Land surface schemes are vital components of general circulation models (GC
Ms) which provide the fluxes of heat, water and momentum at the land-atmosp
here interface. The fluxes simulated by these schemes are especially depend
ent on the way in which the canopy (or 'bulk stomatal') conductance for pla
nt transpiration is modelled. Considerable research has been carried out in
to the dependences of canopy conductance on the local environment, and empi
rical relationships for such dependences have been obtained by fitting the
data collected in field and laboratory experiments. However, observed leaf
level relationships between stomatal conductance and net photosynthesis sug
gest an alternative approach. Given an appropriate algorithm for scaling th
ese values up to canopy level, such relationships allow canopy conductance
values to be derived from (comparatively) well validated models of leaf pho
tosynthesis. This approach is likely to become especially attractive as lan
d surface schemes are extended to simulate CO2 fluxes, since the shared env
ironmental dependences of canopy conductance and photosynthesis reduce the
number of model parameters which require independent specification. This pa
per is concerned with the evaluation of canopy level relationships between
photosynthesis and conductance using data from the first international sate
llite land surface climatology project (ISLSCP) held experiment (FIFE). Sim
ultaneous measurements of CO2 and water vapour fluxes, taken over a predomi
nantly C-4 grassland, are used to independently test the conductance and ph
otosynthesis models. Based on these tests, a simple coupled model of canopy
conductance and photosynthesis is developed which produces a good match to
the experimental data. (C) 1998 Elsevier Science B.V. All rights reserved.