOL CROPS C3 23 42572562000 NORTH PLATTE/ 41.13 -100.68 849 866S 23FLxxno-9A 3COOL IRRIGATED C2 13 42572569000 CASPER/NATRON 42.92 -106.47 1612 1625U 50FLxxno-9A 2COOL GRASS/SHRUBC2 15 42572572000 SALT LAKE CIT 40.78 -111.97 1288 1295U 1072FLxxno-9A 4COOL DESERT C3 59 42572576000 LANDER/HUNT, 42.82 -108.73 1694 1729R -9MVxxno-9A-9COOL GRASS/SHRUBC2 9 42572578000 POCATELLO/MUN 42.92 -112.60 1365 1372S 46HIxxLA-9A 8COOL GRASS/SHRUBC2 14 42572597000 MEDFORD/MEDFO 42.37 -122.87 405 415S 47MVxxno-9A 2WARM FOR./FIELD C3 45 42572605000 CONCORD USA 43.20 -71.50 104 127S 36HIxxno-9A 2COOL FOR./FIELD C2 41 42572635000 GRAND RAPIDS/ 42.88 -85.52 245 229U 601FLxxno-9A 4COOL FOR./FIELD C3 37 42572636000 MUSKEGON/COUNTY ARPT 43.17 -86.23 191 185S 40FLxxLA-9A 3COOL FOR./FIELD C3 29 42572637000 FLINT/BISHOP, 42.97 -83.75 233 238U 522FLxxno-9A 2COOL CROPS C3 59 42572640000 MILWAUKEE/GEN 42.95 -87.90 211 196U 1397FLxxLA-9x-9COASTAL EDGES C3 90 42572641000 MADISON/DANE 43.13 -89.33 264 278U 324HIxxLA-9x-9COOL FIELD/WOODSC3 81 42572644000 ROCHESTER/MUN 43.92 -92.50 402 337U 71FLxxno-9x-9COOL CROPS C2 20 42572645000 GREEN BAY/ 44.48 -88.13 214 194U 96FLxxLA-9A 2COOL FOR./FIELD C3 41 42572651000 SIOUX FALLS/F 43.58 -96.73 435 441U 101FLxxno-9A 2COOL CROPS C3 103 42572654000 HURON/HURON R 44.38 -98.22 393 384S 12FLxxno-9A 1COOL CROPS C3 55 42572677000 BILLINGS/LOGA 45.80 -108.53 1088 1000U 81HIxxno-9x-9COOL GRASS/SHRUBC3 123 42572681000 BOISE/MUN.,ID 43.57 -116.22 874 877U 126HIxxno-9A 2COOL CONIFER C3 37 42572688000 PENDLETON,OR. 45.68 -118.85 456 430S 15HIxxno-9A 4COOL CROPS B2 15 42572698000 PORTLAND/INT. 45.60 -122.60 12 63U 1414HIxxno-9A 1COOL CROPS C3 42 42572712000 CARIBOU/MUN., 46.87 -68.02 190 167S 10HIxxno-9A 1COOL CROPS C3 36 42572734000 SAULT STE MARIE USA 46.50 -84.40 220 209U 100HIxxLA-9A 1COOL MIXED C3 35 42572743000 MARQUETTE UNITED STATES 46.60 -87.40 219 190S 22HIxxLA-9x-9WATER C2 31 42572745000 DULUTH/INT.,M 46.83 -92.18 432 394U 85HIxxLA-9A 8COOL CONIFER C3 54 42572747000 INT.FALLS/FAL 48.57 -93.38 361 345S 10FLxxno-9A 3COOL FOR./FIELD B2 17 42572753000 FARGO/HECTOR 46.90 -96.80 274 271U 74FLxxno-9A 2COOL CROPS C3 146 42572764000 BISMARCK/MUN. 46.77 -100.75 506 514U 50FLxxno-9A 2COOL CROPS C3 32 42572767000 WILLISTON/SLO 48.18 -103.63 581 607S 13HIxxno-9A 2COOL CROPS C3 68 42572773000 MISSOULA / JO 46.92 -114.08 972 1046S 43MVxxno-9A 5COOL GRASS/SHRUBC3 27 42572781000 YAKIMA/YAKIMA 46.57 -120.53 325 385U 55HIxxno-9A 2COOL FIELD/WOODSC3 27 42572793000 SEATTLE WASHINGTON 47.60 -122.33 6 56U 2970HIxxCO 1x-9COOL FIELD/WOODSC3 107 42572797000 QUILLAYUTE,WA 47.95 -124.55 62 64R -9HIxxCO 5A-9WATER A1 0 42591165000 LIHUE, KAUAI, 21.98 -159.35 45 86R -9MVxxCO 1A-9WARM FOR./FIELD C 21 42591182000 HONOLULU, OAH 21.35 -157.93 5 240U 836HIxxCO 1A 1WARM FIELD WOODSC 70 42591285000 HILO/GEN. LYM 19.72 -155.07 11 46S 38HIxxCO 2A 1WARM FOR./FIELD C 19 [chiefio@tubularbells analysis]$
[chiefio@tubularbells analysis]$ cat USA.2008.used
42570026000 BARROW/W. POS 71.30 -156.78 4 0R -9FLMACO 1A-9TUNDRA C 40
42570133000 KOTZEBUE, RAL 66.87 -162.63 5 0R -9FLMACO 1A-9WATER A 10
42570200000 NOME 64.50 -165.43 7 48R -9HIxxCO 1A-9WATER C 23
42570231000 MCGRATH 62.97 -155.62 103 136R -9FLxxno-9A-9MAIN TAIGA B 8
42570261000 FAIRBANKS/INT 64.82 -147.87 138 249S 31HIxxno-9A 4MAIN TAIGA C 46
42570273000 ANCHORAGE/INT 61.17 -150.02 40 8U 173FLxxCO 1A 5WATER C 53
42570308000 ST. PAUL 57.15 -170.22 9 0R -9FLxxCO 1A-9WATER A 0
42570316000 COLD BAY 55.20 -162.72 31 88R -9MVxxCO 1A-9HEATHS, MOORS A 0
42570326000 KING SALMON 58.68 -156.65 15 71R -9FLxxCO30A-9TUNDRA B 7
42570361000 YAKUTAT 59.52 -139.67 9 4R -9FLMACO 3A-9WATER A 0
42570398000 ANNETTE ISLAN 55.03 -131.57 34 24R -9HIxxCO 3A-9WATER A 0
42572202000 MIAMI, FL. 25.82 -80.28 4 3U 1814FLxxCO10A 1WARM CROPS C3 118
42572203000 WEST PALM BEA 26.68 -80.10 6 6U 818FLxxCO 5A 1WARM CROPS C3 80
42572205000 ORLANDO/JETPO 28.43 -81.32 32 23U 971FLxxno-9A 2WARM FOR./FIELD C3 52
42572206000 JACKSONVILLE U/A TO WAYCRO 30.40 -81.70 9 6U 898FLxxCO30x-9WARM FOR./FIELD C3 41
42572211000 TAMPA/INT.,FL 27.97 -82.53 3 7U 1995FLxxCO 2A 1WARM CROPS C3 88
42572217000 MACON/ 32.70 -83.65 110 112U 107HIxxno-9A 8WARM FOR./FIELD C3 17
42572218000 AUGUSTA/BUSH 33.37 -81.97 45 46U 347FLxxno-9A 3WARM FOR./FIELD C3 23
42572219000 ATLANTA/MUN., 33.65 -84.42 315 285U 2960FLxxno-9A 1WARM FOR./FIELD C3 112
42572223000 MOBILE/BATES 30.68 -88.25 67 46U 443FLxxCO21A10WARM FOR./FIELD C3 28
42572226000 MONTGOMERY/DA 32.30 -86.40 62 58U 188HIxxno-9A 5WARM FOR./FIELD C3 15
42572234000 MERIDIAN/KEY, 32.33 -88.75 94 111S 41HIxxno-9A 2WARM CONIFER C2 18
42572235000 JACKSON/ALLEN 32.32 -90.08 101 95U 197HIxxno-9A 3WARM FOR./FIELD C3 26
42572240000 LAKE CHARLES/ 30.12 -93.22 10 5U 70FLxxno-9A 3MARSH, SWAMP B2 13
42572242000 GALVESTON, TX 29.30 -94.80 16 0U 59FLxxCO 2A 1WATER C3 56
42572243000 HOUSTON UNITED 29.97 -95.35 33 22U 3731FLxxno-9A 5WARM CROPS C3 48
42572248000 SHREVEPORT/RE 32.47 -93.78 79 58U 199FLxxno-9A 5WARM DECIDUOUS C3 68
42572250000 BROWNSVILLE/I 25.92 -97.42 7 6U 99FLxxCO25A 4WARM GRASS/SHRUBC3 32
42572254000 AUSTIN/ROBERT 30.30 -97.70 189 166U 846HIxxno-9A 1WARM CROPS C3 73
42572255000 VICTORIA/VICT 28.85 -96.92 36 24U 55FLxxno-9A 4WARM CROPS C3 19
42572256000 WACO,MADISON- 31.62 -97.22 155 142U 103FLxxLA-9A 4WARM FIELD WOODSC2 17
42572259000 DALLAS-FORT W 32.90 -97.03 182 161U 4037FLxxno-9A 5WARM FIELD WOODSC3 61
42572263000 SAN ANGELO/MA 31.37 -100.50 582 585U 84FLxxno-9A 6WARM GRASS/SHRUBB2 14
42572265000 MIDLAND/MIDLA 31.95 -102.18 872 866U 89FLxxno-9A10WARM GRASS/SHRUBC3 18
42572266000 ABILENE/MUN., 32.42 -99.68 546 541U 107HIxxno-9A 4WARM GRASS/SHRUBC2 19
42572267000 LUBBOCK/LUBBO 33.65 -101.82 988 978U 186FLxxno-9A 6WARM GRASS/SHRUBC3 29
42572274000 TUCSON/INT., 32.12 -110.93 779 799U 667HIxxno-9A 1WARM GRASS/SHRUBC3 46
42572278000 PHOENIX/SKY H 33.43 -112.02 337 391U 2395FLxxno-9A 1WARM GRASS/SHRUBC3 85
42572290000 SAN DIEGO/LIN 32.73 -117.17 9 39U 2498FLxxCO 1A 1WATER C3 105
42572295000 LOS ANGELES CALIFORNIA 33.72 -118.27 -999 55U14531FLxxCO 1x-9WATER C3 32
42572306000 RALEIGH/RALEI 35.87 -78.78 134 106U 856HIxxno-9A 5WARM FOR./FIELD C3 29
42572314000 CHARLOTTE/DOU 35.22 -80.93 234 207U 1162FLxxno-9A 2WARM FOR./FIELD C3 55
42572315000 ASHEVILLE/MUN 35.43 -82.55 661 675U 60MVxxno-9A10WARM DECIDUOUS C2 20
42572317000 GREENSBORO/G. 36.08 -79.95 270 257U 184HIxxno-9A 3WARM FOR./FIELD C3 55
42572324000 CHATTANOOGA/L 35.03 -85.20 210 221U 152HIxxno-9A 1WARM FOR./FIELD C3 60
42572326000 KNOXVILLE USA 35.80 -83.98 299 70U 165HIxxno-9A 8WARM CROPS C3 40
42572327000 NASHVILLE/ 36.12 -86.68 180 163U 985HIxxLA-9A 2WARM CROPS C3 83
42572344000 FORT SMITH/MU 35.33 -94.37 141 155U 73HIxxno-9A 2WARM FOR./FIELD C3 35
42572351000 WICHITA FALLS 33.97 -98.48 314 293U 96FLxxno-9A 2WARM CROPS C3 38
42572353000 OKLAHOMA CITY 35.38 -97.60 398 380U 959FLxxno-9A 3WARM CROPS C3 27
42572356000 TULSA/INT., O 36.20 -95.88 195 192U 709FLxxno-9A 1WARM CROPS C3 58
42572363000 AMARILLO/INTL 35.22 -101.72 1098 1086U 158FLxxno-9A 3WARM GRASS/SHRUBC3 31
42572365000 ALBUQUERQUE/I 35.05 -106.62 1620 1587U 589MVxxno-9A 2WARM GRASS/SHRUBC3 48
42572386000 LAS VEGAS/MCC 36.08 -115.17 664 684U 853FLxxno-9A 2HOT DESERT C3 109
42572394000 SANTA MARIA/P 34.90 -120.45 73 120U 62HIxxCO15A 2WARM CROPS C3 23
42572401000 RICHMOND/BYRD 37.50 -77.33 54 30U 866FLxxno-9A 2WARM DECIDUOUS C3 27
42572405000 WASHINGTON/NA 38.85 -77.03 20 37U 3734FLxxno-9A 1WARM FIELD WOODSC3 96
42572406000 BALTIMORE/BLT-WASHNGTN INT’L 39.18 -76.67 45 33U 2342HIxxno-9A 3WARM CROPS C3 62
42572407000 ATLANTIC CITY 39.45 -74.57 20 13U 194FLxxCO15A 4WARM CROPS C3 32
42572408000 PHILADELPHIA USA 40.00 -75.20 9 46U 5892FLxxno-9x-9WARM DECIDUOUS C3 105
42572411000 ROANOKE/MUN., 37.32 -79.97 358 353U 97MVxxno-9A 1WARM FOR./FIELD C3 64
42572412000 BECKLEY(RALEI 37.78 -81.12 766 715S 18HIxxno-9A 3COOL FOR./FIELD C3 23
42572414000 CHARLESTON/KA 38.37 -81.60 299 252U 57HIxxno-9A 2WARM MIXED C3 38
42572421000 CINCINNATI/GR 39.05 -84.67 267 253U 1818HIxxno-9A 4WARM FIELD WOODSC3 37
42572422000 LEXINGTON/BLU 38.03 -84.60 301 286U 406HIxxno-9A 3WARM FOR./FIELD C3 28
42572423000 LOUISVILLE/ 38.18 -85.73 149 148U 949HIxxno-9A 1WARM FIELD WOODSC3 124
42572425000 HUNTINGTON, W 38.37 -82.55 255 199U 55HIxxno-9A 1WARM CROPS C2 21
42572428000 COLUMBUS UNITED 40.00 -82.88 254 250U 1345FLxxno-9A 2WARM FIELD WOODSC3 73
42572429000 DAYTON/. COX, 39.90 -84.20 306 277U 951FLxxno-9A 5WARM CROPS C3 56
42572432000 EVANSVILLE/RE 38.05 -87.53 118 128U 126FLxxno-9A 4WARM FIELD WOODSC3 41
42572434000 ST.LOUIS/LAMB 38.75 -90.37 172 149U 2493FLxxno-9A 2WARM FIELD WOODSC3 104
42572438000 INDIANAPOLIS/ 39.73 -86.27 246 229U 1380FLxxno-9A 2WARM FIELD WOODSC3 88
42572440000 SPRINGFIELD/M 37.23 -93.38 387 381U 140HIxxno-9A 2WARM FIELD WOODSC3 21
42572445000 COLUMBIA/REGI 38.82 -92.22 274 239U 69HIxxno-9A15WARM CROPS B2 13
42572450000 WICHITA/MID- 37.65 -97.43 409 400U 485FLxxno-9A 1WARM CROPS C3 33
42572451000 DODGE CITY/MU 37.77 -99.97 790 776S 21FLxxno-9A 3WARM CROPS C3 28
42572456000 TOPEKA/MUN., 39.07 -95.62 270 286U 120FLxxno-9A 2WARM CROPS C2 26
42572458000 CONCORDIA/BLO 39.55 -97.65 452 439R -9FLxxno-9A-9WARM CROPS C2 10
42572465000 GOODLAND/RENN 39.37 -101.68 1124 1111R -9FLxxno-9A-9COOL GRASS/SHRUBC2 16
42572476000 GRAND JUNCTIO 39.12 -108.53 1475 1491S 29MVxxno-9A 3COOL GRASS/SHRUBC3 29
42572486000 ELY/YELLAND, 39.28 -114.85 1909 2054R -9MVDEno-9x-9COOL DESERT C3 22
42572494000 SAN FRANCISCO 37.62 -122.38 5 102U 6253FLxxCO15A 1COASTAL EDGES C3 66
42572508000 HARTFORD/BRAD 41.93 -72.68 55 43U 755HIxxno-9A 7COOL FOR./FIELD C3 38
42572509000 BOSTON/LOGAN 42.37 -71.03 9 15U 4110FLxxCO 1A 1WATER C3 98
42572513000 WILKES-BARRE- 41.33 -75.73 289 360U 737HIxxno-9A 2COOL FOR./FIELD C2 29
42572520000 PITTSBURGH/GR 40.50 -80.22 373 309U 2395HIxxno-9A 2WARM FIELD WOODSC3 43
42572521000 AKRON/AKRON-CANTON REG AP 40.92 -81.43 378 349U 653FLxxno-9A 6WARM FIELD WOODSC2 40
42572525000 YOUNGSTOWN/WSO AP 41.25 -80.67 365 324U 96FLxxno-9A 5WARM CROPS C2 23
42572530000 CHICAGO/O’HARE, ILLINOIS 42.00 -87.90 205 197U 6216FLxxno-9A 1COOL CROPS C3 125
42572532000 PEORIA/GREATE 40.67 -89.68 202 182U 114FLxxno-9A 2WARM CROPS C3 42
42572533000 FORT WAYNE/MU 41.00 -85.20 252 237U 173FLxxno-9A 3WARM CROPS C3 29
