Multiscale analysis of precipitable water vapor over Africa from GPS data and ECMWF analyses

Abstract

This is the first climatological analysis of precipitable water vapor (PWV) from GPS data over Africa. The data reveal significant modulations and variability in PWV over a broad range of temporal scales. GPS PWV estimates are compared to ECMWF reanalysis ERA40. Both datasets show good agreement at the larger scales (seasonal cycle and inter‐annual variability), driven by large scale moisture transport. At intra‐seasonal (15–40 days) and synoptic (3–10 days) scales, strong PWV modulations are observed from GPS, consistently with ECMWF analysis. They are shown to be correlated with convection and the passage of equatorial waves and African Easterly waves. The high‐frequency GPS observations also reveal a significant diurnal cycle in PWV, which magnitude and spectral content depends strongly on geographic location and shows a seasonal modulation. The diurnal cycle of PWV is poorly represented in ERA40 reflecting weaknesses in the water cycle of global circulation models at this timescale.