Detectability of Falling Snow with NASA CloudSat Satellite: Evaluation from NOAA/NSSL Ground-Radar-Based National Multi-sensor Mosaic QPE (NMQ) System

Snowfall represents a predominant portion of precipitation at mid- and high-latitude regions and greatly contributes to regional atmospheric and terrestrial water budgets. Accurate detection and estimation of snowfall is desirable by various applications in meteorology, hydrology, and climatology.

Nowadays, remote sensing of precipitation has been a major approach for the monitoring of regional and global precipitation. A contemporary instrument is the sate-of-the-art Cloud Profiling Radar (CPR), which is the first spaceborne cloud radar onboard NASA's CloudSat satellite (http://cloudsat.atmos.colostate.edu/). CPR works at W-band (94 GHz), which provides good sensitivity for measuring the vertical structure of cloud liquid/solid water distribution. Combined with CALIOP (Cloud-Aerosol Lidar with Orthogonal Polarization) that is onboard NASA's CALIPSO satellite, CPR has proven to be capable of identifying and retrieving the snowfall.

Verification, refinement, and integration of spaceborne snowfall products require trustworthy ground-based dataset. Routine observations of snowfall have so far mostly been restricted 
to limited stations, with spotty spatial distribution and inconsistent duration of data record. The National Mosaic and Multi-sensor QPE (NMQ or Q2: http://nmq.ou.edu) system, developed by NOAA/NSSL and University of Oklahoma (OU), provides CONUS-wide high-resolution (5min, 1 km) QPE products, including the detection of falling snow. With appropriate data quality control by the radar quality index (RQI), NMQ/Q2 is regarded as an ideal, independent source for the validation of spaceborne products.

The current study evaluates CloudSat-CPR's detectability of falling snow using NMQ-Q2 snowfall products (i.e., solid snowfall precipitation identification) over the CONUS. We have collected all the CloudSat geometric profile data (2B-GEOPROF) and snowfall-profiling data (2C-SNOW-PROFILE) since the launch of CloudSat in 2006. Considering the difference in spatiotemporal resolution and grid consistency, we have also applied suitable interpolation and downscale methods to match CloudSat and Q2 data pairs. The evaluation results show the great potential of W-band cloud radar in detecting the falling snow. The detectability is also affected by the storm type and precipitation intensity. Further enhancement on the snow detection can be expected by incorporating the ground-radar-based NMQ products into spaceborne cloud radar observations.