Expansion of NOAA's national air quality forecast guidance capability to 50 states

The National Oceanic and Atmospheric Administration (NOAA) has produced forecast guidance for surface ozone concentrations and smoke concentrations throughout the lower 48 states (CONUS) for several years and added smoke forecast guidance for Alaska (AK) in August 2009 and Hawaii (HI) in February 2010. These predictions are made in partnership with the US Environmental Protection Agency (EPA) and US Forest Service. During 2010, the NOAA National Weather Service (NWS) has been testing experimental predictions of surface ozone over AK and HI; with successful demonstration of operational readiness criteria, AK and HI ozone predictions will be implemented into NWS' operational product suite in September 2010. Operational forecast guidance is available on the web at http://www.weather.gov/aq/. Experimental prediction guidance is available at http://www.weather/gov/aq-expr.

NOAA's hour by hour forecast guidance at 12 km grid resolution out to 48 hours shows when and where predicted values of ozone and smoke are expected to reach harmful levels in cities, suburbs, and rural areas. Ozone forecasts are produced with a linked numerical prediction system run operationally at the National Centers for Environmental Prediction (NCEP) supercomputing facility: North American Mesoscale (NAM) weather predictions drive the Community Multiscale Air Quality (CMAQ) model. The next upgrade of operational NAM, to the Non-hydrostatic Mesoscale Model on Arakawa B grid (NMM-B), is expected in 2011. Therefore, NOAA is modifying the coupling of NAM to CMAQ: first with a minor adaptation of CMAQ's vertical structure to that of NMM-B, and in follow-on testing, with a new version of CMAQ on the rotated longitude-latitude NMM-B grid. Use of the B-grid in CMAQ will facilitate tighter horizontal coupling between the meteorological and air quality models and it is expected to improve fidelity of air quality predictions at higher horizontal resolution.

NWS is focusing on testing of advanced capabilities that will lead to quantitative predictions of fine particles (PM2.5). Several challenges are being addressed: (1) Chemical mechanisms for inventory-based predictions. More comprehensive chemical mechanisms are needed to account for reactive chemical transport and secondary formation of aerosols from pollutants, but testing has shown these more comprehensive mechanisms over-predict ozone. Compensatory improvements are in testing. (2) Inclusion of intermittent sources. Predictions of PM2.5 from inventoried emissions show substantial seasonal biases that are consistent with missing intermittent sources in the summertime. Smoke from wildfires, and airborne dust are being tested and implemented as components. Standalone experimental testing of dust predictions over CONUS relies on source regions with dust-emissions potential that are estimated from monthly climatology of satellite-observed dust events during 2003-2006 (Ginoux et al. 2010). When surface winds exceed entrainment thresholds, dust is emitted and transported by the HYbrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model driven by NAM meteorology to predict surface and column dust concentrations. Long-range transport of dust impacting CONUS domain is being incorporated through boundary-conditions to help capture events like springtime Asian dust transport and summertime trans-Atlantic transport of Saharan dust. (3) Real-time ingestion of observations. Smoke predictions are based on satellite observations of wildfires location and extent; however surface measurements of the fine particles have not been used in real-time prediction. NOAA is testing a data assimilation capability to reduce biases in predicted surface concentrations of PM2.5. Successful development and testing would lead to advanced developmental guidance for PM2.5 for summer 2011.

To address high bias in experimental ozone predictions in the more comprehensive Carbon Bond 5 (CB-05) gas-phase mechanism employed in newer versions of CMAQ, recent studies examined the sensitivity of ozone predictions to several sources of model uncertainty: (1) specification of lateral boundary conditions, (2) formulation of dry deposition and (3) ways of limiting minimum depth of the planetary boundary layer. Evaluations with hindcasts for August 2009 indicate that changes in all three areas can reduce overprediction biases in most regions. Planned improvements in these areas, along with significant updates to pollutant emission inventories, are being tested for implementation in 2011.

NWS' air quality forecast guidance, experimental and developmental products are being evaluated and tested with a focus group of state and local AQ forecasters. NWS forecasters at the Weather Forecast Offices and NCEP are also encouraged to share their weather expertise and coordinate with their corresponding state and local air quality forecasters.