Simultaneous measurements of wind velocities at two different sites, o
ne over the sea and the other over land, can differ substantially and
therefore cannot be interchanged. In situations where the wind data at
an offshore site are missing while simultaneous measurements from a l
and-based station exist, a linear mean-square estimation (LMSE) techni
que can be used to estimate the missing data. This technique relies on
past wind data gathered simultaneously at the two locations, and it g
enerates from the associated correlation a set of four transfer functi
ons capable of predicting one data set from the other. In the present
case, the LMSE technique is outlined briefly, and is then applied to c
onstruct seasonal transfer functions between a land-based station and
two coastal/offshore sites in Kuwait. Comparisons between the actually
observed wind characteristics and those predicted by the LMSE techniq
ue are favorable, and thus tend to confirm the applicability of the te
chnique under appropriate conditions.