The Impact of Nudging in the Meteorological Model for Retrospective Air Quality Simulations. Part I: Evaluation against National Observation Networks

Abstract It is common practice to use Newtonian relaxation, or nudging, throughout meteorological model simulations to create “dynamic analyses” that provide the characterization of the meteorological conditions for retrospective air quality model simulations. Given the impact that meteorological conditions have on air quality simulations, it has been assumed that the resultant air quality simulations would be more skillful by using dynamic analyses rather than meteorological forecasts to characterize the meteorological conditions, and that the statistical trends in the meteorological model fields are also reflected in the air quality model. This article, which is the first of two parts, demonstrates the impact of nudging in the meteorological model on retrospective air quality model simulations. Here, meteorological simulations are generated by the fifth-generation Pennsylvania State University–National Center for Atmospheric Research Mesoscale Model (MM5) using both the traditional dynamic analysis approach and using forecasts for a summertime period. The resultant fields are then used to characterize the meteorological conditions for emissions processing and air quality simulations using the Community Multiscale Air Quality (CMAQ) Modeling System. As expected, on average, the near-surface meteorological fields show a significant degradation over time in the forecasts (when nudging is not used), while the dynamic analyses maintain nearly constant statistical scores in time. The use of nudged MM5 fields in CMAQ generally results in better skill scores for daily maximum 1-h ozone mixing ratio simulations. On average, the skill of the daily maximum 1-h ozone simulation deteriorates significantly over time when nonnudged MM5 fields are used in CMAQ. The daily maximum 1-h ozone mixing ratio also degrades over time in the CMAQ simulation that uses MM5 dynamic analyses, although to a much lesser degree, despite no aggregate loss of skill over time in the dynamic analyses themselves. These results affirm the advantage of using nudging in MM5 to create the meteorological characterization for CMAQ for retrospective simulations, and it is shown that MM5-based dynamic analyses are robust at the surface throughout 5.5-day simulations.