As a non-intrusive technique with high spatial and temporal resolution lase
r Doppler anemometry is widely used in the study of turbulent flow fields.
The computation of statistical functions of the measured velocity time seri
es presents problems since, by nature of the technique, the data is stochas
tically sampled in time. After a brief review of the LDA measurement princi
ples the essential properties of LDA data sets will be explained and a revi
ew of the methods to handle the special sampling features of the data will
be given. Recent developments will be presented for the estimation of turbu
lence spectra.