42572535000 SOUTH BEND/WSO AP 41.70 -86.32 238 223U 106FLxxno-9A 2COOL FIELD/WOODSC3 50
42572536000 TOLEDO/EXPRES 41.60 -83.80 211 183U 614FLxxno-9A 3WATER C2 32
42572537000 DETROIT MICHIGAN WBAS 42.40 -83.00 191 182U 4352FLxxno-9x-9COOL CROPS C3 93
42572544000 MOLINE/QUAD CITY ARPT 41.45 -90.50 177 195U 364FLxxno-9A 2WARM CROPS C3 36
42572546000 DES MOINES/MU 41.53 -93.65 294 267U 392FLxxno-9A 1WARM CROPS C3 45
42572552000 GRAND ISLAND/ 40.95 -98.32 566 558S 39FLxxno-9A 3WARM CROPS C3 35
42572556000 NORFOLK/KARL 41.98 -97.43 473 487S 21FLxxno-9A 3COOL GRASS/SHRUBB2 11
42572557000 SIOUX CITY/MU 42.40 -96.38 336 332U 81FLxxno-9A 4COOL CROPS C3 23
42572562000 NORTH PLATTE/ 41.13 -100.68 849 866S 23FLxxno-9A 3COOL IRRIGATED C2 13
42572569000 CASPER/NATRON 42.92 -106.47 1612 1625U 50FLxxno-9A 2COOL GRASS/SHRUBC2 15
42572572000 SALT LAKE CIT 40.78 -111.97 1288 1295U 1072FLxxno-9A 4COOL DESERT C3 59
42572576000 LANDER/HUNT, 42.82 -108.73 1694 1729R -9MVxxno-9A-9COOL GRASS/SHRUBC2 9
42572578000 POCATELLO/MUN 42.92 -112.60 1365 1372S 46HIxxLA-9A 8COOL GRASS/SHRUBC2 14
42572597000 MEDFORD/MEDFO 42.37 -122.87 405 415S 47MVxxno-9A 2WARM FOR./FIELD C3 45
42572605000 CONCORD USA 43.20 -71.50 104 127S 36HIxxno-9A 2COOL FOR./FIELD C2 41
42572635000 GRAND RAPIDS/ 42.88 -85.52 245 229U 601FLxxno-9A 4COOL FOR./FIELD C3 37
42572636000 MUSKEGON/COUNTY ARPT 43.17 -86.23 191 185S 40FLxxLA-9A 3COOL FOR./FIELD C3 29
42572637000 FLINT/BISHOP, 42.97 -83.75 233 238U 522FLxxno-9A 2COOL CROPS C3 59
42572640000 MILWAUKEE/GEN 42.95 -87.90 211 196U 1397FLxxLA-9x-9COASTAL EDGES C3 90
42572641000 MADISON/DANE 43.13 -89.33 264 278U 324HIxxLA-9x-9COOL FIELD/WOODSC3 81
42572644000 ROCHESTER/MUN 43.92 -92.50 402 337U 71FLxxno-9x-9COOL CROPS C2 20
42572645000 GREEN BAY/ 44.48 -88.13 214 194U 96FLxxLA-9A 2COOL FOR./FIELD C3 41
42572651000 SIOUX FALLS/F 43.58 -96.73 435 441U 101FLxxno-9A 2COOL CROPS C3 103
42572654000 HURON/HURON R 44.38 -98.22 393 384S 12FLxxno-9A 1COOL CROPS C3 55
42572677000 BILLINGS/LOGA 45.80 -108.53 1088 1000U 81HIxxno-9x-9COOL GRASS/SHRUBC3 123
42572681000 BOISE/MUN.,ID 43.57 -116.22 874 877U 126HIxxno-9A 2COOL CONIFER C3 37
42572688000 PENDLETON,OR. 45.68 -118.85 456 430S 15HIxxno-9A 4COOL CROPS B2 15
42572698000 PORTLAND/INT. 45.60 -122.60 12 63U 1414HIxxno-9A 1COOL CROPS C3 42
42572712000 CARIBOU/MUN., 46.87 -68.02 190 167S 10HIxxno-9A 1COOL CROPS C3 36
42572734000 SAULT STE MARIE USA 46.50 -84.40 220 209U 100HIxxLA-9A 1COOL MIXED C3 35
42572743000 MARQUETTE UNITED STATES 46.60 -87.40 219 190S 22HIxxLA-9x-9WATER C2 31
42572745000 DULUTH/INT.,M 46.83 -92.18 432 394U 85HIxxLA-9A 8COOL CONIFER C3 54
42572747000 INT.FALLS/FAL 48.57 -93.38 361 345S 10FLxxno-9A 3COOL FOR./FIELD B2 17
42572753000 FARGO/HECTOR 46.90 -96.80 274 271U 74FLxxno-9A 2COOL CROPS C3 146
42572764000 BISMARCK/MUN. 46.77 -100.75 506 514U 50FLxxno-9A 2COOL CROPS C3 32
42572767000 WILLISTON/SLO 48.18 -103.63 581 607S 13HIxxno-9A 2COOL CROPS C3 68
42572773000 MISSOULA / JO 46.92 -114.08 972 1046S 43MVxxno-9A 5COOL GRASS/SHRUBC3 27
42572781000 YAKIMA/YAKIMA 46.57 -120.53 325 385U 55HIxxno-9A 2COOL FIELD/WOODSC3 27
42572793000 SEATTLE WASHINGTON 47.60 -122.33 6 56U 2970HIxxCO 1x-9COOL FIELD/WOODSC3 107
42572797000 QUILLAYUTE,WA 47.95 -124.55 62 64R -9HIxxCO 5A-9WATER A1 0
42591165000 LIHUE, KAUAI, 21.98 -159.35 45 86R -9MVxxCO 1A-9WARM FOR./FIELD C 21
42591182000 HONOLULU, OAH 21.35 -157.93 5 240U 836HIxxCO 1A 1WARM FIELD WOODSC 70
42591285000 HILO/GEN. LYM 19.72 -155.07 11 46S 38HIxxCO 2A 1WARM FOR./FIELD C 19
[chiefio@tubularbells analysis]$
Thermometer Records by Year
Normally I would have taken the time to replace the 0.0 entries that are in fact missing data in the early part of this listing with -999, but in this case, the actual temperature numbers are the less important bit. It’s that number of thermometers on the far right that crashes in 2007.
The format here is: Year, 12 monthly averages of daily MIN-MAX averages, the average of those daily MIN-MAX averages for the whole year, and finally the number of thermometer records in that year. We start with one thermometer for the USA in 1743, then ramp up to 1850 in 1963, then suddenly crash to 136 in 2007.
[chiefio@tubularbells analysis]$ cat v2.mean.sorted.425.yrs.GAT
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
--------------------------------------------------------------------------
1743 0.0 0.0 0.0 7.2 15.4 19.9 24.4 0.0 19.0 0.0 5.6 0.0 7.6 1
1744 0.0 0.0 0.0 9.9 16.6 22.2 23.1 0.0 15.8 11.1 5.1 -0.7 8.6 1
1745 0.9 -0.8 2.9 8.8 14.8 20.4 23.4 0.0 19.6 0.0 7.3 3.0 8.4 1
1746 -1.2 1.9 0.7 10.2 0.0 18.3 0.0 0.0 0.0 0.0 0.0 0.0 2.5 1
1747 0.0 0.0 0.0 9.7 0.0 20.5 24.3 0.0 17.8 12.2 6.6 0.0 7.6 1
1748 0.0 0.0 0.0 10.2 15.6 0.0 0.0 0.0 17.9 10.4 4.8 3.6 5.2 1
1749 -4.2 -2.0 2.7 8.0 0.0 18.1 0.0 0.0 14.7 8.4 3.7 -1.9 4.0 1
1750 -2.7 -1.9 2.3 7.4 12.6 0.0 0.0 0.0 15.3 8.3 2.7 1.1 3.8 1
1751 -1.3 0.3 2.8 6.3 0.0 16.4 0.0 0.0 0.0 0.0 4.6 0.0 2.4 1
1752 -6.1 0.0 2.8 5.5 0.0 0.0 0.0 0.0 0.0 8.1 5.9 -0.5 1.3 1
1753 -0.3 0.9 4.1 6.4 12.1 16.1 0.0 0.0 0.0 9.1 3.1 -0.6 4.2 1
1754 1.1 0.2 2.8 0.0 0.0 0.0 17.3 0.0 13.7 10.2 0.0 0.7 3.8 1
1755 0.0 0.7 0.4 6.7 11.7 0.0 0.0 0.0 0.0 0.0 3.1 -0.6 1.8 1
1756 0.4 1.1 0.0 0.0 0.0 15.6 0.0 0.0 14.6 8.6 2.9 -1.1 3.5 1
1757 -1.8 -0.6 1.7 5.6 11.7 17.3 0.0 0.0 0.0 8.8 3.1 1.3 3.9 1
1758 -1.3 -2.4 3.4 8.9 14.1 22.3 20.6 22.5 19.0 0.0 5.4 0.0 9.4 2
1759 -0.4 2.8 4.5 9.3 15.1 18.8 23.2 22.6 19.9 12.2 6.4 -1.8 11.0 2
1760 -0.6 0.6 1.8 7.5 11.2 0.0 0.0 0.0 0.0 8.1 3.3 -0.3 2.6 2
1761 -5.1 -0.9 3.2 7.3 15.7 22.4 24.9 0.0 0.0 0.0 4.8 -3.6 5.7 1
1762 0.0 -2.4 0.1 7.7 13.2 16.3 0.0 0.0 13.9 7.8 3.8 1.3 5.1 1
1763 -5.2 -2.8 1.0 6.6 11.0 0.0 0.0 0.0 13.8 0.0 3.1 0.2 2.3 1
1764 -3.1 0.6 2.3 0.0 0.0 18.6 22.7 21.4 15.1 14.0 6.3 2.4 8.4 1
1765 -5.3 -0.1 3.9 10.0 15.8 0.0 22.8 22.5 17.2 10.6 5.2 -0.7 8.5 1
1766 -3.6 1.2 2.6 8.6 15.6 22.1 23.9 23.7 19.4 10.6 5.3 -1.9 10.6 1
1767 -2.6 -2.2 1.9 9.2 13.6 0.0 23.9 24.1 16.4 10.6 4.7 -3.1 8.0 1
1768 -1.8 3.3 2.2 7.2 14.6 19.2 23.0 22.1 16.8 10.3 4.6 -0.5 10.1 2
1769 -0.1 -1.1 3.2 9.1 14.3 22.0 24.0 22.6 17.6 9.1 3.3 -1.1 10.2 2
1770 -3.7 -0.5 1.7 9.6 16.1 21.1 24.4 23.3 17.6 9.8 5.6 -0.3 10.4 2
1771 1.2 -2.5 4.4 10.0 17.6 21.4 24.0 24.0 17.9 13.9 7.5 -1.9 11.5 2
1772 -3.7 -0.8 -1.1 8.6 14.3 20.7 24.4 21.7 15.4 12.0 5.5 2.6 10.0 1
1773 -0.7 -3.5 4.0 10.6 17.7 22.2 25.3 23.2 18.7 13.2 4.6 1.8 11.4 2
1774 -6.8 -2.2 3.0 0.0 0.0 20.3 23.7 22.5 16.9 14.6 5.6 -0.7 8.1 1
1775 0.0 2.4 5.8 10.3 19.4 19.9 24.4 22.6 18.4 12.1 4.8 -2.4 11.5 2
1776 -3.7 -2.6 1.8 8.3 0.0 21.1 23.0 21.8 19.9 13.5 7.9 1.8 9.4 2
1777 -3.3 -2.4 2.7 7.4 12.4 20.0 21.0 22.7 15.5 11.3 3.2 -1.0 9.1 2
1778 -2.2 0.0 0.0 0.0 13.8 20.3 0.0 0.0 0.0 12.9 2.4 -3.1 3.7 1
1779 -3.0 3.9 1.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.2 1
1781 0.9 0.6 3.1 7.8 14.3 18.8 23.4 22.4 16.9 11.6 3.0 -0.3 10.2 1
1782 -4.2 -3.1 1.2 10.3 15.2 20.8 21.1 22.7 18.5 8.7 3.2 -0.6 9.5 1
1783 -3.5 -0.1 1.7 8.4 14.7 20.8 21.5 21.5 14.3 8.9 2.2 -1.3 9.1 1
1784 -6.2 -6.6 0.8 5.3 13.8 19.7 21.8 22.3 16.8 9.4 6.2 -1.7 8.5 1
1785 -4.6 -2.9 -1.0 7.1 12.2 20.5 21.9 21.4 15.5 10.6 4.5 -0.5 8.7 1
1786 -4.4 -1.4 4.6 7.7 13.6 20.6 22.2 19.8 16.8 11.9 2.2 -3.4 9.2 1
1787 -2.6 -2.4 3.5 8.9 13.1 18.7 20.8 20.7 15.7 8.1 5.8 -0.5 9.1 1
1788 -4.0 -3.3 2.4 9.4 14.3 19.4 23.1 21.9 17.6 14.2 7.9 -1.4 10.1 1
1789 -2.2 -4.8 2.6 8.3 12.7 20.4 23.2 22.8 17.4 8.4 6.1 1.7 9.7 1
1790 0.2 -0.9 2.1 7.1 14.7 20.9 21.6 22.4 17.8 11.2 4.7 -5.3 9.7 1
1791 -2.6 -5.0 4.8 10.0 16.0 20.6 22.4 22.2 16.7 9.0 4.7 -0.5 9.9 2
1792 -7.2 -2.9 4.2 9.3 15.6 18.8 21.9 20.2 15.5 11.8 5.5 -2.5 9.2 2
1793 -1.5 -1.3 4.0 10.2 15.8 21.2 23.3 22.9 17.9 11.1 4.9 -0.7 10.7 2
1794 -2.5 -2.9 4.2 9.9 16.0 19.5 22.4 22.2 18.8 9.7 4.9 4.5 10.6 2
1795 -3.4 -2.8 2.4 9.0 16.0 19.0 23.0 23.0 17.6 11.2 3.9 1.0 10.0 3
1796 -2.6 -3.2 0.1 9.0 13.5 19.6 22.3 20.8 16.5 9.4 3.3 -5.0 8.6 3
1797 -5.4 0.8 2.2 8.1 12.7 19.4 23.2 20.7 15.8 8.7 2.0 -4.0 8.7 2
1798 -1.5 -2.5 3.4 9.6 16.6 21.0 22.8 25.1 18.6 12.2 3.0 -3.3 10.4 3
1799 1.2 0.3 3.0 11.4 16.5 22.0 24.2 23.5 19.3 12.9 7.6 1.4 11.9 4
1800 0.0 0.3 5.5 13.2 16.6 22.1 24.9 23.4 19.6 13.5 6.1 3.3 12.4 4
1801 1.3 3.1 8.0 10.7 18.0 22.0 24.5 22.8 20.9 15.2 7.0 3.1 13.0 4
1802 5.0 2.9 7.2 12.7 16.0 21.7 23.8 23.7 20.4 14.9 7.4 2.1 13.1 4
1803 0.9 3.4 6.8 12.7 15.3 22.7 25.0 24.0 18.9 15.3 6.5 5.0 13.0 4
1804 -3.7 -1.1 2.0 8.4 16.3 20.8 23.3 23.0 20.1 11.0 6.3 -1.3 10.4 3
1805 -5.2 -1.3 4.5 9.9 14.7 20.1 23.5 23.1 19.3 9.3 3.8 3.6 10.4 2
1806 -3.3 0.0 -0.8 5.9 15.0 20.0 21.2 20.6 17.4 10.2 4.4 -1.1 9.1 2
1807 -5.1 -3.7 -0.2 7.2 13.0 18.1 23.7 21.9 16.1 11.1 3.0 2.3 8.9 2
1808 -3.9 -0.8 3.2 8.8 13.0 20.1 22.4 20.2 16.4 9.5 4.8 0.2 9.5 2
1809 -5.7 -4.5 1.0 8.3 14.2 19.3 20.0 20.3 14.8 14.4 1.8 1.8 8.8 2
1810 -3.3 -0.1 1.3 9.6 14.7 20.1 20.9 21.1 17.4 10.5 3.7 -1.6 9.5 2
1811 -3.1 -3.1 4.6 7.7 13.4 20.0 21.7 20.7 17.4 13.0 4.9 -1.0 9.7 2
1812 -5.9 -3.2 -1.1 7.2 10.4 17.8 20.5 19.9 14.6 10.3 4.1 -1.3 7.8 3
1813 -4.5 -2.8 -0.3 8.6 12.4 19.1 21.3 22.2 18.8 10.1 5.4 -1.1 9.1 4
1814 -3.9 -1.0 1.1 8.8 16.1 18.8 20.9 20.9 16.8 11.0 5.4 -1.5 9.4 3
1815 -3.6 -4.0 2.7 6.8 11.2 18.0 21.8 19.2 16.0 10.1 6.0 -1.3 8.6 2
1816 -3.1 -2.0 0.2 6.4 11.3 15.8 18.4 19.6 14.5 10.8 6.3 -0.3 8.2 3
1817 -3.5 -5.7 1.3 8.8 12.8 17.5 20.8 20.9 17.3 9.5 6.1 0.2 8.8 4
1818 -2.1 -4.3 2.7 6.4 13.2 20.3 22.9 20.9 16.4 10.7 7.2 -2.3 9.3 3
1819 0.5 1.3 0.5 8.0 13.2 20.4 22.2 22.4 19.4 9.4 5.3 -1.6 10.1 4
1820 -6.8 0.4 1.8 10.3 14.2 20.6 22.9 21.3 18.0 9.9 3.0 -3.5 9.3 9
1821 -7.3 -0.9 1.3 6.6 14.5 21.2 21.0 23.0 18.1 11.0 4.1 -2.8 9.1 10
1822 -3.6 -0.6 6.0 9.9 17.1 21.2 23.5 22.2 19.1 12.1 7.3 -1.7 11.0 13
1823 -1.1 -4.1 3.4 10.7 15.0 20.0 23.0 22.1 17.2 11.0 3.7 0.5 10.1 14
1824 0.8 -0.6 3.8 9.6 14.9 19.9 23.0 21.4 18.0 11.5 5.3 2.0 10.8 17
1825 0.2 2.5 7.8 12.0 16.7 22.5 25.0 22.8 18.4 13.7 6.4 0.9 12.4 18
1826 -0.5 1.2 5.5 9.1 19.5 21.8 23.5 22.7 18.9 13.4 6.7 1.3 11.9 17
1827 -2.9 2.0 5.3 11.7 15.8 20.0 23.1 21.9 18.2 12.6 4.6 1.7 11.2 21
1828 1.0 4.0 5.5 8.6 16.1 22.3 22.9 22.7 17.7 12.0 7.4 3.9 12.0 24
1829 -1.4 -3.4 2.3 9.8 17.2 20.9 21.7 21.8 15.9 12.0 4.6 3.9 10.4 29
1830 -1.4 0.2 5.9 12.8 16.0 20.4 24.4 22.5 17.8 13.7 10.0 2.4 12.1 30
1831 -3.3 -1.9 6.6 11.3 16.8 22.4 23.2 22.8 18.2 13.1 6.1 -5.2 10.8 31
1832 -0.8 -0.7 4.6 8.9 14.3 19.8 22.0 21.5 17.5 12.5 6.7 2.1 10.7 36
