Precipitable water vapor and its relationship with the Standardized Precipitation Index: ground-based GPS measurements and reanalysis data

Abstract

Monthly means of ground-based GPS measurements of precipitable water vapor (PWV) from six stations in the USA covering the period January 2007–December 2012 are analyzed to investigate their usefulness for monitoring meteorological wet/dry spells. For this purpose, the relationship between PWV and the Standardized Precipitation Index (SPI) on 1-month timescale is investigated. The SPI time series at grid points close to the stations are computed using gridded precipitation records from the NOAA Climate Prediction Center (CPC) unified precipitation dataset (January 1948–April 2012). GPS measurements are first verified against PWV data taken from the latest ECMWF reanalysis ERA-Interim; these PWV reanalysis data, which extend back to 1979, are then used jointly with CPC precipitation to compute precipitation efficiency (PE), defined as the percentage of total water vapor content that falls onto the surface as measurable precipitation in a given time period. The overall results suggest that (i) PWV time series are dominated by the seasonal cycle with maximum values during summer months, (ii) the comparison between GPS and ERA-Interim PWV monthly data shows good agreement with differences less than 4 mm, (iii) at all stations and for almost all months, PWV is only poorly correlated with recorded precipitation and the SPI, while PE correlates highly with the SPI, providing an estimate of the water availability at a given location and useful information on wet/dry spell occurrence, and (iv) long data records would allow, for each month of the year, the identification of PE thresholds associated with different SPI classes that, in turn, have potential for forecasting meteorological wet/dry spells. Thus, it is through PE that ground-based GPS measurements appear of relevance for assessing wet/dry spells, although there is not a direct relationship between PWV and SPI.