2A.3 Quality of mobile observations collected during the 2010 Development Testbed Environment experiment

Monday, 24 January 2011: 4:30 PM
606 (Washington State Convention Center)
Amanda R. S. Anderson, NCAR, Boulder, CO; and M. B. Chapman, S. D. Drobot, A. Tadesse, and P. Pisano
Manuscript (399.2 kB)

The 2010 Development Testbed Environment (DTE10) experiment took place from 28 January 2010 to 29 March 2010 for the purpose of collecting and evaluating mobile data from vehicles. This included both meteorological and vehicle diagnostic data, which was used to assess the quality of the mobile observations and to develop algorithms for road weather hazard products. To examine the quality of the data, over 239,000 air temperature and barometric pressure observations from 9 vehicles were collected during the testing period. These observations were examined and compared to a weather station set up at the testing site. Air temperature and barometric pressure were the focus of the quality assessment because the weather station provided a ground truth with which to compare. The observations from the vehicles were run through the National Center for Atmospheric Research Vehicle Data Translator (NCAR VDT). As part of the VDT, Quality Checks (QCh) were applied and pass rates from these examined by meteorological and non-meteorological factors. Additionally, sensitivity tests were performed with the QCh thresholds to determine if varying the thresholds impacted the pass rates or statistical results of the QCh-ed data. Finally, statistics were calculated for air temperature and barometric pressure in comparison with the weather station, and effects on statistics were examined when separated by different meteorological and non-meteorological factors. Overall, temperature measurements showed consistently good agreement with the ground truth weather station with little of this impact resulting from the QCh process, which demonstrates the feasibility of collecting mobile temperature observations from vehicles. Pressure observations were less well matched with ground truth, the degree of which varied between the make and model of vehicle. Consequently, more work must be done to improve the quality of these observations if barometric pressure from vehicles is to be useful.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner