OBD II - fuel trims (adaptives)

[Joe Chiasson]

If any one could unravel the mystery behind OBD II short and long term fuel
trims (adaptives) it would be greatly appreciated.  My interpretation was
that short term fuel trims dealt with the short term fuel delivery (i.e.
computer increasing or reducing pulsewidth to reach a stoich O2 sensor
reading). My interpretation of long term fuel trims is how out of whack the
fuel delivery to the engine is, either over an OBD II driving cycle or over
the entire driving life of the vehicle, until the ecm is diconnected that
is.  

This is what I thought (now please correct me if I am wrong) untill I ran
into this senario with a 1997 Ford F-150 automatic 4.6 L V-8.  Under
cruising speeds of 55 mph the short term fuel trim was +5% and the long
term fuel -7%.  Under WOT conditions the short term fuel trim was +33% and
the long term -21%.  Now under my intial interpretations on how fuel trims
were calculated I assumed that at WOT the ecm was increasing the fuel
delivery by 33% to reach stoich.  But then why would the long term fuel
trim decrease?  Well this has confused the hell out of me.

If anybody can clue me on on what the fuel trims actually mean, and how
they are determined (i.e. referenced off the front O2, the rear O2, or no
O2 at all)  help would be appreciated.

J.