O3 Forecasts Research Articles

Assessment of Kalman filter bias-adjustment technique to improve the simulation of ground-level ozone over Spain

The CALIOPE air quality modelling system has been used to diagnose ground level O3 concentration for the year 2004, over the Iberian Peninsula. We investigate the improvement in the simulation of daily O3 maximum by the use of a post-processing such as the Kalman filter bias-adjustment technique. The Kalman filter bias-adjustment technique is a recursive algorithm to optimally estimate bias-adjustment terms from previous measurements and model results. The bias-adjustment technique improved the simulation of daily O3 maximum for the entire year and the all the stations considered over the whole domain. The corrected simulation presents improvements in statistical indicators such as correlation, root mean square error, mean bias, and gross error. After the post-processing the exceedances of O3 concentration limits, as established by the European Directive 2008/50/CE, are better reproduced and the uncertainty of the modelling system, as established by the European Directive 2008/50/CE, is reduced from 20% to 7.5%. Such uncertainty in the model results is under the established EU limit of the 50%. Significant improvements in the O3 timing and amplitude of the daily cycle are also observed after the post-processing. The systematic improvements in the O3 maximum simulations suggest that the Kalman filter post-processing method is a suitable technique to reproduce accurate estimate of ground-level O3 concentration. With this study we evince that the adjusted O3 concentrations obtained after the post-process of the results from the CALIOPE system are a reliable means for real near time O3 forecasts.

Assessment of an ensemble of seven real‐time ozone forecasts over eastern North America during the summer of 2004

The real‐time forecasts of ozone (O3) from seven air quality forecast models (AQFMs) are statistically evaluated against observations collected during July and August of 2004 (53 days) through the Aerometric Information Retrieval Now (AIRNow) network at roughly 340 monitoring stations throughout the eastern United States and southern Canada. One of the first ever real‐time ensemble O3 forecasts, created by combining the seven separate forecasts with equal weighting, is also evaluated in terms of standard statistical measures, threshold statistics, and variance analysis. The ensemble based on the mean of the seven models and the ensemble based on the median are found to have significantly more temporal correlation to the observed daily maximum 1‐hour average and maximum 8‐hour average O3 concentrations than any individual model. However, root‐mean‐square errors (RMSE) and skill scores show that the usefulness of the uncorrected ensembles is limited by positive O3 biases in all of the AQFMs. The ensembles and AQFM statistical measures are reevaluated using two simple bias correction algorithms for forecasts at each monitor location: subtraction of the mean bias and a multiplicative ratio adjustment, where corrections are based on the full 53 days of available comparisons. The impact the two bias correction techniques have on RMSE, threshold statistics, and temporal variance is presented. For the threshold statistics a preferred bias correction technique is found to be model dependent and related to whether the model overpredicts or underpredicts observed temporal O3 variance. All statistical measures of the ensemble mean forecast, and particularly the bias‐corrected ensemble forecast, are found to be insensitive to the results of any particular model. The higher correlation coefficients, low RMSE, and better threshold statistics for the ensembles compared to any individual model point to their preference as a real‐time O3 forecast.

Linking the Eta Model with the Community Multiscale Air Quality (CMAQ) Modeling System to Build a National Air Quality Forecasting System

Abstract NOAA and the U.S. Environmental Protection Agency (EPA) have developed a national air quality forecasting (AQF) system that is based on numerical models for meteorology, emissions, and chemistry. The AQF system generates gridded model forecasts of ground-level ozone (O3) that can help air quality forecasters to predict and alert the public of the onset, severity, and duration of poor air quality conditions. Although AQF efforts have existed in metropolitan centers for many years, this AQF system provides a national numerical guidance product and the first-ever air quality forecasts for many (predominantly rural) areas of the United States. The AQF system is currently based on NCEP’s Eta Model and the EPA’s Community Multiscale Air Quality (CMAQ) modeling system. The AQF system, which was implemented into operations at the National Weather Service in September of 2004, currently generates twice-daily forecasts of O3 for the northeastern United States at 12-km horizontal grid spacing. Preoperational testing to support the 2003 and 2004 O3 forecast seasons showed that the AQF system provided valuable guidance that could be used in the air quality forecast process. The AQF system will be expanded over the next several years to include a nationwide domain, a capability for forecasting fine particle pollution, and a longer forecast period. State and local agencies will now issue air quality forecasts that are based, in part, on guidance from the AQF system. This note describes the process and software components used to link the Eta Model and CMAQ for the national AQF system, discusses several technical and logistical issues that were considered, and provides examples of O3 forecasts from the AQF system.

NOGAPS-ALPHA model simulations of stratospheric ozone during the SOLVE2 campaign

Abstract. This paper presents three-dimensional prognostic O3 simulations with parameterized gas-phase photochemistry from the new NOGAPS-ALPHA middle atmosphere forecast model. We compare 5-day NOGAPS-ALPHA hindcasts of stratospheric O3 with satellite and DC-8 aircraft measurements for two cases during the SOLVE II campaign: (1) the cold, isolated vortex during 11-16 January 2003; and (2) the rapidly developing stratospheric warming of 17-22 January 2003. In the first case we test three different photochemistry parameterizations. NOGAPS-ALPHA O3 simulations using the NRL-CHEM2D parameterization give the best agreement with SAGE III and POAM III profile measurements. 5-day NOGAPS-ALPHA hindcasts of polar O3 initialized with the NASA GEOS4 analyses produce better agreement with observations than do the operational ECMWF O3 forecasts of case 1. For case 2, both NOGAPS-ALPHA and ECMWF 114-h forecasts of the split vortex structure in lower stratospheric O3 on 21 January 2003 show comparable skill. Updated ECMWF O3 forecasts of this event at hour 42 display marked improvement from the 114-h forecast; corresponding updated 42-hour NOGAPS-ALPHA prognostic O3 fields initialized with the GEOS4 analyses do not improve significantly. When NOGAPS-ALPHA prognostic O3 is initialized with the higher resolution ECMWF O3 analyses, the NOGAPS-ALPHA 42-hour lower stratospheric O3 fields closely match the operational 42-hour ECMWF O3 forecast of the 21 January event. We find that stratospheric O3 forecasts at high latitudes in winter can depend on both model initial conditions and the treatment of photochemistry over periods of 1-5 days. Overall, these results show that the new O3 initialization, photochemistry parameterization, and spectral transport in the NOGAPS-ALPHA NWP model can provide reliable short-range stratospheric O3 forecasts during Arctic winter.

Aircraft and surface observations of air quality in Puget Sound and a comparison to a regional model

Abstract Three research flights were conducted during August 2001 as part of the PNW2001 field study in the Puget Sound region. Each flight presented a case study of unique sources, meteorological conditions, and spatial distributions of pollutants. On 10 August, an urban smog episode was sampled downwind of the Puget Sound Seattle–Vancouver corridor containing high mixing ratios of O3 (∼80 ppbv) and high aerosol scattering coefficients (σsp) (∼35 M m−1). The ozone and σsp measured during this flight were typical of summer pollution events in the Puget Sound region. The AIRPACT forecast model (MM5–CALMET-CALGRID) showed variable results when compared to surface measurements. In addition, predicted surface O3 maxima were several hours too early and fell off sharply relative to observations. AIRPACT generally under-predicted upper-level O3 concentrations and misplaced the observed O3 maxima. On 16 August, a second pollution region was encountered in a broad layer between 2 and 4.5 km above a well-defined temperature inversion. Meteorological analysis and satellite imagery suggest the high O3 mixing ratios (60 ppbv) and very high σsp (80 M m−1) were from fires burning in Washington, Oregon, and Northern California. IMPROVE aerosol chemistry data from sites in the Cascade Mountains showed evidence of biomass burning emissions reaching the surface 2 days prior to this flight. The last flight, on 20 August, sampled slightly polluted background conditions in the region. Ozone levels reached only 30 ppbv and σsp reached 20 M m−1 at the surface, but was much lower aloft. Therefore, this flight was representative of background conditions during the summer in this region. The results from these research flights provide evidence of enhanced O3 and σsp, extending to 4 km altitude over Puget Sound, during the summer months from varying sources. They further demonstrate the current capability of the AIRPACT O3 forecast system and, in one case, illustrate the need to include biomass-burning emissions in the model forecasts.