Model rain evaporating

Whining here a bit….

These are sad times when you can’t find a model run with at least some AZ rain in it.  The last two USA model runs, compiled from global data taken at 00 Z (5 PM AST) and 06 Z (11 PM AST)  don’t have ANY rain in Arizona.   Prior two runs did, ones that came out during the day yesterday.

The rains in yesterday’s model outputs were late in the month and into early December.  I exulted all day yesterday because the “ensembles of spaghetti” were strongly suggesting those rains were associated with “outlier” model runs, more extreme ones that couldn’t be counted on.

But, those rains kept reappearing.  I unilaterally decided, i.e., had a “hunch”, that the “ensembles” themselves were, in a sense, “outliers”, lying about no rain in AZ.  They were missing something.

Looks like I’ll get burned on that “hunch”, as would be expected by an objective scientist. Now I will talk about something else…

Maybe the costly and more accurate Euro model, which could save thousands of lives by being free and warning poor people of big storms and winter cold way in advance,  has rain here in it here, but I am too poor to afford to look at it.  The Canadian Environment Canada model1, built on the Euro one,  only goes out to 144 hours, not the 5000 hours ahead that we seem to need to see a rain predicted in Arizona someday.


BTW, as an example of that Euro model’s non-availability, if you go to the University of Washington’s model outputs web page, you will see the daunting words for the Euro-UK MET model,  “restricted.”  The letters are in RED to make sure you know you can’t get it.  Don’t hit the latest run link to the right, you will be asked to enter your ID and password, and because you don’t have one, you are punished by not being able to go back to where you started.   Rather, you end up in an endless loop asking for your password and ID to make sure you don’t come back again and mess around thinking you might “get in.”  You will have to close your browser.  If you have a password and ID, then post that model somewhere, oh, man, you are in for fines and, who knows, maybe some jail time.

Its a sad forecasting world out there.


Here are the sad (lotta “sad” today) conflicting rain/no rain model outputs for the SAME time and day, the first panel from yesterday morning when I was happy, and the second panel below, from the very latest run:

Rainy AZ prediction for November 28th based on yesterday’s (11-19) 5 AM AST global data.  See green pixelation in Arizona; that’s model predicted rain.  Was happy to see this late yesterday morning.
18 h later, the model run from last night’s data at 11 PM AST also valid for November 28th.  Look at that huge dry region in the whole West that has replaced a substantial storm! Unbelievable.  There a friggin’ high pressure center almost exactly where there was a big low pressure center!  We call this, “Forecasting hell.”



On to happier things, should have some gorgeous clouds and skies today as a tropical system skirts SE AZ.  Virga likely.  Keep cameras ready for a great sunset.


The End.




Why would they name their organisation, “Environment Canada” when you already have to go to a Canadian site to look at it?