1833 1.5 0.0 3.7 11.6 16.8 18.6 22.3 20.9 18.1 11.0 5.3 2.1 11.0 34
1834 -2.1 4.2 5.6 11.4 15.1 19.7 23.7 22.0 17.9 11.1 5.8 0.3 11.2 36
1835 -0.5 -3.2 3.0 9.1 15.6 19.8 22.0 21.1 15.7 12.6 5.5 -1.9 9.9 34
1836 -1.9 -3.6 0.8 7.6 14.9 18.1 21.6 19.4 17.0 7.6 3.4 -1.5 8.6 38
1837 -3.4 -0.6 2.1 7.7 13.9 19.0 21.6 20.6 16.8 11.0 6.5 0.7 9.7 41
1838 1.4 -4.7 5.3 8.0 13.7 21.1 23.6 21.8 17.7 10.2 3.2 -1.7 10.0 38
1839 -1.5 0.4 3.5 11.3 15.3 18.2 22.6 20.6 16.6 12.6 3.5 -0.2 10.2 39
1840 -4.2 2.4 5.1 11.3 15.9 19.7 22.2 21.6 16.4 11.1 5.1 -0.3 10.5 42
1841 -0.1 -1.5 3.4 7.7 14.1 20.5 21.7 21.2 17.8 9.5 5.1 1.0 10.0 43
1842 0.0 1.5 6.4 10.8 13.6 18.1 21.3 20.2 16.4 10.8 3.2 -0.9 10.1 44
1843 1.8 -3.7 -0.9 9.4 14.6 19.1 21.5 21.4 18.3 9.8 4.4 1.4 9.8 48
1844 -3.6 0.0 4.5 12.7 16.2 19.4 21.8 20.7 17.4 10.8 5.4 0.6 10.5 50
1845 0.6 0.7 5.5 10.9 14.8 20.0 22.5 21.9 17.0 11.4 5.7 -2.6 10.7 47
1846 0.1 -1.7 5.2 11.6 17.3 20.1 23.0 23.2 20.5 11.4 7.9 1.6 11.7 43
1847 -1.6 -0.5 1.8 9.4 15.0 19.6 22.9 21.2 17.4 10.8 6.8 1.7 10.4 43
1848 0.4 -0.2 3.0 9.5 16.4 20.3 21.6 21.5 15.9 11.5 3.5 2.7 10.5 42
1849 -2.8 -2.8 4.9 9.3 14.8 20.9 22.2 22.0 18.1 12.5 9.8 1.3 10.8 48
1850 2.2 2.7 5.1 9.7 14.5 20.4 23.4 22.6 18.9 13.2 8.2 1.1 11.8 48
1851 1.8 3.7 7.2 10.8 16.2 19.9 22.5 21.4 18.8 13.1 5.7 0.0 11.8 54
1852 -1.4 2.3 5.0 9.1 16.6 20.4 22.9 21.5 18.0 14.1 6.2 3.9 11.5 56
1853 2.0 2.8 6.8 12.3 16.5 21.4 22.4 22.0 19.4 13.0 9.7 3.4 12.6 46
1854 1.0 2.5 7.3 11.4 17.0 21.0 24.5 23.1 20.0 14.6 7.7 1.9 12.7 59
1855 2.1 -0.4 5.4 12.9 17.6 20.4 23.7 22.5 20.2 12.7 8.5 1.9 12.3 64
1856 -3.9 -0.8 2.6 12.2 15.8 22.3 24.1 21.5 18.4 12.8 6.5 -0.4 10.9 74
1857 -4.0 4.9 4.6 8.3 15.2 20.0 22.8 22.1 19.1 12.6 6.6 4.9 11.4 82
1858 3.7 -0.2 6.3 11.5 16.0 21.9 23.4 22.2 18.9 14.2 5.3 3.1 12.2 81
1859 1.3 3.6 7.9 9.9 17.4 20.0 22.8 22.0 17.8 11.5 7.9 -0.3 11.8 77
1860 1.2 1.8 7.6 11.5 17.5 21.2 23.3 22.6 18.2 13.4 6.9 0.9 12.2 59
1861 -0.3 3.6 5.8 11.2 14.5 20.5 21.7 21.2 17.8 12.8 6.3 3.0 11.5 71
1862 -1.1 -0.9 4.0 9.5 15.7 19.3 22.2 22.1 18.5 12.6 5.7 2.6 10.8 64
1863 1.5 0.6 3.3 10.1 16.4 18.8 22.1 21.7 16.9 10.5 6.3 0.6 10.7 67
1864 -1.4 1.3 3.7 9.0 16.6 20.4 23.2 22.4 17.5 10.3 5.3 -0.6 10.6 77
1865 -2.6 0.6 5.4 10.7 16.2 21.6 21.6 21.5 20.5 11.5 6.9 0.1 11.2 75
1866 -1.8 -0.5 3.3 11.5 14.4 19.9 23.5 19.8 17.0 12.5 6.7 0.0 10.5 82
1867 -3.3 1.8 0.9 10.2 13.6 21.4 22.5 22.6 18.4 12.4 6.8 0.5 10.7 88
1868 -3.6 -1.1 6.2 8.9 15.2 20.1 24.5 21.3 16.4 10.9 5.1 -1.2 10.2 97
1869 0.8 0.9 2.4 9.6 14.8 19.4 22.1 21.8 17.5 8.7 3.9 0.4 10.2 103
1870 0.6 0.3 2.6 10.9 16.6 21.2 23.9 21.9 18.8 12.5 6.3 -0.4 11.3 125
1871 -0.2 1.2 7.5 11.8 16.7 21.4 22.7 22.9 17.1 13.2 4.3 -0.6 11.5 146
1872 -0.9 1.0 2.8 11.2 17.0 21.5 23.8 23.3 18.7 12.1 3.9 -1.8 11.0 163
1873 -1.3 0.4 4.8 9.8 15.6 21.5 23.3 22.5 18.0 11.3 4.9 2.0 11.1 171
1874 1.1 1.1 5.0 8.3 17.0 21.8 23.8 22.4 19.3 13.0 6.2 2.2 11.8 172
1875 -3.7 -2.1 3.5 9.4 16.5 20.8 23.2 21.8 17.8 12.2 5.3 3.9 10.7 176
1876 2.4 2.5 3.7 10.8 16.2 21.5 23.9 22.7 17.9 11.5 6.0 -2.0 11.4 177
1877 -0.7 4.4 5.2 11.1 16.0 21.3 24.0 23.1 19.6 13.5 7.0 5.2 12.5 179
1878 1.8 4.3 9.7 13.6 16.6 21.1 25.0 24.0 19.2 13.4 7.9 0.5 13.1 192
1879 -0.4 1.7 7.9 11.8 17.6 21.3 24.6 22.9 18.5 15.6 7.1 2.2 12.6 201
1880 5.0 3.2 5.6 11.8 18.7 21.9 23.7 22.8 18.7 12.5 3.2 0.0 12.3 203
1881 -2.0 1.4 5.3 11.0 18.5 21.1 24.1 23.7 20.5 13.8 6.4 4.8 12.4 212
1882 0.9 4.0 6.5 11.9 15.5 21.4 23.1 23.1 19.3 14.6 6.1 1.2 12.3 259
1883 -2.5 0.3 4.5 11.8 16.0 22.0 23.8 22.4 18.7 13.0 7.0 2.1 11.6 277
1884 -2.3 1.2 5.2 11.1 17.1 21.4 23.2 22.7 20.4 14.9 6.9 0.5 11.9 316
1885 -1.7 -1.2 4.1 11.6 16.8 21.2 24.2 22.5 19.1 12.5 7.1 2.3 11.5 335
1886 -3.0 1.3 4.9 12.7 18.0 21.3 24.1 23.5 19.8 13.9 5.4 0.0 11.8 356
1887 -1.1 1.6 5.9 11.4 18.9 21.9 24.8 22.4 19.0 12.5 6.5 1.0 12.1 386
1888 -3.1 1.9 3.6 13.0 16.3 21.7 24.1 22.8 18.7 12.4 7.0 3.1 11.8 441
1889 1.1 0.0 7.4 12.6 17.0 20.8 23.6 22.4 18.4 12.2 6.3 6.0 12.3 511
1890 1.2 3.2 4.5 12.0 16.5 22.2 24.0 22.0 18.4 12.7 8.0 2.2 12.2 536
1891 1.2 1.2 3.4 12.1 16.0 21.0 22.1 22.5 20.0 12.6 5.9 3.7 11.8 595
1892 -1.4 2.8 4.5 10.7 15.5 21.3 23.2 22.8 19.1 13.3 6.1 0.1 11.5 662
1893 -2.9 0.0 4.4 10.8 15.7 21.4 23.8 22.3 18.7 12.8 5.7 1.9 11.2 728
1894 0.3 -0.5 7.4 12.1 17.0 21.4 23.7 22.9 19.4 13.4 6.0 2.4 12.1 765
1895 -1.9 -2.3 5.2 12.5 16.8 21.4 22.5 22.8 20.1 11.3 5.8 1.5 11.3 822
1896 0.5 2.3 3.9 12.8 18.5 21.5 23.8 23.1 17.9 11.9 5.5 2.7 12.0 851
1897 -1.2 1.7 5.2 11.5 16.5 20.8 23.9 22.2 20.2 14.2 6.1 0.5 11.8 886
1898 0.8 2.1 6.6 10.7 16.4 21.6 23.5 23.2 19.7 12.0 4.9 -0.4 11.8 909
1899 -0.3 -3.1 3.6 11.3 16.6 21.3 23.2 22.9 18.5 13.7 8.3 0.8 11.4 927
1900 1.4 -0.7 4.8 11.7 17.2 21.4 23.3 23.9 19.6 15.0 6.4 2.1 12.2 952
1901 0.5 -1.1 5.5 10.3 16.4 21.3 25.3 23.3 18.4 13.8 5.9 0.3 11.7 966
1902 -0.2 -0.3 6.7 11.1 17.6 20.4 23.0 22.1 17.8 13.5 8.3 0.0 11.7 983
1903 0.0 -0.4 7.2 10.9 16.4 19.2 22.7 21.9 18.0 13.1 5.4 -0.3 11.2 1016
1904 -2.1 -0.6 5.8 9.9 16.4 20.3 22.2 21.7 19.0 13.0 7.0 0.7 11.1 1054
1905 -2.3 -2.4 8.1 11.2 16.2 20.8 22.6 22.8 19.4 12.0 6.7 1.1 11.3 1067
1906 1.7 1.3 2.9 12.4 16.2 20.5 22.7 22.9 19.9 12.3 6.0 2.1 11.7 1097
1907 0.2 1.3 8.1 8.7 13.9 19.3 23.0 22.0 18.5 12.4 6.0 2.3 11.3 1127
1908 0.9 0.8 7.2 12.1 15.9 20.1 23.2 21.9 19.3 12.1 7.0 1.8 11.9 1150
1909 0.5 2.3 5.1 10.1 15.1 20.9 22.6 23.0 18.3 12.0 8.3 -2.4 11.3 1179
1910 -0.6 -0.9 10.2 12.3 15.2 20.2 23.4 21.9 19.0 13.8 5.4 0.1 11.7 1191
1911 0.8 1.6 6.7 10.4 17.2 21.9 23.0 22.0 19.4 12.3 3.9 1.5 11.7 1222
1912 -3.9 -0.3 2.7 11.2 16.6 19.6 22.7 21.4 17.9 12.7 6.6 1.8 10.8 1234
1913 0.4 -0.6 4.7 11.4 16.1 20.7 23.2 23.3 18.0 11.7 8.1 2.0 11.6 1252
1914 1.9 -1.0 5.2 11.0 16.8 21.3 23.4 22.3 18.3 13.7 7.1 -1.8 11.5 1264
1915 -1.0 2.8 3.4 13.3 15.0 19.3 21.9 21.2 18.7 13.6 7.1 1.1 11.4 1276
1916 -1.0 0.5 5.3 10.5 15.7 19.1 23.7 22.5 17.8 12.0 5.6 -0.8 10.9 1297
1917 -1.1 -0.8 4.5 9.8 13.2 19.5 23.3 21.6 17.9 10.3 6.6 -1.6 10.3 1313
1918 -4.3 1.1 7.8 9.8 16.7 21.4 22.3 22.9 16.6 13.9 5.7 2.5 11.4 1326
1919 0.7 0.8 5.7 11.0 15.6 21.0 23.5 22.2 19.0 12.4 4.9 -1.3 11.3 1334
1920 -1.4 1.0 5.2 8.6 15.2 20.0 22.4 21.5 18.8 13.3 5.1 1.4 10.9 1342
1921 1.6 2.9 8.6 11.3 16.0 21.6 23.8 22.2 19.7 13.1 6.3 2.0 12.4 1350
1922 -1.9 0.6 5.3 10.8 16.6 21.3 22.5 22.5 19.7 13.3 6.3 1.2 11.5 1353
1923 1.5 -0.9 4.0 10.3 15.4 20.5 23.1 21.9 18.7 11.4 6.6 3.2 11.3 1360
1924 -2.4 1.6 3.6 10.4 14.4 20.3 21.9 22.2 17.1 13.2 6.4 -1.7 10.6 1360
1925 -1.2 3.5 7.0 12.8 15.4 21.3 23.2 22.1 19.8 9.7 5.4 0.7 11.6 1367
1926 -0.1 3.0 4.5 10.2 16.4 20.0 23.0 22.6 18.2 13.0 5.4 0.1 11.4 1370
1927 0.0 3.6 6.2 10.9 15.5 19.5 22.5 20.6 18.7 13.6 7.2 -1.0 11.4 1376
1928 0.5 1.7 6.2 9.3 16.3 18.9 22.9 22.2 17.3 13.2 6.2 1.4 11.3 1383
1929 -2.4 -2.0 6.9 11.1 15.1 19.8 22.9 22.3 17.9 12.4 4.5 1.3 10.8 1386
1930 -3.5 4.3 4.9 12.2 15.6 20.4 23.8 22.7 19.1 11.4 5.9 0.5 11.4 1391
1931 1.2 3.5 4.4 10.9 15.4 21.6 24.2 22.2 20.3 13.9 7.5 2.8 12.3 1398
1932 0.9 2.8 3.3 11.1 16.1 20.8 23.2 22.6 18.4 12.0 5.2 -0.2 11.3 1404
1933 2.1 -0.4 5.5 10.4 16.0 22.1 23.6 22.1 19.9 12.7 6.1 2.3 11.9 1408
1934 1.9 0.7 5.8 12.1 18.1 21.8 24.6 22.9 18.1 13.8 7.9 0.9 12.4 1406
1935 0.0 2.4 7.1 10.1 14.4 19.9 24.1 22.8 18.4 12.5 5.1 -0.4 11.4 1407
1936 -2.2 -3.5 6.7 10.2 17.7 21.4 24.8 23.8 19.4 12.5 5.1 2.0 11.5 1412
1937 -1.9 0.3 4.0 10.3 16.7 20.6 23.5 23.7 18.6 12.3 5.5 0.7 11.2 1414
1938 0.2 2.6 7.7 11.5 15.7 20.4 23.2 23.3 19.2 14.1 5.6 1.7 12.1 1415
1939 1.5 0.0 5.9 10.8 17.1 20.7 23.5 22.5 19.8 13.0 6.1 3.5 12.0 1416
1940 -4.1 1.3 5.2 10.4 16.0 20.9 23.3 22.2 18.6 13.6 4.8 2.8 11.2 1415
1941 0.4 0.9 4.2 12.0 17.1 20.3 23.3 22.3 18.6 13.4 6.6 2.8 11.8 1423
1942 -0.4 -0.1 5.7 12.1 15.6 20.2 23.2 22.0 17.9 12.8 6.3 0.0 11.3 1430
1943 -1.4 2.2 3.8 11.1 15.5 20.9 23.2 22.8 17.7 12.2 5.4 0.7 11.2 1427
1944 0.6 1.7 3.8 9.5 17.0 20.5 22.4 22.1 18.4 12.9 5.8 -0.4 11.2 1435
1945 -0.8 1.8 8.1 10.6 14.4 18.9 22.4 22.0 18.5 12.3 5.8 -1.2 11.1 1482
1946 0.2 1.7 8.4 12.2 14.9 20.1 22.8 21.2 18.1 12.5 6.1 2.2 11.7 1485
1947 0.1 -0.3 3.8 10.7 15.4 19.3 22.2 23.3 19.0 14.9 4.3 1.0 11.1 1507
1948 -2.0 0.0 4.3 11.9 15.6 20.3 22.6 21.9 18.7 11.7 6.1 0.8 11.0 1624
1949 -1.4 0.7 5.5 11.0 16.5 20.8 23.3 22.2 17.6 13.1 7.7 1.5 11.5 1754
1950 0.4 1.7 4.2 9.3 15.5 20.1 21.6 21.2 17.6 14.2 5.2 0.7 11.0 1763
1951 -0.4 1.6 4.0 10.2 16.2 19.5 22.8 22.0 18.0 12.7 4.2 0.5 10.9 1791
1952 0.5 2.4 4.0 11.0 15.8 21.6 23.3 22.4 18.8 11.8 5.7 1.6 11.6 1804
1953 2.3 2.7 6.6 9.8 15.8 21.3 23.2 22.2 19.0 13.7 7.1 1.9 12.1 1818
1954 -0.2 4.7 4.6 12.4 14.8 20.7 23.7 22.3 19.3 13.1 7.2 1.4 12.0 1828
1955 -0.4 0.3 5.1 11.8 16.5 19.2 23.6 23.1 18.9 12.9 4.1 0.1 11.3 1755
1956 -0.2 0.8 4.9 9.9 16.4 21.0 22.5 22.1 18.2 13.5 5.4 2.8 11.4 1757
1957 -1.7 3.5 5.6 11.1 15.8 20.8 23.2 22.0 18.2 11.5 5.9 3.0 11.6 1765
1958 0.3 0.1 4.1 10.9 16.7 19.9 22.5 22.6 18.6 12.8 6.8 0.2 11.3 1769
1959 -1.0 1.1 5.2 11.1 16.5 21.0 22.8 22.9 18.5 12.3 4.6 2.6 11.5 1767
1960 -0.3 0.3 2.1 11.6 15.5 20.4 22.8 22.2 19.0 12.9 6.5 0.0 11.1 1763
1961 -0.6 3.0 6.3 9.4 15.0 20.5 22.6 22.3 18.2 12.5 5.6 0.0 11.2 1761
1962 -1.8 1.9 3.9 11.0 17.0 20.0 22.0 22.1 17.6 13.5 6.4 0.9 11.2 1801
1963 -3.0 0.8 6.7 11.3 16.1 20.5 22.8 21.9 18.9 15.2 7.0 -1.5 11.4 1850
1964 0.4 0.5 4.4 10.9 16.4 20.2 23.3 21.3 17.9 12.2 6.6 0.5 11.2 1841
1965 0.0 0.4 3.1 11.2 16.5 19.5 22.3 21.7 17.3 12.7 7.3 2.6 11.2 1835
1966 -2.4 0.3 6.1 10.2 15.4 20.2 23.6 21.5 18.1 11.9 6.7 1.0 11.1 1830
1967 1.1 0.5 6.5 11.1 14.4 20.1 22.2 21.6 17.8 12.4 5.7 1.3 11.2 1823
1968 -1.2 0.4 6.8 10.9 14.8 20.3 22.7 21.8 18.1 12.8 5.8 -0.4 11.1 1822
1969 -1.2 0.8 3.1 11.7 16.5 19.8 23.1 22.5 18.7 11.5 5.8 1.3 11.1 1813
1970 -2.6 2.0 4.3 10.5 16.5 20.5 23.0 22.7 18.6 12.1 6.0 1.4 11.2 1797
1971 -1.7 0.9 4.0 10.2 14.7 20.9 22.1 22.0 18.5 13.6 5.8 2.0 11.1 1693
1972 -0.8 0.6 5.9 10.2 16.0 19.8 22.3 22.0 18.2 11.4 4.7 -0.1 10.8 1689
1973 -0.8 1.0 7.3 10.1 15.1 20.7 22.7 22.4 18.3 13.5 6.4 1.3 11.5 1685
1974 0.0 1.3 6.7 11.2 15.7 19.8 23.1 21.4 17.1 12.1 6.2 1.4 11.3 1679
1975 0.4 0.7 3.9 8.7 16.4 19.9 22.8 22.1 17.2 12.9 6.5 1.3 11.1 1670
1976 -1.1 3.8 6.2 11.3 15.0 20.1 22.5 21.5 17.9 10.4 4.1 -0.4 10.9 1669
1977 -4.3 2.3 6.7 12.3 16.9 20.9 23.5 22.1 18.9 12.2 6.3 0.6 11.5 1660
1978 -2.8 -2.0 5.0 11.0 15.7 20.5 22.9 22.2 19.1 12.4 6.0 0.0 10.8 1660
