The Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report found that changes in extreme events have occurred and the frequency of such events is expected to increase. Precipitable water vapor (PWV) is a useful measure of the moisture content of the atmosphere. This paper combines the predicted GCM trends in PWV from 2000 to 2100 with uncertainty estimates from infrared spectrometers, NASA Atmospheric Infrared Sounder (AIRS) and EUMETSAT Infrared Atmospheric Sounding Interferometer (IASI), to estimate minimum trend detection times on regional and global spatial scales. The minimum detection time (MDT) is the number of years before the multimodel GCM trend exceeds a fractional change equal to the uncertainty in the observed product, plus the width of the time window used to smooth out natural variability. Results indicate that the median value of PWV has an MDT of 15 yr or less over all scales, while extreme dry (5th) and wet (95th) PWV conditions (percentiles) have higher measurement uncertainty and corresponding larger MDTs. Product providers have done a relatively good job tuning results to the mean atmospheric state but more attention should be given to improving the satellite estimates for extreme PWV. A fractional measurement error of 3% is desirable to detect predicted climate trends within 15 years or less for the entire PDF of PWV. This paper presents an important case study for the design of observing systems directly linking the estimated uncertainty of the PWV products to the detectability of long-term trends. If there is a need to decrease detection times over the existing weather observation system then necessary changes to the climate observational system design can be understood quantitatively.
All Science Journal Classification (ASJC) codes
- Atmospheric Science