GaryM
“In a nutshell Kriging does attempt to take into account nearby values (observed) when estimating a value at a given location.”
But you have to pay attention to the whole statistical house of cards. “Nearby” is often defined as 750 miles.
The quoted sentence above applies to kriging and is not specific or restricted to the BEST implementation. Above* you wrote:
So surrounding topography, proximity to lakes, forests, and mountains, and susceptibility to impacts of various oceanic oscillations, are irrelevant to your model? Or its laughably inflated claims to precision?
* http://judithcurry.com/2015/08/25/jcs-conscience/#comment-727491
These factors affect other stations in that locale. Kriging weights closer location more than distance. Kriging is one attempt at incorporating Tobler’s Law into the estimation. Kriging is an estimation technique. Your comment suggests to me that either you still somewhat unfamiliar with the conceptual basis of the technique you are criticizing or choose to ignore it just to chew on a leg. As for 750 miles and greater: given coverage in some areas who is surprised? That aspect is of course subject to criticisms and comment, yes that affects the quality of estimates. However, were I you I would raise my game—maybe dig into the analyses and in particular the error analyses. As for BEST, the fact is one does not work for hours and hours on such tasks and not notice issues and on a number of occasion it has been noted that there are priorities.
As for But you have to pay attention to the whole statistical house of cards. That is a pretty vague comment. Just what is your ‘statistical house of cards’ here? That is, for you what statistical sensibilities are being tweaked?
Somewhere up above I made a comment showing differences of several degrees on average yearly temperature (which is the CAGW standard – “warmest year ever!!!!”). I assume here you are referring to the comparison between locations with similar latitudes and how they illustrate local and regional influences. Well, again kriging preferentially weights/weighs nearer observations (Tobler’s Law). BTW Steven adequately caveated his discussion up front:
I will keep it simple imagine the function looks like …
Monthly, annual? Ha! Give me a break! Better yet, just what is your point? Just messin’ with Steven or an important criticism?
I could care less about ‘record’ or ‘no record’. For reasons well beyond BEST, C&W, etc., etc., etc. that ‘highest recorded temperature meme IMO is total silliness that I generally ignore in the media. Don’t waste your time with it.
Mosher’s comments sound great, if you ignore their intended purpose. Which is of course his obscurantist point.
Absent language cues I am not as quick to attribute purpose to Steven’s and others remarks. Where others may look at the remarks looking for alarm I try to look for caveats and technical aspects. One also has to look at context.