Comments on: Help Robert Rohde Locate Argentina

By: Earle Williams

Earle Williams — Fri, 04 Nov 2011 23:38:38 +0000

Peter,

‘props’ is an abbreviated form of ‘proper respect.’ The ‘mad’ modifier indicates that the recipient is due excessively proper respect.

By: Geoff Sherrington

Geoff Sherrington — Thu, 03 Nov 2011 10:32:31 +0000

Sorry Simon, my error, wrong line of spreadsheet. Try Point Lonsdale. -38.3 144.62

By: Jeff Wood

Jeff Wood — Wed, 02 Nov 2011 23:04:43 +0000

Sinan, many thanks for that. Goodness knows what I was doing last year to miss it.

We laymen appreciate that sort of effort. You elucidate what gives us a headache when trying to understand difficult and technical descriptions.

The later frames give a sad sense of something dying. I am relieved that the stations are mostly still live. Proper science will insist on their inclusion in future datasets.

By: LearDog

LearDog — Wed, 02 Nov 2011 22:56:23 +0000

Ha ha ha! ROFL..! Ha ha ha!

By: kuhnkat

kuhnkat — Wed, 02 Nov 2011 19:15:08 +0000

Maybe one with the other inserted?

By: John Whitman

John Whitman — Wed, 02 Nov 2011 16:08:54 +0000

Steve,

Hilarious. Even more so because it is not a joke.

The humor is like twisting the knife a little bit after your initial bold thrust at their metadata?

Belated Halloween cheers.

John

By: Simon Greener

Simon Greener — Wed, 02 Nov 2011 06:36:47 +0000

Put -41.07,145.95 into Google Maps on the iPad and it lands on Latteene Rd, Round Hill, just south of Burnie. Nothing wrong with the long/lat. Sen plenty of point objects inthe oceans around Tasmania over the years (transposition of numbers) but not in this case.
Simon

By: Peter O'Neill

Peter O'Neill — Wed, 02 Nov 2011 03:31:49 +0000

mad props to him

unfortunately I need a translation. This looks like predictive text gone wrong!

By: Geoff Sherrington

Geoff Sherrington — Wed, 02 Nov 2011 03:24:33 +0000

In April 2011 I sent Steven Mosher a corrected spreadsheet for 589 Australia stations that presumably originated from BEST. Of these, 32 had problems of location, name or no available data, e.g
91009 50194958001 -41.07 145.95 BURNIE (ROUND HILL) Error- station coords at sea. That is 543 ppm eror rate from these causes alone. At the time I thought that Steven was working with BEST, but I now know that he was not, so this is not a criticism of him. I simply make the point that cleaned data were prepared over many, many hours for Australia. I do not know if BEST used them.

By: Peter O'Neill

Peter O'Neill — Wed, 02 Nov 2011 03:22:42 +0000

Resurfacing shortly. Have been working on quantifying the effect on GISTEMP of station location errors in the GHCN metadata (and less than impressed to find similar location errors in the BEST metadata, as uploaded by Nick Stokes). I had submitted a comment on Hansen et al (2010) to Reviews of Geophysics, but it turns out that Reviews of Geophysics publishes reviews only, and does not accept comments or replies, and a stand-alone paper submission to GRL or JGR, quantifying the size of the impact of the corrections was suggested. The unfortunate side effect of this AGU policy is of course that a review describing a method based on metadata containing numerous substantial errors remains without comment drawing attention to the existence of such errors until such time that the impact of the corrections has been quantified, a non-trivial task in view of the number of stations involved. As resolution claimed

"Station location in the meteorological data records is provided with a resolution of 0.01 degrees of latitude and longitude"

is not matched by the actual metadata used, and as a result the urban/rural classification derived is substantially in error, a reader surely deserves to be made aware of this even before quantification of the impact of the corrections required becomes available. Some mention of these metadata problems is long overdue at the GISTEMP "Updates to analysis" page as well - Hansen et al are aware of these metadata shortcomings. I first brought a few such errors to their attention in December 2009, before the adoption of nightlights as the basis of classification, and suggested

It looks to me as if changing to global lights is likely to introduce as many new misclassified stations as it corrects, and that the only way this classification can really be improved is by actually examining each location, painful as that exercise may be.

(and painful indeed it proves to be: if you are aware of any GHCN station location corrections, perhaps you might check my page at http://oneillp.wordpress.com/2011/08/20/are-you-acquainted-with-any-of-these-ghcn-stations/ to see if the stations in question are listed there) I hope to post a related comment later today at clearclimatecode.org/just-440-stations, which may be of interest.