1979 -4.3 -2.6 6.0 10.4 15.5 20.0 22.6 21.7 18.9 12.9 5.7 2.4 10.8 1657
1980 -0.3 0.4 4.4 11.0 16.1 20.2 23.9 22.7 19.0 11.6 6.1 1.1 11.4 1650
1981 0.0 2.7 6.2 12.9 15.3 21.1 23.0 22.0 18.1 11.4 7.1 0.8 11.7 1623
1982 -3.3 0.4 5.7 9.4 16.5 19.2 22.8 21.9 17.9 12.1 5.9 2.9 10.9 1605
1983 0.5 2.7 6.1 9.0 14.8 19.9 23.4 23.8 18.7 12.8 6.6 -2.8 11.3 1594
1984 -1.5 3.1 4.6 10.2 15.6 20.7 22.6 22.8 17.6 13.1 5.7 2.5 11.4 1592
1985 -2.5 -0.2 6.7 12.4 16.8 19.9 23.1 21.7 17.7 12.8 5.4 -1.1 11.1 1594
1986 1.2 2.1 7.5 11.7 16.6 21.3 23.2 21.7 18.3 12.7 5.7 2.0 12.0 1590
1987 0.0 2.9 6.4 11.9 17.6 21.4 23.2 22.3 18.6 11.5 7.2 2.1 12.1 1589
1988 -1.8 0.9 6.0 11.2 16.6 21.4 23.8 23.3 18.4 11.6 6.8 1.4 11.6 1598
1989 1.5 -0.8 5.9 11.6 16.0 20.5 23.3 22.1 18.1 12.9 6.2 -2.1 11.3 1597
1990 3.0 2.8 7.4 11.7 15.5 21.4 23.1 22.7 19.7 12.8 7.7 0.7 12.4 1572
1991 -0.6 4.4 7.1 12.4 17.9 21.4 23.7 23.1 18.8 13.3 5.2 2.8 12.5 1549
1992 1.5 4.4 7.1 11.7 16.5 20.2 22.4 21.3 18.5 12.7 5.7 1.0 11.9 1536
1993 0.1 0.0 5.6 10.7 16.8 20.5 23.3 22.9 18.0 12.3 5.3 1.9 11.4 1529
1994 -1.5 0.3 7.1 12.3 16.4 22.1 23.4 22.4 19.0 13.0 7.2 3.1 12.1 1519
1995 1.0 2.6 6.9 10.6 15.8 20.7 23.7 24.0 18.6 13.3 5.7 1.1 12.0 1495
1996 -0.9 1.9 4.2 10.9 16.6 21.4 23.1 22.6 18.1 12.7 4.6 1.8 11.4 1464
1997 -0.6 3.0 7.3 9.7 15.5 20.7 23.3 22.3 19.4 12.8 5.5 1.7 11.7 1431
1998 2.1 4.4 5.8 11.4 17.9 20.8 24.2 23.5 20.9 13.5 7.8 2.8 12.9 1428
1999 0.8 4.1 5.9 11.7 16.4 20.8 24.1 23.0 18.4 12.8 9.2 2.6 12.5 1447
2000 0.6 4.3 8.2 11.5 17.7 21.0 23.2 23.3 18.9 13.4 4.2 -2.1 12.0 1429
2001 -0.2 1.3 5.2 12.3 17.4 21.0 23.6 23.6 18.7 12.7 9.2 3.1 12.3 1434
2002 2.1 2.8 4.8 12.4 15.6 21.8 24.5 23.1 20.0 11.7 6.0 2.1 12.2 1421
2003 -0.1 0.5 6.6 11.6 16.4 20.4 24.0 23.8 18.5 13.5 6.7 1.9 12.0 1411
2004 -1.4 0.9 8.2 11.8 17.4 20.5 22.9 21.5 19.3 13.4 7.3 1.7 12.0 1382
2005 0.3 3.2 5.5 11.8 15.6 21.4 24.2 23.4 20.1 13.4 7.5 0.2 12.2 1214
2006 4.1 1.4 6.1 13.5 17.3 21.7 24.6 23.4 18.3 12.6 7.7 4.1 12.9 1177
2007 1.0 0.5 8.5 11.1 17.7 21.6 23.6 24.2 20.2 15.0 7.5 2.4 12.8 134
2008 0.3 2.1 6.2 11.5 16.2 21.6 23.5 22.6 19.3 12.7 6.8 1.7 12.0 136
-0.4 1.2 5.6 11.0 16.1 20.6 23.1 22.3 18.6 12.7 6.1 1.0 11.5
-1.1 0.4 4.5 10.3 15.6 20.4 23.0 22.2 18.1 12.1 5.8 0.7
[chiefio@tubularbells analysis]$
Notice the “order of magnitude” fall off in number of thermometers from 2006 to 2007.
The last two lines are the monthly averages for all the data calculated two different ways to test sensitivity to the order of calculation (just days data, or months averages then average that average; basically does a month/year with one thermometer get the same weight as a month/year with 1000 thermometers). This is an example of why I say the “Global Average Temperature” has little meaning in the 1/10 C place. It depends as much on programmer decisions as it does on thermometers.
It is remotely possible that I just got a bum copy of the data and it has been fixed by now (But I’ve heard nothing in the news; and their web sites seems to confirm this data).
It is relatively easy to test. Download the v2.mean file. Search for those records starting with “425” (the USA “country code”). Then take those records and split out the ones with “2008” for the year. Count them. That’s it.
Geek Sidebar
In Linux tools that would be:
grep ^425 v2.mean > USA.mean grep "2008 " USA.mean > USA.mean.2008 grep "2009-" USA.mean >> USA.mean.2008 wc -l USA.mean.2008
(In fact there is a slightly more elegant way to do the second “grep” that I won’t go into here. The grep for “2008-” is for those cases where there is data missing for a January since that would be a -9999 and the “-” would butt up against the 2008 as “2008-“. You grep for “2008-” to be sure that case is counted.)
Or you can just pull the whole thing into an editor and go looking for the 2008 records.
You can download a copy of the GHCN files from:
ftp://ftp.ncdc.noaa.gov/pub/data/ghcn/v2
And the USHCN files are here:
ftp://ftp.ncdc.noaa.gov/pub/data/ushcn
The “by Latitude” set for the USA
The format is year, then 9 latitude bands of 5 degrees each from “less than 30” to 30-35, etc up to “over 65N”.
[chiefio@tubularbells analysis]$ cat USA.bylat LAT pct: 1743 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1744 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1745 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1746 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1747 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1748 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1749 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 DecLatPct: 1749 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1750 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1751 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1752 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1753 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1754 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1755 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1756 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1757 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1758 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1759 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 DecLatPct: 1759 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1760 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1761 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1762 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1763 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1764 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1765 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1766 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1767 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1768 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1769 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 DecLatPct: 1769 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1770 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1771 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1772 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1773 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1774 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1775 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1776 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1777 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1778 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1779 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 DecLatPct: 1779 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1781 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1782 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1783 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1784 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1785 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1786 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1787 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1788 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1789 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 DecLatPct: 1789 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1790 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1791 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1792 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1793 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1794 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1795 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1796 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1797 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1798 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1799 0.0 25.0 0.0 75.0 0.0 0.0 0.0 0.0 0.0 DecLatPct: 1799 0.0 4.2 0.0 95.8 0.0 0.0 0.0 0.0 0.0 LAT pct: 1800 0.0 25.0 0.0 75.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1801 0.0 25.0 0.0 75.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1802 0.0 25.0 0.0 75.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1803 0.0 25.0 0.0 75.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1804 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1805 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1806 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1807 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1808 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1809 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 DecLatPct: 1809 0.0 13.8 0.0 86.2 0.0 0.0 0.0 0.0 0.0 LAT pct: 1810 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1811 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1812 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1813 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1814 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1815 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1816 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1817 0.0 0.0 25.0 75.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1818 0.0 0.0 33.3 66.7 0.0 0.0 0.0 0.0 0.0 LAT pct: 1819 0.0 0.0 25.0 75.0 0.0 0.0 0.0 0.0 0.0 DecLatPct: 1819 0.0 0.0 10.0 90.