A statistical comparison of cirrus particle size distributions measured using the 2-D stereo probe during the TC4, SPARTICUS, and MACPEX flight campaigns with historical cirrus datasets

摘要

This paper addresses two straightforward questions. First, howsimilar are the statistics of cirrus particle size distribution (PSD)datasets collected using the Two-Dimensional Stereo (2D-S) probe to cirrus PSD datasetscollected using older Particle Measuring Systems (PMS) 2-D Cloud (2DC) and 2-DPrecipitation (2DP) probes? Second, how similar are the datasets whenshatter-correcting post-processing is applied to the 2DC datasets? To answerthese questions, a database of measured and parameterized cirrus PSDs –constructed from measurements taken during the Small Particles in Cirrus(SPARTICUS); Mid-latitude Airborne Cirrus Properties Experiment (MACPEX); andTropical Composition, Cloud, and Climate Coupling (TC) flight campaigns– is used.Bulk cloud quantities are computed from the 2D-S database in three ways:first, directly from the 2D-S data; second, by applying the 2D-S data to icePSD parameterizations developed using sets of cirrus measurements collectedusing the older PMS probes; and third, by applying the 2D-S data to asimilar parameterization developed using the 2D-S data themselves. This is doneso that measurements of the same cloud volumes by parameterized versions ofthe 2DC and 2D-S can be compared with one another. It is thereby seen – giventhe same cloud field and given the same assumptions concerning ice crystalcross-sectional area, density, and radar cross section – that theparameterized 2D-S and the parameterized 2DC predict similar distributionsof inferred shortwave extinction coefficient, ice water content, and 94 GHzradar reflectivity. However, the parameterization of the 2DC based onuncorrected data predicts a statistically significantly higher number of totalice crystals and a larger ratio of small ice crystals to large ice crystalsthan does the parameterized 2D-S. The 2DC parameterization based onshatter-corrected data also predicts statistically different numbers of icecrystals than does the parameterized 2D-S, but the comparison between thetwo is nevertheless more favorable. It is concluded that the older datasetscontinue to be useful for scientific purposes, with certain caveats, andthat continuing field investigations of cirrus with more modern probes isdesirable.