New paper compares water vapor measurements from satellite instruments and GPS receivers

Date Added: 
Thursday, June 18, 2015


RSS Scientists, in collaboration with Junhong Wang from SUNY Albany, have written a paper describing a comparison of total column water vapor measurements made by satellite microwave radiometers and ground-based GPS receivers. The paper is titled "Intercomparison of total precipitable water measurements made by satellite-borne microwave radiometers and ground-based GPS instruments" and is  published in the Journal of Geophysical Research: Atmospheres, vol 120, 2492-2504.


What is in the paper?

The paper compares measurements of total column water vapor (or total precipitable water) measurements made by two very different observing systems.

  1. Microwave radiometers on satellites, such as SSM/I, SSMIS, AMSRE and WindSat can measure total column water vapor.  This is the method used for all RSS water vapor products.
  2. Global Positioning System (GPS) instruments are typically used to deduce location based on very precise measurements of the time required for a radio signal to propagate from a GPS satellite to the receiver.  If measurements made by a stationary GPS receiver are analyzed, the small additional delays due to water vapor can be used to deduce the amount of water vapor in the atmosphere.  This method is not sensitive to things that can cause problems with the microwave radiometer method, such as the presence of rain or calibration errors in the satellite instruments.  GPS measurements of total column vapor are thought to be absolutely accurate to within 1.0kg/m^2 (or 1 mm liquid water equivalent).

These two types of measurements were compared and were found to agree very well in general.  A close look at the data revealed minor problems in both types of measurement systems.  Several GPS stations showed small jumps in bias, presumably due to changes in instrumentation or observing practices over time, and there are small differences in bias from station to station that are difficult to assign to the satellite measurements.  The satellite data showed a small error as a function of wind speed that had not been previously detected, and in some cases, errors as a function of rain rate were identified.

Why is it important?

  • The excellent agreement between the two datasets adds to the credibility of both.
  • We showed that satellites are a valuable tool for finding small problems in GPS data, and vice-versa.
  • We found that neither system showed any evidence of long-term drifts.