In article ,
Steve Loft writes:
This topic has come up a few times now, and Metcheck have now started
doing them, for example (you might need to select an area if you
haven't looked at it before):

http://www.metcheck.com/V40/UK/FREE/...ility_rain.asp

Could someone who understands this sort of thing confirm to this
neophyte that what they're saying is nonsense? Basically, they take the
10 GFS ensembles, and each one contributes 10% to the probability. So,
for Moray for Thursday morning next week - seven days away - all the
ensembles are currently saying it's going to be dry, so Metcheck's
probability of it being dry next Thursday morning is 100%. So it's an
absolutely certainty that it's going to be dry here next Thursday
morning? Is this tosh, or is this some other use of 100% that I'm not
familiar with?

Sounds like nonsense. Apart from the unreliability of the GFS 7 days
out, which makes the whole thing a nonsense, there is also the point
that the operational run as I understand it uses a finer grid than the
other ensemble members and therefore should be given a somewhat higher
weighting.

They also seem to be using the term tercile for their 10 bands of
possible outcomes (Dry, 1-2mm, 2-3mm etc). I thought a tercile was one-
third of something. More technobabble?

If they have ten bands, then those should presumably be deciles.
--
John Hall
"Honest criticism is hard to take,
particularly from a relative, a friend,
an acquaintance, or a stranger." Franklin P Jones