New Technique Improves Cirrus Cloud Characterization

NASA’s Jet Propulsion Laboratory

A new technique for retrieving cirrus properties from radiometric measurements at submillimeter wavelengths has been developed. The technique can accurately measure the amount of ice present in cirrus clouds, determine the median crystal size, and constrain crystal shape. The retrieval algorithm improves upon prior algorithms by also retrieving middle and upper tropospheric water-vapor profiles in concert with cloud properties. This joint-analysis method corrects for retrieval errors introduced by water vapor in and near the cloud.

Errors Associated With Retrieval of IWP and Dme are significantly reduced with the new retrieval technique.

Submillimeter-wave cloud ice radiometry is a relatively new technique. In 1995, two theoretical papers were published describing the use of radiometry to characterize ice clouds. These studies indicated that cirrus clouds scatter the upwelling flux of submillimeter-wavelength radiation emitted by lower atmospheric water vapor back towards the Earth, thus reducing the upward flux of energy. (In the submillimeter-wave spectral region, ice particles primarily scatter radiation rather than emitting or absorbing it.) From space, this effect makes clouds look radiatively cold against the warm emissions of water vapor in the lower troposphere. The ability of cirrus ice to scatter radiation is primarily a function of the amount of ice and the distribution of crystal sizes. Scattering induced by changes in crystal size is distinguished from scattering induced by changes in the total ice content, termed the ice water path (IWP), by making measurements at widely spaced frequencies. Additionally, crystal shape can be constrained by determining the crystal height-to-width aspect ratio, which is derived from off-nadir measurements at orthogonal polarization angles. Key assumptions underlying the theoretical predictions were validated by a set of airborne measurements in 1996.

The new retrieval algorithm corrects for middle and upper tropospheric water vapor that degrades retrieval accuracy via two mechanisms. First, water vapor emits radiation, reducing the apparent fraction of the radiation scattered by an underlying cloud. Second, water vapor absorbs radiation, also reducing the apparent scattered fraction. Thus, there is a need for a retrieval technique that corrects for these watervapor- induced screening effects. The new algorithm builds on previous work by simultaneously retrieving water-vapor profiles and cirrus properties.

A Bayesian algorithm is used to invert a mathematical model of the radiometric properties of both cloud ice and water vapor. The model is statistical in nature relying on a combination of an in-situ cirrus measurement database, assumptions about vertical cloud inhomogeneity, and estimates of cloud temperature. The in-situ cirrus database consists of measurements from four sets of field measurements including three sets taken over a tropical site (CEPEX) and one over a midlatitude, Midcontinent site (FIRE II). The assumptions about cloud inhomogeneity are based on the observed relationship between IWP and the median crystal diameter, Dme.

The accuracy of this method has been assessed in computational simulations using the complement of radiometric channel planned for a new airborne instrument, the submillimeter- wavelength cloud ice radiometer (SWCIR) currently being developed by JPL. The instrument will have the capability to make radiometric measurements at ten frequencies spanning from 183 to 643 GHz. The simulations have quantified the accuracy of expected cirrus retrievals and have also quantified improvements that could be expected with the addition of an 880-GHz channel. The table presents selected results from these simulations. These results illustrate the dramatic improvement in accuracy that is achievable with the new analysis technique.

This work was performed by Steven Walter of Caltech (now employed by Aerojet in Azusa, CA), and K. Franklin Evans and Aaron Evans at the University of Colorado for NASA’s Jet Propulsion Laboratory. For further information, access the Technical Support Package (TSP) free on-line at www.nasatech.com/tsp under the Physical Sciences category. NPO-21016