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1820 0.0 0.0 33.3 66.7 0.0 0.0 0.0 0.0 0.0 LAT pct: 1821 0.0 0.0 30.0 70.0 0.0 0.0 0.0 0.0 0.0 LAT pct: 1822 0.0 15.4 15.4 69.2 0.0 0.0 0.0 0.0 0.0 LAT pct: 1823 0.0 14.3 21.4 64.3 0.0 0.0 0.0 0.0 0.0 LAT pct: 1824 0.0 17.6 17.6 64.7 0.0 0.0 0.0 0.0 0.0 LAT pct: 1825 5.6 16.7 11.1 66.7 0.0 0.0 0.0 0.0 0.0 LAT pct: 1826 5.9 11.8 11.8 70.6 0.0 0.0 0.0 0.0 0.0 LAT pct: 1827 4.8 9.5 14.3 71.4 0.0 0.0 0.0 0.0 0.0 LAT pct: 1828 4.2 12.5 8.3 70.8 0.0 0.0 4.2 0.0 0.0 LAT pct: 1829 3.4 10.3 10.3 69.0 3.4 0.0 3.4 0.0 0.0 DecLatPct: 1829 2.9 11.6 15.1 68.6 0.6 0.0 1.2 0.0 0.0 LAT pct: 1830 6.7 10.0 13.3 66.7 3.3 0.0 0.0 0.0 0.0 LAT pct: 1831 6.5 9.7 12.9 67.7 3.2 0.0 0.0 0.0 0.0 LAT pct: 1832 8.1 10.8 10.8 62.2 2.7 0.0 5.4 0.0 0.0 LAT pct: 1833 5.7 14.3 8.6 62.9 2.9 0.0 5.7 0.0 0.0 LAT pct: 1834 5.4 13.5 10.8 62.2 2.7 0.0 5.4 0.0 0.0 LAT pct: 1835 5.7 11.4 8.6 65.7 2.9 0.0 5.7 0.0 0.0 LAT pct: 1836 5.1 12.8 15.4 59.0 2.6 0.0 5.1 0.0 0.0 LAT pct: 1837 7.1 9.5 16.7 59.5 2.4 0.0 4.8 0.0 0.0 LAT pct: 1838 5.1 10.3 17.9 59.0 2.6 0.0 5.1 0.0 0.0 LAT pct: 1839 2.5 10.0 15.0 62.5 2.5 0.0 7.5 0.0 0.0 DecLatPct: 1839 5.8 11.2 13.2 62.5 2.7 0.0 4.7 0.0 0.0 LAT pct: 1840 7.0 9.3 16.3 58.1 2.3 0.0 7.0 0.0 0.0 LAT pct: 1841 6.8 6.8 15.9 61.4 2.3 0.0 6.8 0.0 0.0 LAT pct: 1842 6.7 6.7 15.6 62.2 2.2 0.0 6.7 0.0 0.0 LAT pct: 1843 6.1 10.2 16.3 59.2 2.0 0.0 6.1 0.0 0.0 LAT pct: 1844 3.9 9.8 17.6 60.8 2.0 0.0 5.9 0.0 0.0 LAT pct: 1845 4.2 10.4 16.7 62.5 2.1 0.0 4.2 0.0 0.0 LAT pct: 1846 4.7 9.3 23.3 62.8 0.0 0.0 0.0 0.0 0.0 LAT pct: 1847 2.3 9.1 25.0 59.1 0.0 0.0 4.5 0.0 0.0 LAT pct: 1848 2.3 7.0 25.6 60.5 0.0 0.0 4.7 0.0 0.0 LAT pct: 1849 10.2 12.2 20.4 51.0 2.0 0.0 4.1 0.0 0.0 DecLatPct: 1849 5.4 9.2 19.2 59.7 1.5 0.0 5.0 0.0 0.0 LAT pct: 1850 9.8 15.7 17.6 51.0 2.0 0.0 3.9 0.0 0.0 LAT pct: 1851 10.5 15.8 15.8 52.6 1.8 0.0 3.5 0.0 0.0 LAT pct: 1852 9.5 12.7 17.5 55.6 1.6 0.0 3.2 0.0 0.0 LAT pct: 1853 9.4 15.1 22.6 47.2 1.9 0.0 3.8 0.0 0.0 LAT pct: 1854 9.0 16.4 23.9 46.3 1.5 0.0 3.0 0.0 0.0 LAT pct: 1855 8.5 16.9 25.4 46.5 1.4 0.0 1.4 0.0 0.0 LAT pct: 1856 7.3 15.9 22.0 51.2 1.2 0.0 2.4 0.0 0.0 LAT pct: 1857 7.8 15.6 24.4 47.8 2.2 0.0 2.2 0.0 0.0 LAT pct: 1858 7.9 14.6 29.2 43.8 2.2 0.0 2.2 0.0 0.0 LAT pct: 1859 7.1 16.5 30.6 38.8 4.7 0.0 2.4 0.0 0.0 DecLatPct: 1859 8.5 15.5 23.6 47.6 2.1 0.0 2.7 0.0 0.0 LAT pct: 1860 6.2 18.5 26.2 41.5 4.6 0.0 3.1 0.0 0.0 LAT pct: 1861 5.1 14.1 32.1 42.3 3.8 0.0 2.6 0.0 0.0 LAT pct: 1862 1.4 13.9 33.3 45.8 2.8 0.0 2.8 0.0 0.0 LAT pct: 1863 1.3 10.7 33.3 49.3 2.7 0.0 2.7 0.0 0.0 LAT pct: 1864 1.2 11.6 29.1 53.5 2.3 0.0 2.3 0.0 0.0 LAT pct: 1865 1.2 14.3 29.8 51.2 1.2 0.0 2.4 0.0 0.0 LAT pct: 1866 0.0 14.3 28.6 52.7 2.2 0.0 2.2 0.0 0.0 LAT pct: 1867 0.0 14.3 27.6 51.0 4.1 0.0 3.1 0.0 0.0 LAT pct: 1868 0.0 15.1 28.3 48.1 4.7 0.0 3.8 0.0 0.0 LAT pct: 1869 0.9 14.0 28.1 47.4 3.5 0.0 6.1 0.0 0.0 DecLatPct: 1869 1.5 14.0 29.5 48.6 3.2 0.0 3.2 0.0 0.0 LAT pct: 1870 2.9 11.5 26.6 51.1 2.9 0.0 5.0 0.0 0.0 LAT pct: 1871 2.5 13.0 31.5 47.5 3.1 0.0 2.5 0.0 0.0 LAT pct: 1872 1.7 15.0 32.2 45.0 2.8 0.6 2.8 0.0 0.0 LAT pct: 1873 3.0 13.9 31.8 42.3 6.5 0.0 2.5 0.0 0.0 LAT pct: 1874 2.5 13.8 32.5 42.4 6.9 0.0 2.0 0.0 0.0 LAT pct: 1875 2.4 13.3 31.9 42.4 7.6 0.0 2.4 0.0 0.0 LAT pct: 1876 2.4 13.3 32.7 41.7 7.6 0.0 2.4 0.0 0.0 LAT pct: 1877 2.8 15.7 31.0 41.2 7.4 0.0 1.9 0.0 0.0 LAT pct: 1878 2.9 17.2 31.1 41.2 6.7 0.4 0.4 0.0 0.0 LAT pct: 1879 2.8 17.0 30.4 42.1 6.9 0.4 0.4 0.0 0.0 DecLatPct: 1879 2.6 14.6 31.3 43.2 6.1 0.1 2.0 0.0 0.0 LAT pct: 1880 2.8 17.1 30.3 41.0 8.4 0.4 0.0 0.0 0.0 LAT pct: 1881 2.7 17.0 28.4 40.5 8.3 0.4 2.7 0.0 0.0 LAT pct: 1882 2.9 23.5 26.0 37.6 7.4 0.3 1.9 0.3 0.0 LAT pct: 1883 2.7 22.9 27.7 36.1 7.8 0.3 2.1 0.3 0.0 LAT pct: 1884 2.2 25.1 28.0 34.5 8.1 0.3 1.6 0.3 0.0 LAT pct: 1885 2.3 24.4 28.0 35.5 8.2 0.3 1.0 0.3 0.0 LAT pct: 1886 2.2 23.2 29.5 35.7 8.0 0.2 1.0 0.2 0.0 LAT pct: 1887 2.5 22.0 29.7 36.9 8.1 0.0 0.9 0.0 0.0 LAT pct: 1888 2.8 22.5 28.9 37.3 8.2 0.0 0.2 0.0 0.0 LAT pct: 1889 2.8 22.2 28.9 36.6 9.2 0.0 0.4 0.0 0.0 DecLatPct: 1889 2.6 22.3 28.6 36.9 8.2 0.2 1.1 0.1 0.0 LAT pct: 1890 3.0 20.9 27.8 37.4 10.1 0.0 0.7 0.0 0.0 LAT pct: 1891 3.1 20.1 27.7 37.7 11.0 0.0 0.5 0.0 0.0 LAT pct: 1892 3.3 20.0 28.6 36.2 11.5 0.0 0.4 0.0 0.0 LAT pct: 1893 3.0 19.9 29.2 35.8 11.7 0.0 0.3 0.1 0.0 LAT pct: 1894 2.8 19.2 29.2 35.5 12.7 0.0 0.5 0.1 0.0 LAT pct: 1895 3.1 18.1 29.7 35.7 12.8 0.0 0.6 0.1 0.0 LAT pct: 1896 3.1 17.9 29.9 35.6 13.0 0.0 0.4 0.1 0.0 LAT pct: 1897 2.8 17.8 29.3 37.1 12.5 0.0 0.2 0.1 0.2 LAT pct: 1898 2.8 17.1 29.7 36.5 12.9 0.0 0.6 0.1 0.2 LAT pct: 1899 2.9 16.7 29.8 36.0 12.8 0.0 1.1 0.4 0.2 DecLatPct: 1899 3.0 18.6 29.2 36.3 12.2 0.0 0.5 0.1 0.1 LAT pct: 1900 2.9 16.8 30.2 35.9 12.3 0.0 1.3 0.5 0.2 LAT pct: 1901 3.0 17.4 30.8 35.1 11.9 0.0 1.2 0.5 0.3 LAT pct: 1902 2.9 17.5 30.8 35.1 12.0 0.0 1.1 0.3 0.3 LAT pct: 1903 2.9 17.5 30.5 34.8 12.5 0.0 1.1 0.3 0.4 LAT pct: 1904 2.8 17.3 31.2 34.4 12.5 0.0 1.2 0.2 0.4 LAT pct: 1905 3.1 17.2 30.4 34.4 13.2 0.1 1.0 0.4 0.1 LAT pct: 1906 3.1 17.2 31.1 33.9 13.1 0.2 0.9 0.5 0.1 LAT pct: 1907 3.1 17.1 30.6 33.7 13.6 0.2 1.0 0.7 0.2 LAT pct: 1908 3.2 17.4 30.3 33.6 13.4 0.2 1.1 0.7 0.2 LAT pct: 1909 3.1 17.2 29.7 33.4 14.2 0.2 1.1 1.0 0.2 DecLatPct: 1909 3.0 17.3 30.6 34.4 12.9 0.1 1.1 0.5 0.2 LAT pct: 1910 3.1 17.0 29.7 33.3 14.2 0.2 1.1 1.3 0.2 LAT pct: 1911 3.1 16.8 30.2 33.1 14.2 0.2 1.1 1.0 0.2 LAT pct: 1912 3.2 16.9 30.6 32.7 14.2 0.2 1.0 1.1 0.2 LAT pct: 1913 3.1 17.1 30.5 32.6 14.3 0.1 1.0 1.0 0.1 LAT pct: 1914 3.1 17.0 30.1 33.0 14.4 0.1 1.0 1.1 0.1 LAT pct: 1915 3.1 17.1 29.9 33.0 14.3 0.1 1.2 1.0 0.2 LAT pct: 1916 3.1 16.9 30.0 32.7 14.6 0.1 1.3 1.1 0.2 LAT pct: 1917 3.1 16.6 29.8 32.3 14.6 0.2 1.7 1.3 0.4 LAT pct: 1918 3.0 16.8 29.5 32.2 14.7 0.2 1.7 1.5 0.4 LAT pct: 1919 3.0 16.8 29.6 32.4 14.6 0.2 1.7 1.5 0.3 DecLatPct: 1919 3.1 16.9 30.0 32.7 14.4 0.2 1.3 1.2 0.2 LAT pct: 1920 3.0 16.8 29.3 32.4 14.7 0.2 1.7 1.5 0.3 LAT pct: 1921 3.1 16.8 29.7 32.0 14.6 0.2 1.7 1.3 0.5 LAT pct: 1922 3.1 16.9 29.5 32.2 14.6 0.1 1.7 1.3 0.5 LAT pct: 1923 3.1 17.0 29.6 32.1 14.6 0.1 1.6 1.4 0.5 LAT pct: 1924 3.1 17.0 29.5 32.1 14.6 0.1 1.5 1.6 0.5 LAT pct: 1925 3.1 16.8 29.5 32.1 14.6 0.1 1.7 1.6 0.5 LAT pct: 1926 3.2 16.9 29.5 32.0 14.7 0.1 1.7 1.5 0.5 LAT pct: 1927 3.2 16.9 29.4 32.2 14.7 0.1 1.6 1.5 0.5 LAT pct: 1928 3.2 16.9 29.1 32.2 14.6 0.1 1.7 1.5 0.7 LAT pct: 1929 3.2 16.9 28.9 32.2 14.7 0.1 1.7 1.7 0.7 DecLatPct: 1929 3.1 16.9 29.4 32.1 14.6 0.2 1.7 1.5 0.5 LAT pct: 1930 3.3 16.9 28.8 32.2 14.6 0.1 1.7 1.7 0.7 LAT pct: 1931 3.2 16.7 28.9 32.4 14.5 0.1 1.7 1.8 0.7 LAT pct: 1932 3.2 16.7 28.8 32.3 14.5 0.1 1.8 1.8 0.7 LAT pct: 1933 3.2 16.7 28.8 32.4 14.5 0.1 1.8 1.8 0.7 LAT pct: 1934 3.2 16.9 28.9 32.5 14.5 0.1 1.7 1.6 0.6 LAT pct: 1935 3.2 16.9 29.0 32.4 14.4 0.1 1.7 1.6 0.7 LAT pct: 1936 3.2 16.8 28.9 32.3 14.4 0.1 1.9 1.6 0.7 LAT pct: 1937 3.2 16.7 29.1 32.3 14.4 0.1 1.9 1.6 0.7 LAT pct: 1938 3.2 16.7 29.1 32.3 14.4 0.1 1.8 1.6 0.7 LAT pct: 1939 3.2 16.6 28.9 32.4 14.4 0.2 2.0 1.6 0.7 DecLatPct: 1939 3.2 16.8 28.9 32.4 14.4 0.1 1.8 1.6 0.7 LAT pct: 1940 3.2 16.9 29.0 32.3 14.4 0.1 1.9 1.6 0.7 LAT pct: 1941 3.1 16.6 28.7 32.0 14.2 0.2 2.1 2.3 0.7 LAT pct: 1942 3.1 16.4 28.4 31.9 14.0 0.3 2.4 2.8 0.7 LAT pct: 1943 3.2 16.3 28.5 31.9 14.0 0.3 2.4 2.8 0.7 LAT pct: 1944 3.1 16.3 28.4 32.0 13.9 0.3 2.4 2.9 0.8 LAT pct: 1945 3.4 17.0 28.3 31.6 13.6 0.3 2.4 2.7 0.8 LAT pct: 1946 3.6 17.0 28.1 31.5 13.6 0.5 2.2 2.8 0.8 LAT pct: 1947 3.5 17.2 27.9 31.3 13.7 0.5 2.3 2.7 0.9 LAT pct: 1948 3.4 17.9 28.2 30.6 13.4 0.4 2.3 2.9 0.9 LAT pct: 1949 3.8 18.2 28.2 30.4 13.2 0.4 2.2 2.8 0.9 DecLatPct: 1949 3.4 17.0 28.4 31.5 13.8 0.3 2.2 2.6 0.8 LAT pct: 1950 4.0 18.4 28.0 30.2 13.2 0.4 2.1 2.8 0.9 LAT pct: 1951 4.2 18.3 27.7 30.2 13.0 0.4 2.1 3.1 0.9 LAT pct: 1952 4.4 18.3 27.9 30.1 12.9 0.3 2.1 3.0 0.9 LAT pct: 1953 4.4 18.4 27.9 30.0 13.0 0.3 2.1 3.1 0.9 LAT pct: 1954 4.5 18.3 28.0 29.9 12.9 0.4 2.1 2.9 0.9 LAT pct: 1955 4.5 18.4 27.7 30.0 12.9 0.3 2.2 3.1 0.9 LAT pct: 1956 4.6 18.5 27.6 30.0 12.8 0.3 2.2 3.1 0.9 LAT pct: 1957 4.6 18.4 27.7 29.9 12.8 0.3 2.2 3.1 0.9 LAT pct: 1958 4.5 18.5 27.7 29.9 12.8 0.3 2.2 3.1 0.9 LAT pct: 1959 4.5 18.4 27.6 30.0 12.9 0.2 2.3 3.1 0.9 DecLatPct: 1959 4.4 18.4 27.8 30.0 12.9 0.3 2.2 3.1 0.9 LAT pct: 1960 4.5 18.4 27.6 30.0 12.9 0.2 2.3 3.1 0.9 LAT pct: 1961 4.7 18.5 27.3 29.8 12.8 0.3 2.5 3.1 1.0 LAT pct: 1962 4.8 18.4 27.8 29.7 12.6 0.3 2.4 3.0 1.0 LAT pct: 1963 4.9 18.5 28.0 29.3 12.8 0.3 2.4 2.9 0.9 LAT pct: 1964 4.7 18.4 28.2 29.5 12.8 0.3 2.4 2.9 0.8 LAT pct: 1965 4.7 18.4 28.2 29.6 12.8 0.3 2.3 2.8 0.9 LAT pct: 1966 4.7 18.4 28.1 29.7 12.9 0.3 2.3 2.7 0.9 LAT pct: 1967 4.7 18.3 28.1 29.7 12.8 0.3 2.3 2.8 0.9 LAT pct: 1968 4.8 18.4 28.1 29.7 12.8 0.3 2.3 2.7 0.9 LAT pct: 1969 4.8 18.3 28.1 29.7 12.8 0.3 2.4 2.8 0.9 DecLatPct: 1969 4.7 18.4 28.0 29.7 12.8 0.3 2.4 2.9 0.9 LAT pct: 1970 4.7 18.4 28.1 29.8 12.9 0.3 2.4 2.6 0.8 LAT pct: 1971 4.5 17.6 28.1 30.5 13.0 0.3 2.5 2.7 0.9 LAT pct: 1972 4.5 17.5 28.1 30.5 13.1 0.3 2.5 2.8 0.9 LAT pct: 1973 4.5 17.5 28.1 30.4 13.1 0.3 2.5 2.8 0.9 LAT pct: 1974 4.6 17.6 28.2 30.3 13.1 0.3 2.4 2.7 0.9 LAT pct: 1975 4.6 17.4 28.3 30.3 13.1 0.3 2.4 2.8 0.9 LAT pct: 1976 4.6 17.4 28.3 30.3 13.0 0.3 2.5 2.7 0.9 LAT pct: 1977 4.5 17.5 28.1 30.6 13.0 0.3 2.5 2.7 0.9 LAT pct: 1978 4.5 17.5 28.2 30.5 12.9 0.3 2.5 2.7 0.9 LAT pct: 1979 4.3 17.5 28.3 30.5 13.0 0.3 2.5 2.7 0.9 DecLatPct: 1979 4.5 17.6 28.2 30.4 13.0 0.3 2.5 2.7 0.9 LAT pct: 1980 4.3 17.3 28.4 30.6 13.0 0.3 2.5 2.7 0.9 LAT pct: 1981 4.2 17.3 28.5 30.6 13.2 0.3 2.3 2.7 0.9 LAT pct: 1982 4.1 17.5 29.1 31.1 13.2 0.2 1.8 2.3 0.7 LAT pct: 1983 4.1 17.6 28.8 31.2 13.2 0.3 1.8 2.3 0.6 LAT pct: 1984 4.7 17.5 29.0 30.7 12.7 0.3 2.1 2.5 0.7 LAT pct: 1985 4.7 17.5 29.2 30.7 12.7 0.2 1.9 2.4 0.7 LAT pct: 1986 4.7 17.7 29.3 30.7 12.7 0.2 1.9 2.3 0.7 LAT pct: 1987 4.8 17.4 29.2 30.9 12.6 0.1 1.9 2.3 0.7 LAT pct: 1988 4.8 17.6 29.1 30.8 12.6 0.1 1.9 2.3 0.7 LAT pct: 1989 4.8 17.7 29.1 30.9 12.5 0.1 1.9 2.3 0.6 DecLatPct: 1989 4.5 17.5 29.0 30.8 12.8 0.2 2.0 2.4 0.7 LAT pct: 1990 4.9 17.8 29.1 30.8 12.6 0.1 1.8 2.3 0.7 LAT pct: 1991 4.8 18.2 29.5 31.6 12.9 0.0 1.2 1.4 0.4 LAT pct: 1992 4.8 18.4 29.6 31.8 12.9 0.0 1.0 1.2 0.3 LAT pct: 1993 4.7 18.4 29.5 31.8 12.9 0.0 1.0 1.3 0.3 LAT pct: 1994 4.3 18.7 29.4 32.4 13.7 0.0 0.5 0.9 0.2 LAT pct: 1995 4.2 18.7 29.3 32.4 13.7 0.0 0.5 0.9 0.2 LAT pct: 1996 4.2 18.9 29.3 32.5 13.4 0.0 0.5 0.9 0.2 LAT pct: 1997 4.3 19.0 29.3 32.4 13.4 0.0 0.6 0.9 0.2 LAT pct: 1998 4.3 18.9 29.1 32.5 13.5 0.0 0.6 0.9 0.2 LAT pct: 1999 4.3 18.6 29.5 32.5 13.5 0.0 0.6 0.9 0.2 DecLatPct: 1999 4.5 18.5 29.4 32.0 13.2 0.0 0.9 1.2 0.3 LAT pct: 2000 4.3 18.7 29.3 32.7 13.4 0.0 0.6 0.9 0.2 LAT pct: 2001 4.4 18.6 29.6 32.4 13.4 0.0 0.6 0.9 0.2 LAT pct: 2002 4.3 18.5 29.5 32.5 13.5 0.0 0.6 0.9 0.2 LAT pct: 2003 4.2 18.6 29.4 32.4 13.6 0.0 0.6 0.9 0.2 LAT pct: 2004 3.9 18.2 29.8 32.8 13.6 0.0 0.6 0.9 0.2 LAT pct: 2005 3.7 18.5 29.4 33.2 14.3 0.0 0.4 0.3 0.2 LAT pct: 2006 3.7 18.3 29.5 33.2 14.4 0.0 0.4 0.3 0.2 LAT pct: 2007 8.2 17.2 28.4 26.9 11.2 0.0 3.7 3.0 1.5 LAT pct: 2008 8.8 16.9 28.7 26.5 11.0 0.0 3.7 2.9 1.5 LAT pct: 2009 8.1 17.8 28.1 26.7 11.1 0.0 3.7 3.0 1.5 DecLatPct: 2009 4.3 18.4 29.5 32.5 13.6 0.0 0.7 0.9 0.3
Notice the large jump in the lower latitude band in 2007 as the three Hawaiian Airports become a much larger percentage when the rest of the USA evaporates…
This, I think, is a problem.
