A Representation of the Close Tie Between Water Vapor, Clouds, and Climate
This figure from Miloshevich et al.  shows altitude profiles of relative humidity (RH) as measured by 3 balloon-borne instruments: 2 Vaisala RS80-A radiosondes (solid) and the reference-quality NOAA cryogenic frostpoint hygrometer (dashed). Superimposed on the hygrometer profile are representative ice crystals collected at the same altitude by the NCAR balloon-borne ice crystal replicator. All were flown on the same balloon on 10 Nov 1994 near Boulder, Colorado. The ice-saturation curve (RHi) separates the region where ice crystals are stable or will grow (RH>RHi) from the region where ice crystals will sublimate and eventually disappear (RH<RHi). The ice crystals growing in ice-supersaturated air in the cloud top region (Z>10.5 km) are smaller and have pristine shapes and sharp edges, whereas the larger crystals become increasingly rounded as they fall through ice-subsaturated air and finally sublimate completely at the cloud base. Upper tropospheric water vapor and cirrus clouds have a significant impact on the Earth's radiation balance by virtue of their low temperature and high frequency of occurrence, and are therefore important to climate research.
The very large dry bias exhibited by the RS80-A radiosondes, the most widely used operational radiosonde in the world at the time, is clearly not consistent with the presence of a cloud, and is far from accurate enough for cloud physics, climate, or water vapor remote sensing research, unlike the much more accurate frostpoint hygrometer. However, analysis of a dataset of hygrometer and RS80-A dual soundings showed that the bias is calibration-related and is a function of temperature and RH, therefore an empirical correction could be developed.