Thermodynamic Atmospheric Profiling During the 2010 Winter Olympics Using Ground-Based Microwave Radiometry

• Published in: IEEE transactions on geoscience and remote sensing 2011-12, Vol.49 (12), p.4959-4969
• Main Authors: Cimini, D., Campos, E., Ware, R., Albers, S., Giuliani, G., Oreamuno, J., Joe, P., Koch, S. E., Cober, S., Westwater, E.
• Format: Article
• Language: English
• Subjects: Applied geophysics; Atmospheric measurements; Bayesian variational methods; Clouds; Earth sciences; Earth, ocean, space; Exact sciences and technology; Humidity; Internal geophysics; Meteorological satellites; Meteorology; Microwave radiometry; Ocean temperature; radiometry; Temperature measurement; Temperature sensors
• ISSN: 0196-2892; 1558-0644
• DOI: 10.1109/TGRS.2011.2154337

Ground-based microwave radiometer profilers in the 20-60-GHz range operate continuously at numerous sites in different climate regions. Recent work suggests that a 1-D variational (1-DVAR) technique, coupling radiometric observations with outputs from a numerical weather prediction model, may outperform traditional retrieval methods for temperature and humidity profiling. The 1-DVAR technique is applied here to observations from a commercially available microwave radiometer deployed at Whistler, British Columbia, which was operated by Environment Canada to support nowcasting and short-term weather forecasting during the Vancouver 2010 Winter Olympic and Paralympic Winter Games. The analysis period included rain, sleet, and snow events (~235-mm total accumulation and rates up to 18 mm/h). The 1-DVAR method is applied "quasi-operationally," i.e., as it could have been applied in real time, as no data were culled. The 1-DVAR-achieved accuracy has been evaluated by using simultaneous radiosonde and ceilometer observations as reference. For atmospheric profiling from the surface to 10 km, we obtain retrieval errors within 1.5 K for temperature and 0.5 g/m 3 for water vapor density. The retrieval accuracy for column-integrated water vapor is 0.8 kg\m 2 , with small bias (-0.1 kg\m 2 ) and excellent correlation (0.96). The retrieval of cloud properties shows a high probability of detection of cloud/no cloud (0.8/0.9, respectively), low false-alarm ratio (0.1), and cloud-base height estimate error within ~0.60 km.