Pages

▼

Sunday, May 31, 2015

GISS Anomaly Estimate for May: A Record High

From the month's U Maine reanalysis data, I estimate GISS's global anomaly for May will be +0.83°C.

That'd be a record-high for the month, smashing last year's record of +0.77°C, and the 6th-warmest of any month since GISS's records began in 1880.

It would also bring the warming in 10 years up to +0.08°C, in 15 years to +0.14°C, and in 30 years to +0.49°C. Seems the Pause is itself pausing.

My estimate last month for April (0.66°C) wasn't very close to the actual value (0.75 0.71°C), and but the relationship between the estimate from reanalysis data and the actual GISS anomaly isn't a strict linear fit:


so don't take my number to the bank.

Nota bene: Turns out GISS corrected their numbers a few days after they first came out (May 15th), which I didn't catch, which explains why I had some incorrect some numbers above. Their new numbers don't change my estimate for May.
at

4 comments:

  1. Jon11:22 AM

    My estimate last month for April (0.66°C) wasn't very close to the actual value (0.75°C)...

    I thought the April anomaly was 0.71°C - that's still what's showing on the GISTEMP site and it doesn't appear, looking at my records, as though there was a big adjustment in the temperatures for Aprils past in the last update there. Am I missing something or is that 0.75°C just a typo?

    ReplyDelete
  2. David Appell12:11 PM

    Jon, you're right, thanks.

    http://data.giss.nasa.gov/gistemp/tabledata_v3/GLB.Ts+dSST.txt

    No sure now how I got 0.71 C....

    ReplyDelete
  3. David Appell12:20 PM

    Turns out there was a problem, and GISS published corrected data on May 15:

    "May 15, 2015: Due to an oversight several Antarctic stations were excluded from the analysis on May 13, 2015. The analysis was repeated today after including those stations."

    http://data.giss.nasa.gov/gistemp/updates_v3/

    ReplyDelete
  4. David Appell12:25 PM

    Turns out the newer data doesn't change my estimate for GISS's May anomaly.

    ReplyDelete