The composite characteristics of cirrus clouds: Bulk properties revealed by one year of continuous cloud radar data

The properties of midlatitude cirrus clouds are examined using one year of continuous vertically pointing millimeter-wave cloud radar data collected at the Atmospheric Radiation Measurement Program Southern Great Plains site in Oklahoma. The goal of this analysis is to present the cloud characteristics in a manner that will aid in the evaluation and improvement of cirrus parameterizations in large-scale models. Using a temperature- and radar reflectivity-based definition of cirrus, the occurrence frequency of cirrus, the vertical location and thickness of cirrus layers, and other fundamental statistics are examined. Also the bulk microphysical properties of optically thin cirrus layers that occur in isolation from other cloud layers are examined. During 1997, it is found that cirrus were present 22% of the time, had a mean layer thickness of 2.0 km, and were most likely to occur in the 8.5-10-km height range. On average, the cirrus clouds tended to be found in layers in which the synoptic-scale vertical velocity was weakly ascending. The mean synoptic-scale vertical motion in the upper troposphere as derived from Rapid Update Cycle model output was +0.2 cm s^-1. However, a significant fraction of the layers (33%) were found where the upper-tropospheric large-scale vertical velocity was clearly descending (w < -1.5 cm s^-1). Microphysical properties were computed for that subset of cirrus events that were optically thin (infrared emissivity < 0.85) and occurred with no lower cloud layers. This subset of cirrus had mean values of ice water path, effective radius, and ice crystal concentration of 8 g m^-2, 35 μm, and 100 L^-1, respectively. Although all the cloud properties demonstrated a high degree of variability during the period considered, the statistics of these properties were fairly steady throughout the annual cycle. Consistent with previous studies, it is found that the cloud microphysical properties appear to be strongly correlated to the cloud layer thickness and mean temperature. Use of these results for parameterization of cirrus properties in large-scale models is discussed.