It was not so easy as I thought it might be, to capture the start of
the negative vibes in the North Atlantic. My method of forecasting the
weather from the time of the phases of the moon seems to have broken
down toward the third week in June -if not earlier.

I trolled throught the UKMO FAX at WetterZentrale (Archiv der 00 UTC
UKMO-Bracknell-Bodenanalysen (ab 27.01.1998)
http://www.wetterzentrale.de/topkarten/tkfaxbraar.htm ) to see if there
was anything like a negative NAO around then and there doesn't at first
glance appear to be anything indicative for a few more days, with a
relatively deep Low of 985 deepening mid Atlantic by the 29th June,
replacing (or at least, pushing it off the map) a relatively high
anticyclone of 1035.

Whilst it all gets rather tight on the 7th July on, it is still very
much a positive NAO with a nice warm Low of 978 appearing off Greenland
on the 11th.

I seem to recall that something struck me as different about the lunar
phases for June, but I can't remember what it was:

JUNE 18 14:08 should have been sunny with not much cloud.
JUNE 25 16:05 should have been wet.
JULY 3 16:37. This is the one that threw me IIRC.
JULY 11 3:02 and this should have been a classically thundery one with
a positive NAO.

A month later:
By the 18th, a pattern of sending cyclones through the Norwegian Sea up
toward the Berents Sea seems to have become adopted. And the Low
pressures from then on are fairly high in the NA except for a blip over
Newfoundland. And even that seemed to push off into the Arctic by the
23rd.

I suppose it's a case of "Keep zeigen and ye shall find."