Musings from the Chiefio 
It’s not just a problem. It’s criminal.
It’s days like this I wish I were a Fox News field reporter.
I’d hound that GISS sob homer simpson lookalike till he broke down in tears.
AHHHHHHHHHHHHHHHHHHHHHHHHH!!!!!!!
The last time Watts caught Hansen with his thumb on the scale, GISS used the excuse that they get their raw data from NOAA. You might get a cross check thingie from them.
Check http://www1.ncdc.noaa.gov/pub/data/ushcn/
If you haven’t already.
Could this be the V2 you were talking about?
http://www1.ncdc.noaa.gov/pub/data/ushcn/v2/
Reading this I start to wonder whether the climate change movement is just misguided ( as i had thought) or whether it is deliberately leading to something. How odd to lose all the cold temperature stations in the 2 year run up to Copenhagen. And the lost stations will have their data amalgamated into the other nearest stations… ie the ones on the coast in the case of California.
I hate conspiracy thoerists, but i find myself being lured into believing in one myself! I think I just prefer unscrupulous though.
But if youwant to know how Government can behave just look at this article in today’s UK papers
http://www.mailonsunday.co.uk/news/article-1222769/Dishonest-Blair-Straw-accused-secret-plan-multicultural-UK.html
If this was a product of a conspiracy, there would be two classes of conspirator. One class, could be the one person who didn’t post ALL the reports in the list. Yes, you do need more than one person for a conspiracy, but read on.
The other class is those people who use these data, recognized the deletions, AND didn’t squawk. It’s this second group that I wonder about. Maybe they did squawk. Can the only users of these data really be that sleepy?
Where would the squawks, if there were any, show up?
It is very hard for me to understand how the element complement in these time series could change as much as E.M has demonstrated without someone making a stink. Or is it that no-one really uses them?
With Anthony’s logging of individual station siting biases, and now the discovery of locational bias in the listing of reporting stations, one might conclude that anything done with these data sets for years after the first mass deletions can have little validity.
Appalling. But publishable.
Thanks, E.M. for these very useful investigations.
Well, a peek at:
http://www1.ncdc.noaa.gov/pub/data/ushcn/
and the other link PaperTiger gave above (h/t!) has a station list in it. For California, there are many stations (though I’ve not yet found a list of “when last reported” for each station; but on my list for just after “coffee, croissant, and the cat” will be to download and unpack the latest monthly file and inspect the contents for California…)
As yes, IMHO:
Hanlon’s Razor is a very important tool, but it does say: Never attribute to malice that which is adequately explained by stupidity. In this case, stupidity is just inadequate.
I know of no accidental behaviour that can delete all the cold thermometers while leaving the major metro thermometers on the beaches in the series…
Oh, and per the “users not catching deletions”: That, to me, does look adequately covered by stupidity. It took me darned near a year to figure it out and I’m staring at the GHCN data set directly. This is what feeds into GIStemp, what goes into the “mouth”. Most “users” look at the the anomaly maps and the “homogenized” data food product that comes out the, um, er, ah, “other end” of GIStemp.
While good practice ought to include looking up stream at the quality of what goes into any data set you use for your work, GIStemp is a nearly impenetrable jungle (though I’ve been whacking it down to size) and the GHCN data are encoded in a way that most folks will “glaze” over. (I know it isn’t “sexy” and that only 1 in 1000 blog readers will even look past one of the charts of GHCN data I put up while 1 in 100,000 might actually look at the patterns of numbers… but I do it because that is how a good forensic audit is done. Down in the weeds, belly of the snake land…) Given that, a typical user is unlikely to discover this.
Though, yes, I would have expected someone who, say, specialized in the study of Redwoods or Conifer Forests to have noticed that not a single thermometer record existed in the GHCN for their study area. (Kind of makes any dendrochronology study in California after 2007 a moot point.)
FWIW, the link above for USHCN lists these stations in California:
So the steps of a forensic exam would now “bound the bounder” by confirming those stations have data in the USHCN set. Then you would have the “is here – isn’t there” and you work in from both directions to “evaporated right here”. That’s were you start lifting fingerprints and looking for who has sweaty palms.
And yes, at this point IMHO this has turned from a “Characterize GIStemp and audit how it works” into a “Forensic audit for establishing guilt”. (With a small possibility that the guilt will only be for an astounding level of stupidity – probably rising to the standard for dismissal.)
How much of this I can do from “outside”, and why the GAO or equivalent is not doing it from the “inside” are both TBD – To Be Determined.
One useful tool for this is an Org chart. You put up the folks who control both USHCN and GHCN along with GIStemp and GISS. Then you find the lowest level where they all have common management. The implementation is most likely below that point. The “decision to do the deed” may be above that point (and if so, there ought to be management foot prints in the snow. email, memos, meeting schedules with the next layer down); or it may be below that point if a group of underlings took it on themselves (in which case, few foot prints in the snow, but folks had to work inside their areas and without authority and that, too, leaves footprints… though a bit muddier…)
An interview with the “management team” and watching how they react to the news of a forensic audit is often informative. Needs someone with a good “people reader” running the meeting. Who blusters too much. Who is bored and uninterested. Who is a bit too stiff. Who defends the decisions. etc. It all paints a picture and can often tell you exactly where to “Dig Here!”. You see two folks give a quick glance, then their boss defends it as “perfectly reasonable” while HIS boss is looking quietly pissed and making notes; you probably have the two who ‘did it’, the one who told them to, and the “clueless uberboss” suddenly getting clue…
God what I’d give for a subpoena and a badge right now.
Well, looks like USHCN is collecting the data and does have it in their data set. From:
http://www1.ncdc.noaa.gov/pub/data/ushcn/v2/monthly/
File: 9641C_200907_raw.avg.gz dated: 07-Aug-2009 16:18 Size: 3.3M
Has data described in part as:
These variables have the following definitions:
ID is the station identification code. Please see “ushcn-stations.txt”
for a complete list of stations and their metadata.
ELEMENT is the element code. There are four values corresponding to the
element contained in the file:
1 = mean maximum temperature (in tenths of degrees F)
2 = mean minimum temperature (in tenths of degrees F)
3 = average temperature (in tenths of degrees F)
4 = total precipitation (in hundredths of inches)
YEAR is the year of the record.
And that file has an entry for Berkely:
04069332008 -9999 514 542 556 596 625 623 635 638 636 591 493 -9999
who’s description from above is:
040693-04 37.87 -122.27 299 CA BERKELEY
So the question now comes down to:
Who decided to remove these USHCN sites from the GHCN and why does the GIStemp USHCN file “cut off” at 2007?
If those two decisions were made by the same person or group, there is your “smoking gun”.
“DIG HERE!” – with a coal powered steam shovel…
BWT, other sites, too, have entries in that monthly data file. I just chose Berkeley as an example.
Yreka is in too:
04986632008 308 365 413 463 575 639 730 718 649 531 437 323 512
049866-01 41.72 -122.63 2625 CA YREKA
Though I notice that for both of them there is an odd “discontinuity” just after the first station ID numbers.
Yreka is 049866-3 in the data record and
and it is 049866-01 in the description.
Similar thing for other records.
So a “sanitation issue” prior to deciding this is fraud is to look for some programatic error that did not catch and adjust for a change of minor modification number. If it can be shown that this is a “failure to do maintenance on the update code”, then the blame “only” rises to the level of “dismissal with prejudice”… And it would be very interesting to see who set the priorities of the “maintenance team”…
IMHO, the issue is now at the interface between the USHCN dataset (which is clearly being maintained) and two objects:
1) GHCN that is not getting the new USHCN data.
2) GIStemp that is not updating using the new USHCN data and so has complete crap as its output.
Anyone know who has management responsibility for that interface? I’d really like to start collecting names for my org chart…
UPDATE: (After 2nd cup of coffee and feeding cat ;-)
Curiously, the files at: http://www1.ncdc.noaa.gov/pub/data/ushcn/ two levels up show a cutoff date of 2007 for last update:
and it is an hcn_doe_mean_data type file that is the input to GIStemp. That probably explains why GIStemp cuts off USHCN data in 2007. These files are no longer updated. That’s one “decision maker point”. So the other one is: Who decided to remove the selected entries from GHCN?
So I guess the next step is: “Who owns GHCN?” and “Who decided in 2007 that these records would be eliminated?”
One More Thing…
Since STEP2 can delete station records, I was a tiny bit worried that some of these 136 “kept” records might not be “kept” by the time GIStemp was done with them. So I checked.
STEP2 makes a “deleted short records” log. Here are the records that it felt ought to be discarded:
Now a very interesting thing about this list. It is longer than the total number of records kept for 2008.
$grep "425 dropped" short.station.list | wc -l 139 $This is easily explained. There are about 1850 station records going in and GIStemp STEP2 will be tossing out records from some time periods that do not include 2007 or newer. It will take a comparison of this list to the list of kept stations (and perhaps going so far as to look at the exact time interval of the “tossed” segment) to sort out if any of the “kept 136” are “unkept” in their entirety.
I’m not sure it’s worth it right now. But as an example:
has a “1” in the “substation field” while
has a “0”.
So at least part of the history of one of the stations that was “kept” was “tossed”… Though this graph does not make it look like a very important piece (but you never know…):
http://data.giss.nasa.gov/cgi-bin/gistemp/gistemp_station.py?id=425911650000&data_set=0&num_neighbors=1
IMHO, the “jump” from 1950 to 1980 just shows the transition of the airport from a little bit of nowhere in WWII (probably not more than grass and dirt, but a history lesson would be nice…) to a major tourist destination Jet Port (as when my spouse and I went there on our honeymoon just about then when Kauai was a “new undiscovered” place in the islands. Not much change before that time, not much change after.
Yup: http://hawaii.gov/lih/airport-information/airport-history
shows the temp rise is more or less directly proportional to the growth history. Modulo the occasional dips that seem to be within a year or two of significant US economic slowdowns and recessions.
There is a possible explanation (not particularly exculpatory IMO).
Watts went on a cross country trip to the headquarters of NOAA early in 2008 to make a presentation / slide show of the surface station project. During that trip Dr. Bruce Baker announced the near completion of the U. S. Climate Reference Network.
http://wattsupwiththat.com/2008/04/24/road-trip-update-day-2-at-ncdc-and-press-release/
NOAA today [april 24th. 2008] announced it will install the last nine of the 114 stations as part of its new, high-tech climate monitoring network. The stations track national average changes in temperature and precipitation trends. The U.S. Climate Reference Network (CRN) is on schedule to activate these final stations by the end of the summer.
NOAA also is modernizing 1,000 stations in the Historical Climatology Network (HCN), a regional system of ground-based observing sites that collect climate, weather and water measurements. NOAA’s goal is to have both networks work in tandem to feed consistently accurate, high-quality data to scientists studying climate trends.
Maybe GISS gradually stopped using the HCN in favor of the new automated CRN.
Still troubling – a system of thermometers that rely on the power of averages rather then precision for it’s accuracy being downsized to include only the most populated cities of California, guarantees a global warming signal no matter what sort of electronic readout gizmo they use.
Seems a lot of Air Force (and some Army) bases are being dropped and I would think those would have among the most accurate records. I also noticed that several F.S. Ranger Station thermometers dropped out in Montana and Idaho, most in reliably cold locations (perhaps these dropped out when automatd RAWS stations started up, though). Also, stations such as Mullen Pass (above) which is just outside of Couer d’Alene, Idaho, is a reliably cold site. Ditto Drummond, MT and Whitehall, MT. Seems a lot of cold sites drop.
Also, the remaining sites in Idaho (Pocatello and Boise) are very close to the southern state boundary and leave no measurements of North Idaho which is mountainous (as opposed to desert). Same for Billings and Missoula, MT. I live about 10 miles from Missoula and have monitored temps and ppt. for about 6 years. I always have colder temps and higher ppt. than Msla. Leaving out the other major cities of Great Falls, Kalispell and Bozeman will give a decidedly slanted view of Montana weather.
But, then, you already knew all that.
GIStemp uploads the GHCN data directly. It comes “en block” as a total history (not monthly updates).
So the observation about CRN might be more usefully stated as:
Did GHCN decide to only include CRN stations in about 2006 or 2007 ?
If so, then we have the 2 decisions that have broken GIStemp for any use after 2006:
USHCN moved to a V2 data format and no “maintenance” programming was done on GIStemp to merge “old” USHCN with the “new” ignored V2 USHCN data. (A fault I’m going to fix in my “SmithTemp” version Real Soon Now ;-)
GHCN also does not move to incorporate V2-USHCN and instead moves to CRN. Effectively eviscerating the US temperature record as far as GHCN is concerned from 2007 forward.
The two, together, leave GIStemp completely and functionally broken for any date after December, 2006. (And “dodgy” for any date before that too…)
It fits.
Still need to know “who made what decision” (no GIStemp maintenance, GHCN discontinuity introduction, USHCN format change and non-backward compliance on the major dataset into GIStemp).
Still need to confirm a “GHCN going CRN” decision (ought not be too hard).
I’d count as at least “Termination Stupidity”. Proving malice would take a bit..
I had thought (hoped) that a change in equipment might be the answer (as papertiger suggests). When a new record becomes available is it may be included in the raw data but will not actually be used in the final set for 20 years. This is apparent with new records in Antarctica, for example, but does also seem to happen sometimes with overlapping records. I was hoping to see that many stations changed to new instruments and the old instruments, which tend to be operated in parallel for a few years, were abandoned in 2007, while the new instrument record, which GIStemp regards as a separate record, is not condisered long enough to pass QC.
Using the NOAA station locator (http://lwf.ncdc.noaa.gov/oa/climate/stationlocator.html ) I have checked a few dropped stations
For example, here is the list of data products for Pasadena:
http://mi3.ncdc.noaa.gov/mi3qry/dataProductGrid.cfm?fid=1818&stnId=1818&PleaseWait=OK
I have also looked at locations, updates and other information.
This may still be an explanation, but there seems to be no valid reason for for the mass extinction of records in the US in 2007 (or in Australia in 1992). Perhaps GIStemp goes back further than we think and is resoponsible for the demise of the dinosaurs.
The decrease in stations is 93%.
The station in Houston is at Bush Intercontinental Airport, a spot that should be disqualified. They shouldn’t be using any thermometers anywhere near major urban areas, much less that airport, which has grown and expanded over the years.
Oh, but wait. This would give them all sorts of reasons to “adjust” the data. Muhahaha…
btw, does anyone know how much uncertainty is inserted into the final product by all of the adjustments? Every adjustment by definition involves some degree of uncertainty, and these uncertainties will accumulate and make the data less certain than it was before. There should be a full accounting of these uncertainties and how they affect the data.
@E.M.Smith: that certainly fits. Each one of the records I’ve checked terminates in May 2007 and the last actual Annual Mean in the record is 2006.
This may be out of GISS’ control, but it is very convenient.
As you wrote in a post some time ago, “Use the anomaly Luke; the anomaly will save us” will be the defence – that it is not a problem.
However, it is still a problem. The issue is proving it.
This has not been independently validated yet. I could have just gotten a bum download of the file. I would like another party to confirm this prior to everyone making a big Hoopla over it. But if it holds up to scrutiny and independent verification…
I’m seeing things about checks and station drop outs from Dennis Ellie and Harold. Is it safe to call this a thing?
Cause four thermometers to cover California, three of them in the three largest population centers of the state, that’s so ridiculous even a Marin County resident is going to shake their head in disgust.
I’ll wait on word from you.
REPLY: Well, since I first said that: I’ve shown that the data really are not making it into the USHCN files that GIStemp uses. I’ve shown it isn’t in the GHCN data at the GIStemp site nor in the copy I’ve got. And we’ve racked up a “reasonable” set of steps to reach this point (CRN subsititution, non-maintenance on GIStemp in the USHCN-v2 transition). So, well, I’m feeling comfortable that I’ve not made a “foobar” here. While I’d still like some 2nd party to hit the ftp link, and count the 2008 records in v2.mean, well, I’m ready to “call it a thing”. -ems
OK, since nobody has yet given me an independent validation, I’ve done a 2nd one myself:
Context is that I’m on a different network, on a different machine, and a different architecture and operating system. I’ve downloaded the file via a browser rather than via a Linux ftp command and decompressed it with a different utility. Here is an “ls” showing the two files:
v2.mean
v2.mean.Z
Here are the commands, as executed in a Mac rather an Linux environment. (Yes, they are “unix style” commands, but the Mach base of Mac OS is not the same as the Linux base):
Snow-Book:~/Desktop chiefio$ grep ^425 v2.mean > v2.USA Snow-Book:~/Desktop chiefio$ grep "2008-" v2.USA | wc -l 5 Snow-Book:~/Desktop chiefio$ grep 2008- v2.USA > v2.USA.missing Snow-Book:~/Desktop chiefio$ cat v2.USA.missing 4257229000042008-9999 132 148 167 175 195 209 224 217 206 181 136 4257229500042008-9999 138 155 171 178 203 211-9999 203 206 178 127 4257236300002008-9999 56 89 137 191 252 253 237 194 145 90 80 4257246900002008-9999-9999-9999-9999-9999-9999-9999-9999-9999-9999-9999 -29 4259119000002008-9999-9999-9999-9999-9999-9999-9999-9999-9999-9999-9999 231 Snow-Book:~/Desktop chiefio$ grep "2008 " v2.USA > v2.USA.2008 Snow-Book:~/Desktop chiefio$ cat v2.USA.missing >> v2.USA.2008 Snow-Book:~/Desktop chiefio$ wc -l v2.USA.2008 136 v2.USA.2008 Snow-Book:~/Desktop chiefio$So we have a fresh trail from download to extraction to record selection to counting.
Still 136 thermometer records in GHCN for the USA.
