Comparing CMAQ to Observations

• Activity Number: 466
• Type: Contributed
• Date/Time: Wednesday, August 9, 2006 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics and the Environment
• Author(s): Li Chen*+ and Michael L. Stein
• Companies: The University of Chicago and The University of Chicago
• Keywords: aggregation ; fractional bias ; root normalized mean square error ; variation decomposition

Abstract

This talk compares observed hourly ozone concentrations to the CMAQ modeled ozone concentrations at three different spatial resolutions, 36km, 12km and 4km. Performance measures, e.g., fractional bias and root normalized mean square error, are calculated. The results show that higher resolution CMAQ model output does not necessarily provide smaller FB and RNMSE than the lower ones. Aggregation is performed to obtain new versions of lower resolution model output based on higher resolution model output. The aggregated lower resolution model output predicts better than either the unaggregated high resolution run or the low resolution run in terms of RNMSE. Variation decomposition is used to understand the statistical behavior of CMAQ model output at different resolutions. The temporal variation is well captured by CMAQ output, but spatial variation and space-time interaction are not.