Abstract
Actual data are used in one case to investigate the nature and source of vertical velocity errors resulting from analyses of multiple-Doppler radar measurements. Consistent with earlier analytical works, larger errors are found than would be expected from previous theoretical studies. It is shown that the reconstructed maximum updraft speed in strong updrafts (>20 m s−1) is accurate, on the average, to within about 10% (standard deviation of 10%). Storm advection, incomplete sampling of low-altitude divergence caused by the radar horizon, top boundary errors, and uneven terrain are studied and all are dismissed as dominant sources of error in the case considered here. The inability to determine a dominant error source has important consequences for the formulation of vertical velocity adjustment schemes.
