SNOW

Measurement of ice-phase precipitation in the Arctic remains a challenge (e.g. Goodison et al. 1998, Benning and Yang 2005). Wind, drifting snow, and the accumulation of snow on and behind obstacles all combine to make quantitative winter-time precipitation measurements very difficult. The traditional method of measuring precipitation with the simple bucket gauge suffers from systematic bias because the gauge disrupts the normal flow of snow. This problem is shared by even the most sophisticated shielded gauges in the climate reference network (CRN; http://www.ncdc.noaa.gov/crn/). This flow disruption results in an underestimation of the total snow rate. In order to improve on precipitation estimates, simple gauge measurements have been compared to a well-tested reference gauge and precipitation pits. Regression function correction factors based on meteorological parameters, such as wind speed and temperature, are applied to correct gauge errors of up to 80-120% (Yang et al. 1998, 2000, 2005). Because measuring snowfall rates with gauges remains problematic, other methods of estimating snow fall rates are used. These include direct measurements such as the hot plate Total Precipitation Sensor (TSP; Cherry et al. 2009) deployed in Barrow and disdrometers (Brandes et al. 2007), or combination remote sensing/direct measurements (Xie and Arkin 1997, Huffman et al. 1997, Huang et al. 2010).

The lack of accurate ice precipitation measurements has been identified as an impediment to efficient improvement of climate models (i.e. in the ARM program see Xie et al. 2006). As a result, improving measurements of frozen precipitation is a priority for several multi-agency programs including the Global Energy and Water Cycle Experiment (GEWEX, http://www.gewex.org/), Study of Environmental Arctic Change (SEARCH) (Morison et al. 2001), the Northern Eurasia Earth Science Partnership Initiative (NEESPI) (http://neespi.gsfc.nasa.gov/science/science.html), and the Arctic Community Wide Hydrologic Analysis and Monitoring Program (http://arcticchamp.sr.unh.edu/).

The acquisition and deployment of the new X- and Ka/W-band radars in Barrow opens up an opportunity for ARM to obtain spatial estimates of snowfall rates using the polarimetric X-band measurements and dual-frequency measurements (using different combinations of the three wavelengths). However, calculations of X- and Ka-band radar back-scattering of ice crystal aggregates with their complex structure, suggest that the commonly used T-matrix approach (Matrosov et al. 2007) for modeling the radar back-scattering underestimates the reflectivity by several decibels, with errors increasing with increasing radar frequency (Botta et al. 2010, 2011). Moreover, the X-band polarimetric measurements and the Ka/W-band measurements are sensitive to the assumed shape of the snow (Botta et al. 2011).

A preliminary investigation of ice crystal aggregate shapes approximated as ellipsoids was conducted using 24,000+ individual aggregate images collected with 2-D video disdrometers on 39 days. These data included snow events in Boulder, CO, Egbert, Ontario and Helsinki, Finland. These results combined with the Botta et al. (2010, 2011) simulations suggest that combined radar/disdrometer measurements will be needed to make progress on radar estimation of snowfall rates.

Timeline

Campaign Data Sets