So the only thing “still in common” is me. I did both of the examinations. (Well, all three if you include the 2nd test of the original download on the Linux box). But I had an EE in Computer Science with over 30 years experience with writing software in a UNIX / Linux environment “desk check” my work on this test. 2nd set of eyes. 2nd brain. 2nd set of experience behind it all.
While I’d still like a second completely independent validation, I’m willing at this point to “call it a wrap”.
GIStemp has a fatal flaw in the initial data load that makes it completely useless for any date after 2006. Since the GLOBAL average temperature and the GLOBAL anomaly maps depend on the high percentage of US thermometers in the data set in the past, changing that number by a factor of 10 in the present, and with clear geographical misrepresentation, makes any present usage of GIStemp products invalid.
Let the Hoopla Begin.
E.M., I can import the data tomorrow and run some verification for all of your yearly station counts. (I’ve been developing applications and databases for about 25 years.)
If you need help with verification in the future, you are welcome to email me directly. I’m fascinated in general with the study of trends (stocks, hurricanes, temps or whatever).
btw, I had a hard time this summer with tomatoes. The night time temps would not drop below 80 degrees for many weeks and the buds just would not set. I’ve also had four waves (generations?) of horn worms work the plants over. I’ve never seen so many horn worms. They are bright green and gooey on the inside.
This must be a joke or maybe a test to see if airport sites have a different profile from the total US thermometer population.
Almost all these latitude and longitudes turn up as airports! The rest are within a couple of miles of an airport (lat long roundoff error?) HONOLULU might be the exception.
Could this be an airport subset for site quality testing, with the full data set stored at a different file location?
@opa
HONOLULU is the main airport at Oahu. It has a “U” for urban and an “A” for airport. The population is give as 836,000 (if I’m parsing it right…).
No joke though. FWIW on the “Islands in the sun” posting a commenter linked to an official pronouncement that ALL the sites in the Pacific Ocean were to be airports. All others are to be closed…
@Harold
Thanks. It may be paranoia on my part, but this is a rather important claim and I really would like some other party to test it just to be sure I’m not being some kind of Clever Hans behind…
Also, FWIW, since I stopped spraying my garden and leave the wasps nests alone, I now have about a dozen wasps that patrol the garden eating bugs. Haven’t had a tomato horn worm since (though on 2 occasions I did see a about a “hand sized” patch of the slightly chewed stuff the baby worms give you and did see one worm about the size of a rice grain; but the next day the plant had been completely rid of them by the wasps. And they don’t bother me at all, just constantly search the plants for bugs.
Don’t know if your wasps would be the same kind, though. These are some kind of California yellow jackets (paper nest); not the “mud dauber” kind we had in the central valley when I was a kid. Oh, and I got a crop of lady bugs this year that cleaned up a late summer aphid outbreak on some collards. It’s not a big “green faith” thing with me. Just one year “gave up” from too many bugs; and the system took care of itself about 2 years after that.
It took a while to find, but I think I found “who owns GHCN” and “who manages it”.
From: http://gcmd.nasa.gov/records/GCMD_GA_CLIM_GHCN.html
We find that:
GHCN data is produced jointly by the National Climatic
Data Center, Arizona State University, and the Carbon Dioxide
Information Analysis Center at Oak Ridge National Laboratory.
The NCDC is a part of NOAA. So I’m not seeing NASA on this list. But…
It goes on to say:
Personnel
SCOTT A. RITZ
Role: DIF AUTHOR
Phone: 301-614-5126
Fax: 301-614-5268
Email: Scott.A.Ritz at nasa.gov
Contact Address:
NASA Goddard Space Flight Center
Global Change Master Directory
City: Greenbelt
Province or State: Maryland
Postal Code: 20771
Country: USA
So it looks to me like it has NASA staff assigned, part of Goddard (though it isn’t clear to me if G. Space Flight Center and G.I.S.S. are siblings or if one is a parent of the other. I suspect GSFC is an underling to GISS. That would have Scott Ritz reporting to Hansen IFF I have this figure out… (And all that personal data is at the other end of the link anyway so I’m not publishing any private data NASA has not already published.)
Still have not been able to find any reference to CRN replacing USHCN in GHCN or anything like that.
It’s looking to me like GISS has their fingerprints all over the GHCN deletions, with NOAA ether as patsy or passive cooperator.
The v2.mean file is chock full of -9999 values. The data set has more missing values than I would have expected to see. I was surprised.
I haven’t yet looked at the documentation for the v2.mean layout. Here is my guess, though:
1 to 10: station identifier
11 to 12: some sort of status code (pertaining to the station?)
13 to 16: year
17+: 12 columns of five characters each containing the monthly means in celsius with an implied decimal in tenths (masked decimal 1 or “MD1” in Pick/Revelation lingo)
Stats for GHCN v2.temperature.inv:
Total records: 7,280
Stations in U.S. (beginning with “425”): 1,921
26% of the stations in this table are in the U.S.
Stats for GHCN v2.mean:
Total records: 595,748
Records for U.S. (station ID beginning with “425”): 195,576
Some record counts for the U.S.:
2005: 1214
2006: 1177
2007: 134
2008: 136
33% of all records in the v2.mean table (a global data set?) are from the U.S.
Results were generated by Advanced Revelation v3.12 running on a Windows XP Pro workstation and a database server running Windows Server 2003. The filing system uses linear hashing, and the records are all variable length. I’m guessing that this system (the technology) is dissimilar to the one that you are using.
Other stats from v2.mean for the U.S.:
Total monthly mean values: 2,098,944
Total missing monthly mean values: 84,564
In double-checking the total record counts, I counted 179,382 records for the U.S. in v2.mean. (The total record counts in my previous post were based on a counter that my subroutine displayed during the import process. I will have to check the subroutine to see what it was doing.)
E.M., can you verify total records for the U.S. in v2.mean? 179,382?
Ok, now I get it. The station ID in v2.mean is eleven characters, but there is a significant character (maybe a version number?) in the 12th position just before the year. It looks like a sequential counter.
Here is an example of the same station (location: ARANYAPRATHET) with a bunch of versions (or locations?):
22848462000 0 (1951 to 1990)
22848462000 1 (1951 to 1991)
22848462000 2 (1951 to 1975)
22848462000 3 (1961 to 1970)
22848462000 4 (1971 to 1980)
22848462000 5 (1987 to 2009)
The record keys have to be unique in Pick databases so I will have to revise the import subroutine to include the counter in the second position of the key.
Old key: 22848462000*1951
New key: 22848462000*0*1951
This explains the discrepancies in my previous counts.
Note that I have yet to read any of the documentation for this data table. Maybe I should look before I leap next time. lol.
E.M., how do you process a station with many versions or locations? Are they counted as the same station?
Regardless of what happens, I will still get the same counts that you got for the U.S. (134 in 2007 and 136 in 2008). A station is still a station, right?
@Harold Vance:
First off, thank you so much for doing this validation. I really appreciate a completely independent “from scratch” reconstruction. I’m also quite comforted that you, too, found the same result. That it was done “without the docs” actually prevents an error in the docs from leading us both to the same place. Believe it or not, that is one of the more ‘advanced’ forms of QA… Doing it “blind”…
The layout of the v2.inv (station inventory) file is described in the link here:
https://chiefio.wordpress.com/2009/02/24/ghcn-global-historical-climate-network/
The start of this record is very similar to that of the v2.mean file, so you can use these descriptions for the first few fields to know what you’ve got.
About half way down:
c ic=3 digit country code; the first digit represents WMO region/continent
c iwmo=5 digit WMO station number
c imod=3 digit modifier; 000 means the station is probably the WMO
c station; 001, etc. mean the station is near that WMO station
About here, it starts to diverge from the v2.mean file:
c name=30 character station name
c rlat=latitude in degrees.hundredths of degrees, negative = South of Eq.
c rlong=longitude in degrees.hundredths of degrees, – = West
c ielevs=station elevation in meters, missing is -999
c ielevg=station elevation interpolated from TerrainBase gridded data set
c pop=1 character population assessment: R = rural (not associated
c with a town of >10,000 population), S = associated
So the first 3 are the country, then an 8 digit unique Thermometer Identifier that can be subdivided into 5 digit WMO location, 3 digits for a sublocation (i.e. 2 at the same airport get the same 5 digits, but the last 3 will be different if one is on the roof and the other on the tarmac).
Now here is where it gets just a bit different. The v2.inv file has information about stations. The v2.mean file has records for each thermometer each year AND any different “modification history” for any given thermometer. So, say, Bob reads the thermometer every morning at 8 am, but Sally reads it every night at 9pm. Both records could be recorded, but both would get a different Time Of Observation Bias modfication. So you could get a 425xxxxxyyy0 and 425xxxxxyyy1 record for the same thermometer.
Because of this I will often differentiate between number of thermometers (3+8=11 long identifier) vs the number of thermometer records (3+8+1=12 long identifier) in a given analysis.
Then you get the 4 digit year, then the 12 temperature fields.
Harold Vance
In double-checking the total record counts, I counted 179,382 records for the U.S. in v2.mean. (The total record counts in my previous post were based on a counter that my subroutine displayed during the import process. I will have to check the subroutine to see what it was doing.)
E.M., can you verify total records for the U.S. in v2.mean? 179,382?
I get:
Snow-Book:~/Desktop chiefio$ grep “^425” v2.mean | wc -l
194576
Snow-Book:~/Desktop chiefio$
Total records starting with 425 (i.e. unique lines for each year, stationID of 8 char and modification flag of 1 char) which is remarkably similar to your import record counter modulo a “194” vs “195”. Is it possible you mis-typed a digit?
The last few records in the file with a 2009 year are:
Snow-Book:~/Desktop chiefio$ grep “^425” v2.mean | grep “2009 ” | tail
4257275300002009 -168 -113 -44 56 121 176 192 189 184-9999-9999-9999
4257276400002009 -130 -103 -58 51 124 164 194-9999 184-9999-9999-9999
4257276700012009 -135 -123 -63 52 111 160 188 185 180-9999-9999-9999
4257277300002009 -42 -9 10 66 120 161 207 196 167-9999-9999-9999
4257278100002009 -9 17 40 92 145 197 233 215 168-9999-9999-9999
4257279300002009 40 54 55 96 135 177 209 187 169-9999-9999-9999
4257279700002009 45 46 48 75 101 138 157 153 145-9999-9999-9999
4259116500002009 214 210 209 211 236 258 261 259 261-9999-9999-9999
4259118200002009 226 230 232 237 256 281 277 275 281-9999-9999-9999
4259128500002009 211 208 205 206 236 241 244 245 239-9999-9999-9999
Snow-Book:~/Desktop chiefio$
which shows they have data through September (i.e. it is the current released copy).
and the number of records total in the file are:
Snow-Book:~/Desktop chiefio$ wc -l v2.mean
595748 v2.mean
Snow-Book:~/Desktop chiefio$
which would indicate that we got the same file, but are perhaps counting “records” differently.
And yes, your equipment, approach, software, and methodology are entirely different from mine.
Yes, I mistyped a digit. The count for prefix “425” should have read 194,576.
I fixed the import routine to include the “thermometer” designation in the key. I got the same totals that you got with the grep utility. Although I use Ubuntu at home, I come from the DOS/Win camp and haven’t gotten up to speed on all of the gnu tools and utilities.
With Advanced Revelation (or OpenInsight or Pick) the command would look something like this:
COUNT GHCN_V2_MEAN WITH STATION_ID “425]”
(GHCN_V2_MEAN is the table that I created before importing the v2.mean table. Of course, the fields also have to be defined before they can be used with a SELECT or a COUNT command but this is all very trivial work.)
Here are my updated counts for total values (means) and missing values (aka “-9999”) in v2.mean:
U.S.
Total: 2,334,912
Missing: 87,522 (3.7%)
Global
Total: 7,148,952
Missing: 376,808 (5.3%)
Many thanks for the explanation of how the station ID schema works. Yes, I did approach this topic blind, and I also documented my goofs (mistaken assumptions) for everyone to see. I was too lazy to read the documentation, and anyway, the decoding of the layout from scratch is a lot more fun. ;-)
Harold Vance
Yes, I mistyped a digit. The count for prefix “425″ should have read 194,576.
I fixed the import routine to include the “thermometer” designation in the key. I got the same totals that you got with the grep utility.
Great! So I’m comfortable at this point calling it “confirmed”.
Although I use Ubuntu at home, I come from the DOS/Win camp and haven’t gotten up to speed on all of the gnu tools and utilities.
I have a love / hate relationship with the Unix / Linux tools. I love what they let me do in a few keystrokes and I hate how they make me do it. Oh well, it’s built into my brain now ;-) FWIW, noone is “up to speed on all” of it. There is more in a Unix / Linux box than any one person can ever learn. Folks learn all the bits they need for what they do, and let the rest sit. (There are hundreds of folks extending it every day, so every day the capabilities extend by more than you can learn… Heck, I only know about 10% of the grep options and I’ve been using it for 30 years!)
With Advanced Revelation (or OpenInsight or Pick) the command would look something like this:
COUNT GHCN_V2_MEAN WITH STATION_ID “425]”
I don’t know how much “programming” it takes to describe your “test rig”, but if you would like to “put the code up” for others to use, let me know. IMHO, the more folks who are able to take this data and “kick it around” the better.
Yes, I did approach this topic blind, and I also documented my goofs (mistaken assumptions) for everyone to see. I was too lazy to read the documentation, and anyway, the decoding of the layout from scratch is a lot more fun. ;-)
And that is exactly how this kind of QA test is supposed to be done. Each mistake shows a likely place that others might have made a similar mistake, it is valuable information. Running without the docs avoid the docs guiding two teams into a consistent error. I’d thought of asking someone to do it that way but figured I would have been asking too much 8-0.
One of the hardest things for folks to do, and yet one of the best to do, is to simply embrace any “oddity” or “error” and showcase it along with the parts that “went right”. Sometimes you find out you made a typo, sometimes it is the key to finding where the other guy has a broken analysis. The only “wrong” thing is to hide it or be embarrassed by it. Often the most interesting stuff comes out of the statement: “I don’t know why, but I got {foo} while you got {bar}.”
And in a very strange way, when you have a “fail to confirm” then study the issue even more getting an even stronger “confirm” and show that the “fail” was an early error, that actually makes the final “confirm” all that much stronger. While I have some faith in a “confirmed right out the gate in one go”; I have much more faith in the “almost matched, oh, wait, found this little typo and that off-by-one-char in the load script, now it confirms”. It tells me the code and process was gone through a couple of times; not just copied and run from the original, errors and all replicated.
So again, thanks, and a hearty: Well Done!
Now that you have your own climate database, just think what analysis you can do ;-)
If you find anything interesting, we can post it here.
E.M.,
I sure hope you’ve got a saved copy of the GISSTemp output for comparison in case of sudden, unannounced changes!
At this point, I’ve saved a couple of copies of the input files from different months, a bunch of intermediate results, a few copies of the source code, the output, …
Any “unannounced changes” will be met with “unannounced analysis of the changes” 8-|
When you have a copy of the initial state, any changes become forensic foot prints in the snow. I like footprints in the snow. They take you to interesting places ;-)
I would suggest that anyone familiar with their local BOM download and save any data sets they have public. Under the “Aussie” thread, we’ve found out that the Aussie BOM has decided to re-write the past and has announced their pre-2007 (gee, that’s a familiar date…) data sets are now “obsolete”…
I’d do it, but there just isn’t enough of me to cover every country in the world. And yes, I do think we need to start preserving the oldest copies of data available. I think it is going to be needed as “evidence” and I think the folks in the crosshairs are going to start trying to burn the evidence.
But it’s too late to change the data for the USA… I already have my copies…
A good thing to do now would be to find what GISS has to say about the station dropout.
Do they give a rational? Was it announced in advance? Is the topic covered in FAQ?
So I took a look.
Some answers are found via IceCap US_AND_GLOBAL_TEMP_ISSUES.pdf.
Joe D’Aleo is all over this topic. Like for instance did you know that GHCN stations don’t get any UHI adjustment?
Explains the “why” for using San Diego (+2 C), Los Angeles (+4 C), and San Francisco (+2 C), pretty clearly.
What you have stumbled on by accident is the “human caused” warming. It all comes from an NCDC office in North Carolina, has zero to do with CO2, and it’s definitely deliberate.
Reflecting on the list of cities included in the US portion of the GHCN, I’m struck by one thing.
THEY ARE ALL CITIES! HUGE CITIES> THE LARGEST URBAN AREAS AVAILABLE>
This is fraud. Straight up. I want lawsuits against these clowns.
Harold Vance E.M., how do you process a station with many versions or locations? Are they counted as the same station?
FWIW, this is entirely up to the designer of the data series processing programs. The way GIStemp does it is truly bizarre and involves much of the more questionable parts of it.
For my tests, I just combine all the data. Since I’m making an average of the records, I see no reason to toss one and keep another. The whole point started out as making a benchmark to measure the changes done by GIStemp from the “raw” data, so the best thing for that purpose is just “average it all together”. Then you can see what GIStemp does and ask “is that better?”.
Along the way I started seeing these patterns…
What would I do if I were making a temperature history product? Probably something very similar (and unlike GIStemp). Though I think I would start with taking ALL data series at a single location (5 digit code) and averaging them together to get one series for that place. If it is OK to average a whole state, it ought to be ok to average 3 thermometers at one place…
Then, with discontinuities between segments, you have a choice. To splice or not to splice. GIStemp goes for a strange mix of “toss some segments, interpolate other segments, average some other bits, fill in gaps based on other stations 1000 km away, and a bit more”. I can see little justification for that level of data fabrication other than to try and hide just how horrid the base data are and how unsuited they are to the intended use of making global ‘anomaly maps’.
I would either just accept the gaps, and annotate the data quality in each year (% dropout figure); or I’d need to cook up a valid splicing method (which is where I think Hansen lost it.) This, IMHO, is the worst part of the process. How to fabricate valid data where there are none.
