Airborne estimates of precipitation during the Mixed-Phase Arctic Cloud Experiment

Realistic modeling of the Arctic radiation budget depends in part on the accurate portrayal of cloud physical properties, including precipitation rate. Atmospheric models need to keep track of the mass that is removed from the clouds as precipitation, and how much of that precipitation reaches the ground or remains in the atmosphere as water vapor. This study looks at snowfall rates derived from airborne measurements during the Mixed-Phase Arctic Cloud Experiment (MPACE) in 2004. It also compares these rates to those measured at the surface from the National Weather Service at Barrow, AK, and those predicted from model output.

During MPACE, there were three different synoptic regimes observed, with multi-layer and single-layer cloud conditions occurring during the second regime, under consideration here. Multiple cloud layers were present from October 4 through early October 8, while a single cloud layer persisted from late on the October 8 through October 12. Airborne measurements of ice-water content, particle size, and crystal habit from the University of North Dakota Citation aircraft were used to compute the snowfall rate. Only flight segments from below cloud base were used for these estimates. For multi-layer cloud days, the average snowfall rate derived at cloud base was 2.6 mm d^-1, but only 0.7 mm d^-1 measured at the surface. Surface rates predicted by a variety of single column and cloud resolving models ranged from 0.3 to 1.6 mm d^-1. The single-layer cloud snowfall rate averages were 4.1 mm d^-1 at cloud base, 0.25 mm d^-1 at the surface, and 0.4 to 0.7 mm d^-1 from the models.

The finding that surface precipitation rates were less than cloud base rates was expected, due to sublimation of falling ice particles. Theoretical estimates of sublimation losses were obtained using sub-cloud profiles of relative humidity and found to be less than the measured fractional decrease in precipitation rate. Measurement uncertainties for both the airborne and surface values are discussed relative to these differences. Snowfall rates for single and multi-cloud days were also compared; particle size distributions and sublimation effects were examined for possible contribution to differences in estimates of these rates. While particle size distributions were found to be similar on single and multi-layer days, there were significant differences in the sub-cloud relative humidity profiles for single (relatively dry) and multi-layer conditions. Finally, the measured snowfall rates are compared to the model calculations reported for these cloud conditions as a check on model performance.