The amount of precipitable water vapor can be calculated from observations in two channels 870 and 905 nm, where only one (905 nm) is sensitive to water vapor absorption. As with other wavelength-pair analysis, radiative-transfer analysis (LOWTRAN) is used to generate tables (illustrated in Figure 19) for W, equal to the ratio
of measured radiances at the Triana observing angles.
The precipitable-water tables are directly based on analysis done for the MODIS instrument onboard the Terra satellite scheduled for launch in the near future (Kaufman and Gao, 1992). The method was successfully applied to data obtained from AVIRIS (Airborne Visible Infrared Imaging Spectrometer), used as a MODIS simulator.
Triana-EPIC will be able to see clouds form and dissipate against a background of water vapor, thereby showing atmospheric modelers the processes that they cannot see at present. This will lead to an improved representation of clouds and cloud formation in general circulation models.
Since the water vapor measurement using W depends on backscatter of sunlight, it is able to detect total column water vapor. Infrared sounders depend on thermal contrast, and cannot give information on water vapor down to the planetary boundary layer. Yet much more water vapor is contained within this low-altitude layer, per millibar, than the layers above it. The EPIC I905 / I870 nm will give the only complete sunrise to sunset water-vapor data obtained from space, and will be able to match up with a similar once per day measurement from MODIS on the polar orbiting Terra satellite. With measurements of water vapor throughout each day, we can improve our estimates of latent heat transport, and improve our understanding of climate. A regional application of the effect of water vapor on the radiative forcing of dust aerosols has been discussed by Hsu et al. (1999a).
A further novel application to cloud height determination can be made with the total water vapor measurements. If the scene is cloud filled to at least a cloud fraction of 0.5, then the observed amount of total water vapor is greatly reduced since there is a large altitude gradient for water-vapor content above the Earth‚s surface. When the water vapor measurements from AVIRIS above a cloud are compared with AVIRIS cloud-top temperature measurements (made in the 12 µm infrared), there is a very good correlation as shown in Figure 20.
The Triana estimates of total precipitable water over cloud filled scenes will be used to estimate cloud heights and compared with the same scenes observed in the solar Fraunhofer line channel (393.5 nm). Cloud heights determined by the two methods (water vapor and the Ring effect from Raman scattering) will be compared. Cloud top temperature is a standard technique used to estimate cloud height, and is the basis for the ISSCP cloud height climatology database determined from AVHRR and GEO satellite data.
Validation and calibration of the two methods of cloud-height determination will be made by comparing with infrared temperature based determinations using matched scenes from MODIS, AVHRR, and GEO imagers.
