.... Be a bit wary about relating 850 hPa temperatures to expected
maxima at sea-level. They are often a good guide but you need to know
what the pitfalls are as well:-

.... Given an *average* lapse rate between msl and 5000ft .. (the
approx. 850 hPa level), of 2degC/1000ft, then this would imply a
near-surface temperature of around 10C higher than the 850 level, or
about 30degC in this case. Given strong summer heating (as in this
situation) and assuming a near Dry Adiabatic Lapse Rate at T(max),
then the lapse rate (assuming the atmosphere is heated uniformly to
the 5000ft level - see below), would be around 3degC/1000ft, so 3 * 5
= 15C, and for 20C at 850hPa, this indicates 35 degC - and local
heating of the lowest few metres would / could add something on to
this - in the early stages of the current 'heat wave', this will
probably be good guidance.

.... However, in stagnant anticyclonic situations (as this may become),
an inversion will form (or air with a low inversion will be advected
around the equator-ward side of the high), which could well descend/be
BELOW the 850 hPa level, and these relationships will fall apart as
the atmosphere below the inversion will behave as if it is isolated
from that at 850hPa and above. The 850hPa temperature will obviously
have some relevance, but there won't be a direct link.

.... See also the link below

http://www.booty.demon.co.uk/metinfo...opics.htm#850T

Martin.