And this ties into the point PaperTiger made about ALL CITIES.
GIStemp does a “look aside” at other places to fill in the missing records. When all those places are major cities, well…
EM,
I was going to take a look at the siting of the San Fran GHCN station, and noticed these surface stations are not on Watt’s surface station survey.
Which is suggestive that they are members of the new Climate Reference Network.
It wouldn’t surprise me one single bit to find the LA station planted on the median of the Ventura freeway.
Where do you get the listing of what is in / not in the CRN?
I haven’t looked into it yet and don’t have a good starting point. I can see a “plausible scenario” where in 2007, the USHCN is changed in format (while GIStemp does not get maintenance to add the USHCN.v2 new data file), a decision is taken at GHCN to add in the CRN stations (and not bother adding in the USHCN.v2 records, since CRN is “the new way”); and both end up dropping 93% of the USA. It would just be a “convenient accident” that this breaks GIStemp in a way that causes 115 year records to fall… So a listing of what IS in CRN would let me do a nice A/B comparison.
It looks like it might be embedded in the 29 MB “Master Station History” here:
http://www.ncdc.noaa.gov/oa/climate/surfaceinventories.html
but it would be nice to figure out where to download just that list, and if there is “yet another file format” with “yet another file” of just CRN, that would tend to support the thesis that the breakage occurred at the USHCN.v2 / CRN transition interface (and the meetings related to it, where decisions were made to drop USHCN.v2 or not fund the maintenance to keep it in… gee, “management decisions”? …)
Unfortunately, that “convenient accident” would still amount to terminal incompetence AND it would not provide “cover” for the deletions elsewhere in the world. (I’ve done the charts, but not yet posted, for Canada and Russia. Canada has a huge drop out, yet all the talk of Siberia turns out to just be talk… Wonder why it is that all the English Speaking countries are doing one thing and the Russians are not…)
UPDATE: from a link at Climate Audit:
http://www.climateaudit.org/?page_id=1686
I found this link:
http://www.ncdc.noaa.gov/crn/newstations?sort_by=loc_state
That gives a CRN station list. No Station ID numbers, but it’s a start.
Most notable for this posting is that these stations are NOT in the 4 used for California:
CA Bodega 6 WSW 20071025 University of California – Davis (Bodega Marine Laboratory)
CA Fallbrook 5 NE 20071019 San Diego State Univ’s Santa Margarita Ecological Reserve (Old Mine Road)
CA Merced 23 WSW 20040325 Kesterson Reservoir (US Bureau of Reclamation)
CA Redding 12 WNW 20030325 Whiskeytown National Recreation Area (RAWS Site)
CA Santa Barbara 11 W 20071019 Univ. of California – Santa Barbara (Coal Oil Point Reserve)
CA Stovepipe Wells 1 SW 20040505 Death Valley National Park (Stovepipe Wells Site)
CA Yosemite Village 12 W 20070919 Yosemite National Park, (Crane Flat Lookout)
While I think this list is way too short and still over represents the coastal areas (where is there anything from the Cascades? Whiskeytown is close, but hardly a Mt. Shasta or Weed. I was stuck on the road all night in a blizzard in Weed once… it gets damn cold. ) At least it isn’t all airports on the beach…
And the GCOS listing from here:
http://www.ncdc.noaa.gov/hofngsn/HOFNGsnStn
has places like Fresno and Eureka on it, so the “high quality airports” listing is not the source.
OK, I tried to “give these guys cover” with plausible alternatives to a deliberate act; but I’m just not seeing it. The two most plausible alternative sources have more data than those 4 California stations. At this point, the presumption shifts to “deliberate act” until shown otherwise as far as I’m concerned.
E.M., thanks for the answer on the stations with more than one thermometer.
I can’t seem to post a comment on your latest thread “Blame Canada.”
Anyway:
EUREKA,N.W.T. 79.98 -85.93 10 222R -9MVxxCO 1A-9WATER A 0
This station is located in Nunavat. It’s pretty close to northern Greenland and not anywhere near the N.W.T.
Maybe their GPS equipment was broken when they were filling out the station ID form?
Here is a link for your viewing pleasure. The clip has lots of cute white bunnies. All they’re missing are big, pointy teeth:
http://lidar.ssec.wisc.edu/Eureka_movie_v3.avi
REPLY: Don’t know why, but found your “code” posting in the spam queue. Approved it and it is on it’s way. And yes, the metadata can be kind of screwed up in a lot of cases. Glad they don’t use it for anything other than adjusting the temperature history of the planet 8-} -ems
Here is the code that I used to import the GHCN v2.mean data. I put all sorts of comments in it to show people what I was doing. The language is R/BASIC, which is a free form, relatively high level language that’s great for processing strings and arrays (records). This subroutine will create a record with a three-part key (station*thermometer*year) and twelve fields that contain the mean temp values.
SUBROUTINE IMPORT.GHCN.V2.MEAN
DECLARE SUBROUTINE MSG
* Initialize source file variables. These were hardcoded for simplicity.
* Source file name must use traditional dos file naming convention (8.3).
* User may need to rename file before running this subroutine. This caveat
* applies to Advanced Revelation only, not OpenInsight.
SOURCE.FILE.NAME = “C:\DATA\GHCN\200910\V2_MEAN”
SOURCE.FILE.INFO = DIR(SOURCE.FILE.NAME)
SOURCE.FILE.SIZE = SOURCE.FILE.INFO
!
* Open data files.
!
OSOPEN SOURCE.FILE.NAME TO SOURCE.FILE ELSE
MSG(“Unable to open file “:SOURCE.FILE.NAME:”.”,”A”,””,””)
RETURN
END
OPEN “GHCN_V2_MEAN” TO FILE.GHCN.V2.MEAN ELSE
MSG(“Unable to open file GHCN_V2_MEAN.”,”A”,””,””)
RETURN
END
* Clear the data file that will be used to store the GHCN V2.MEAN data.
CLEARFILE FILE.GHCN.V2.MEAN
!
* Initialize variables.
!
EQUATE True$ TO 1
EQUATE False$ TO 0
* Set record delimiter to ASCII character 10 (the line feed).
RECORD.MARK = CHAR(10)
* Set byte location to 0 (the start of the source file).
BYTE.LOC = 0
* Set row count to 0.
SOURCE.ROW.COUNT = 0
* Set records written to 0.
RECORDS.WRITTEN = 0
* Set chunk size to any value longer than the actual source record.
CHUNK.SIZE = 1024
DONE = False$
!
* Process the source file.
!
LOOP UNTIL DONE
* Read some data into variable Chunk.
OSBREAD Chunk FROM SOURCE.FILE AT BYTE.LOC LENGTH CHUNK.SIZE
* Find the end of the record (the line feed).
End.of.Record = INDEX(Chunk,RECORD.MARK,1)
* Look for unexpected values for end of record.
IF End.of.Record LT 1 THEN
End.of.Record = 1
GOTO BUMP.BYTE.LOC
END
* Display row counter on screen to indicate activity.
SOURCE.ROW.COUNT += 1
PRINT @(40,10):SOURCE.ROW.COUNT:
* Extract the record.
Source.Record = CHUNK[1,End.of.Record-1]
* Extract the station ID.
Station.ID = Source.Record[1,11]
*** * Use this code when only importing country code 425 (U.S.).
*** IF Station.ID[1,3] NE “425” THEN GOTO BUMP.BYTE.LOC
* Extract “thermometer version” and year.
Thermometer.Version = Source.Record[12,1]
Year = Source.Record[13,4]
* Build the record key. All keys must be unique in the Pick universe.
V2.Mean.Key = Station.ID:”*”:Thermometer.Version:”*”:Year
* Initialize the record.
V2.Mean.Record = “”
* Initialize byte loc for monthly means. Temps begin at byte 17.
Mean.Temp.Byte.Loc = 17
* Initialize the field number in which to store the monthly means.
Field.Number = 1
FOR MONTH.LOOP = 1 TO 12
* Extract mean temp.
Mean.Temp = Source.Record[Mean.Temp.Byte.Loc,5]
* Convert spaces to nulls in mean temp. Only numeric data should
* be recorded in the record.
SWAP CHAR(32) WITH “” IN Mean.Temp
* Store mean temp in record.
V2.Mean.Record = Mean.Temp
* Increment byte loc by five bytes.
Mean.Temp.Byte.Loc += 5
* Increment the field number counter.
Field.Number += 1
NEXT MONTH.LOOP
* Write the record to the data file (Revelation file handle).
WRITE V2.Mean.Record ON FILE.GHCN.V2.MEAN, V2.Mean.Key THEN
RECORDS.WRITTEN += 1
PRINT @(40,11):RECORDS.WRITTEN:
END ELSE
MESSAGE = “Houston, we have a problem. ”
MESSAGE:= “A write error was detected for key “:V2.Mean.Key
MESSAGE:= ” in file GHCN_V2_MEAN.”
MSG(MESSAGE”T#50″,”A”,””,””)
END
* Set byte loc to start of next record.
BUMP.BYTE.LOC:
BYTE.LOC += End.of.Record
PRINT @(40,12):BYTE.LOC:
* Check byte loc to see if it exceeds the source file byte count.
* If not, go get the next record
IF (BYTE.LOC+1) GE SOURCE.FILE.SIZE THEN DONE = True$
REPEAT
!
* End of Job.
!
* Close the source file.
OSCLOSE SOURCE.FILE
* Display the results of the import process.
MESSAGE = “”
MESSAGE = “IMPORT PROCESS FINISHED”
MESSAGE = “Source file name: “:SOURCE.FILE.NAME
MESSAGE = “File size: “:SOURCE.FILE.SIZE
MESSAGE = “Rows processed: “:SOURCE.ROW.COUNT
MESSAGE = “Records written: “:RECORDS.WRITTEN
MSG(MESSAGE,”A”,””,””)
RETURN
Okay, my first comment here.
I noticed right off with the kept station list that some states have massive representation. Alaska (where I live), Texas and Florida being the primary examples. The first two make some amount of sense if the goal were to apportion coverage based on area, but FL pretty much wrecks that theory.
One thing that amazes me if the Annette Is. station is actually at the airport that measurement is being taken at an airport that has been shut down for decades. The hangar building is use for boat storage and the runway as a drag strip. When I lived there the runway was in no condition for use by planes, you used float planes or boat if you wanted to leave.
P.S. It would not surprise me if you could actually get a reasonable idea of U.S. temperature history with 135 or so stations, but you would have to ignore state boundaries while doing so. Having both Los Angeles and San Diego in such a creation is loony. Concentrate on the geographic regions.
Here is a map of California with “some” temperature measuring stations. These are taken from NOAA Western Regional Climate Center site. It appears (by rough eye-ball count) to contain about 200 stations, and about half of those “reporting” and the other half “missing.”
The map: http://tinypic.com/r/oiztxf/4
The website:
It is very curious, that only 4 stations are now active for GISS purposes. What’s wrong with the other 196?
REPLY: “Nice links! Yes, that is the $Trillion Dollar Question… Realize, though, that it isn’t just GIStemp that is buggered by this. It is anyone using GHCN data as input. -ems”
RE “Are 4 near-ocean locations enough to measure California? Would they be warmer than the snowy mountains?”
Are you suggesting there are only four currently-reporting stations in CA? I found 15 listed within a 222 km. radius of Bakersfield alone – and 15 within 140 km. of Sacramento. (Just go to http://data.giss.nasa.gov/gistemp/station_data/ , type in “Sacramento” and hit “Search”. When the Sacramento record is found, just slick the (*) to its left and you’ll get a list of nearby stations, from which you can find those whose records extend to 2009.)
REPLY: [ Note the title of this report. GHCN. That is the input data for most of the worlds temperature series. Yes, it has only 4 left in California. What you are looking at is GISS. That is after GIStemp has been run. That is NOT GHCN.
Until about a month ago, GIStemp merged in the USHCN data set. This Dataset “cut off” in May of 2007. It includes USA stations not in GHCN. As of about a month ago, GIStemp has put the USHCN Version 2 data set in. While this includes data to present, it does NOT fix GHCN (that is still used by other temperature series ‘as is’).
Further, the USHCN.v2 data set is NOT THE SAME as the USHCN data set. It has a newer and different set of changes done to the data that look, on first examination, to have the effect of inducing an artificial warming trend to the data.
So GIStemp now gets “1/2 points” for putting the USA back in, but NOAA / NCDC get “minus 1/2 points” for re-cooking the data. Net effect, it’s still broken. And there are still thermometer deletions in Californis in GHCN.
And your comment on the mountains suggests a misunderstanding of what is being measured here, and why. From http://data.giss.nasa.gov/gistemp/ : “Our analysis concerns only temperature anomalies, not absolute temperature.
REPLY: You are confused. You have confounded “measuring” with “analysis”. GHCN reports the temperatures. These are averages, but not anomalies. These averages of temperatures run through several steps of GIStemp before, toward the very end, GIStemp PRODUCES an anomaly map as the PRODUCT of their analysis. The nature of the product can not protect you from the process…
Temperature anomalies are computed relative to the base period 1951-1980. The reason to work with anomalies, rather than absolute temperature is that absolute temperature varies markedly in short distances, while monthly or annual temperature anomalies are representative of a much larger region. Indeed, we have shown (Hansen and Lebedeff, 1987) that temperature anomalies are strongly correlated out to distances of the order of 1000 km.”
RE “The Thermometer Langoliers have eaten 9/10 of the thermometers in the USA”
As of 2008, the USHCN data-set had 1221 stations – are you saying it now has only 122? Or that the original version had 12210 (it actually had 1219)? As mentioned above, I can find 15 with records (to 2009) within 2 hours drive of Bakersfield CA alone!
REPLY: Again, you have confounded different things. USHCN is NOT GHCN. If you can’t keep that straight, you will stay confused. This posting is about GHCN, not USHCN.
RE “. . . including all the cold ones in California.”
Again, what GISTEMP calculates is CHANGES in temp rather than absolute temperature – “cold” stations can be just as indicative of warming anomalies as “warm” ones if they become less cold over time.
REPLY: [ And equally AGAIN: What GIStemp does has no influence on GHCN. And a product can not protect you from the process. Long before the anomaly step GIStemp has used averages of temperatures to “fill in” missing data (of which there is a great deal) and to calculate Urban Heat Island corrections (which it does badly, often “correcting” in the wrong way). None of that will be reversed by calculating an anomaly “at the end”. ]
RE “This is EVERYTHING fed into GIStemp for the USA from the GHCN file. ( I still need to check USHCN . . .”
You’re joking, right? Almost all of the US data comes from USHCN!
REPLY: No, I’m not joking. I’m accurately reporting and you are confused. Perhaps understandably, since things have changed over time, but confused none the less. When this report was written, USHCN “cut off” in May of 2007. So it was absolutely correct that for the USA, everything from then to date was only from GHCN. In November, IIRC, GIStemp moved to using USHCN.v2 (a few months after this report…) and put the present records back in for the USA. Though, as noted above, it’s a ‘re-cooked’ version that has different data in it (often 1/2 F different, and in a pattern that on first inspection seems to show artificial warming.)
RE your comments on October 25, 2009 at 1:23 pm about deleted records:
I searched for several of these at http://data.giss.nasa.gov/gistemp/station_data/ and found every one. Each one’s data ended before 2009, so they may not be appropriate for the GISTEMP calculations (until they’re updated?). But when I clicked the (*) to the left of each returned record I got lists of nearby stations, lots of them with data to 2009, meaning that there should be no problem constructing anomalies for the respective regions near the older-data stations.
REPLY: Again, you have missed the changes over time noted above.
RE “Did GHCN decide to only include CRN stations in about 2006 or 2007 ?” (another of your Oct 25/09 comments)
Why the confusion? http://data.giss.nasa.gov/gistemp/sources/gistemp.html states very explicitly what land data sets are used by GISTEMP: GHCN, USHCN, SCAR and a single German station.
REPLY: I can not explain to you why you are confused. I clearly stated GHCN, you then reply GIStemp. There is still no answer to why GHCN deleted the stations. There are partial answers to why GIStemp put USHCN.v2 recent data back in.
Are you a skeptic – or a denier?
REPLY: I am a searcher after truth, nothing more. And, mirroring your tone: Are you, Sirah, a troll or a lier? Seriously, if you want to start flinging insulting words at folks, you can earn a rapid place in the SPAM filter. I run a polite place here. Those who are not polite will get “response in kind” and a trip to SPAM land. Now there is no way I can tell if you are a “true believer” who just happens to have an exact match to troll talking points and a well thought out deliberate plan of deception via feigned confusion, or if you are really just so impolite as to not understand that your words show both poor reading comprehension and use of loaded language. But frankly, I don’t need to.
The rest of your talking points get a SNIP!
Small clarification for Harold Vance (October 27, 2009 posting) re the location of the station “Eureka, NWT”.
That was in the NorthWest Territories until relatively recently when Nanavut was created by splitting off the eastern part of NWT.
So Frobisher Bay (now named Iqaliut), Resolute, Cambridge Bay, and Alert for example are now in Nanavut while Yellowknife, Hay River, Inuvik, and Tuktuk are still in the NWT. (Rattling off big airport locations important to trans-polar flights and locals – Very Good Things up there, not many NIMBYs about airports. :-)
The location Harold quotes looks right. As he says it is near Greenland.
Coincidentally, I’ve been close to it for a few hours – long enough to get the Hercules airfreighter loaded and leave. My visit was to an airstrip serving a petroleum drilling rig in the 1970s, a bit south of Eureka on the same peninsula of Ellesmere Island. Flying over Axel Heidberg island to the west was interesting – quite mountainous, whereas the western area of the High Arctic that I visited more is quite low. (Not that I was worried but it crossed my mind that it was nice to hear four engine-prop combinations purring out there on the wings.)
Pingback: ClimateGate: 30 years in the making « The Ninth Law
Pingback: Snows of Mount Hamilton | Musings from the Chiefio