Application of a statistical post-processing technique to a gridded, operational, air quality forecast L.S. Neal * , P. Agnew, S. Moseley, C. Ord  o ~ nez, N.H. Savage, M. Tilbee Met Office, Fitzroy Road, Exeter, EX1 3PB, UK highlights  An automated bias correction scheme for air quality forecasting is described.  Site specific biases are converted to a gridded field using Kriging.  Bias reduced from 7.02 to 0.53 mgm 3 for O 3 , from 4.00 to 0.13 mgm 3 for PM 2.5 .  Post-processing scheme provides improved model performance out to five days ahead. article info Article history: Received 21 May 2014 Received in revised form 29 August 2014 Accepted 3 September 2014 Available online 4 September 2014 Keywords: Bias correction Air quality forecast Ozone Particulate matter abstract An automated air quality forecast bias correction scheme based on the short-term persistence of model bias with respect to recent observations is described. The scheme has been implemented in the oper- ational Met Office five day regional air quality forecast for the UK. It has been evaluated against routine hourly pollution observations for a year-long hindcast. The results demonstrate the value of the scheme in improving performance. For the first day of the forecast the post-processing reduces the bias from 7.02 to 0.53 mgm 3 for O 3 , from 4.70 to 0.63 mgm 3 for NO 2 , from 4.00 to 0.13 mgm 3 for PM 2.5 and from 7.70 to 0.25 mgm 3 for PM 10 . Other metrics also improve for all species. An analysis of the variation of forecast skill with lead-time is presented and demonstrates that the post-processing in- creases forecast skill out to five days ahead. Crown Copyright © 2014 Published by Elsevier Ltd. All rights reserved. 1. Introduction Regional air quality forecasts have improved significantly over the last decade or so, due to factors such as (i) the availability of near-real-time boundary fluxes provided by improved global composition models; (ii) increased computing power, allowing improved resolution and greater sophistication in the representa- tion of chemical processes; (iii) improved pollutant emission in- ventories. For a review of air quality forecast modelling in Europe see Kukkonen et al. (2012). However despite these advances, the spatially and temporally detailed prediction of atmospheric composition at a given site re- mains a challenging problem and it is not uncommon for forecasts to contain large errors (see Solazzo et al., 2012). These may arise due to errors in inputs of key model parameters such as actual emissions (as opposed to annual mean values: see Pouliot et al., 2012), initial and boundary conditions for chemical species (Schere et al., 2012) as well as meteorology (Vautard et al., 2012). In such circumstances human forecaster intervention may be required to modify the model predictions, based on recent observations and judgement about how conditions are evolving. Alternatively, automated methods may be employed which offer the possibility of improving forecasts and may minimise or completely remove the need for human intervention (e.g. Rouil et al., 2009). A simple daily persistence forecast (i.e. that today's observed values should be the same as yesterday's) is often used as a reference forecast in mete- orological verification (e.g. Jolliffe and Stephenson, 2012). This basic idea can be further developed with varying degrees of so- phistication. The most advanced methods of using observations to improve air quality forecasts are those of data assimilation (DA, see * Corresponding author. E-mail address: lucy.neal@metoffice.gov.uk (L.S. Neal). Contents lists available at ScienceDirect Atmospheric Environment journal homepage: www.elsevier.com/locate/atmosenv http://dx.doi.org/10.1016/j.atmosenv.2014.09.004 1352-2310/Crown Copyright © 2014 Published by Elsevier Ltd. All rights reserved. Atmospheric Environment 98 (2014) 385e393