The authors have compared isentropic trajectories computed from meteor
ological fields from different analysis centers. The analysis was perf
ormed for the South Atlantic, where a springtime maximum in tropospher
ic ozone has sparked considerable interest in the transport meteorolog
y. Using the model of Schoeberl et al., isentropic forward trajectorie
s were computed from an array of points over southern Africa and backw
ard trajectories from an array of points over the South Atlantic. The
model was run for an 8-day period in October 1989 with input taken fro
m the twice-daily global gridded data fields from the National Meteoro
logical Center (NMC) and from the European Centre for Medium-Range Wea
ther Forecasts (ECMWF). There were large differences between the traje
ctories based on the two fields in terms of travel distance, horizonta
l separation, and vertical separation. Best comparisons for individual
trajectories were found in the low-latitude easterlies, and the poore
st comparisons were found in the westerlies and in the vicinity of the
center of the South Atlantic subtropical anticyclone. Significant dif
ferences in wind speeds between the two analyses also led to large tra
jectory differences. Trajectories were also computed using once-daily
NMC fields. The effect of this degradation of the data was small. Traj
ectories computed from balanced winds computed from the NMC geopotenti
al height and temperature fields showed the largest differences when c
ompared with the ECMWF trajectories. The balanced wind fields should n
ot be used in trajectory construction in the tropical lower tropospher
e. It is difficult to make a definitive recommendation concerning whic
h set of fields should be used in future transport analyses in this re
gion due to the very large trajectory differences found in this analys
is and the lack of any independent verification data. Any extensive an
alysis of transport in this region should be done only in conjunction
with considerable additional data collection.