Search

Abstract: In this paper we investigate the added value of assimilation of ozone in-situ measurements for both re-analysis and forecasting purposes. Various simulations were performed using the LOTOS-EUROS chemical transport model and an Ensemble Kalman Filter (EnKF) to assimilate measured ozone surface concentrations over Europe, for spring and summer of 2007. The results for the re-analysis of ozone show a significant improvement in the LOTOS-EUROS performance score when compared to measurements. The average correlation coefficient for the daily maximum ozone concentration improves from 0.72 to 0.83. Similarly, the average Root Mean Square Error (RMSE) for the daily maximum ozone concentration is reduced from 20.8 to 16.9 μg m−3. The free running model performs well in forecast mode and the agreement between in-situ and modeled ozone concentration is good. The average temporal correlation coefficient ranges from 0.62 for the first day forecast to 0.61 for the third day forecast. Based on these results, assimilated fields were used to initialize the forecast. As for the re-analysis a better comparison between model and observation was observed. The mean correlation coefficient increased by 0.07 and the averaged RMSE decreased by 0.65 μg m−3. However, the addition of an inheritance scheme to import additional information from the data assimilation to the forecast did not significantly improve the concentration fields. [Copyright &y& Elsevier]

Abstract: In this paper, we examine the extent to which an ensemble generated from a single air quality model correctly represents the global uncertainty of ozone simulations with a regional scale chemistry-transport model and whether it is suitable for ozone data assimilation. An ensemble of 30 members is constructed from a reference simulation in which the model parameters that are the most uncertain for determining ozone concentrations are randomly perturbed. Comparisons of the simulated ensemble using the model CHIMERE with observations are examined over the summer season both at the surface and higher in the troposphere. Although the ensemble overestimates vertical and horizontal correlation between errors, the average model error is well represented both at the surface and in the vertical dimension (after an adjustment for the latter). A variability of about 7–8ppb is found. The results found in this study with respect to tropospheric ozone model errors and error correlation lengths both from ensemble simulations and model to observation comparisons are of particular interest for data assimilation when constructing model error covariance matrices. [ABSTRACT FROM AUTHOR]

Summary: This paper compares two Monte Carlo sequential data assimilation methods based on the Kalman filter, for estimating the effect of measurements on simulations of state error variance made by a one-dimensional hydrodynamic model. The first method used an ensemble Kalman filter (EnKF) to update state estimates, which were then used as initial conditions for further simulations. The second method used an ensemble transform Kalman filter (ETKF) to quickly estimate the effect of measurement error covariance on forecast error covariance without the need to re-run the simulation model. The ETKF gave an unbiased estimate of EnKF analysed error variance, although differences in the treatment of measurement errors meant the results were not identical. Estimates of forecast error variance could also be made, but their accuracy deteriorated as the time from measurements increased due in part to model non-linearity and the decreasing signal variance. The motivation behind the study was to assess the ability of the ETKF to target possible measurements, as part of an adaptive sampling framework, before they are assimilated by an EnKF-based forecasting model on the River Crouch, Essex, UK. The ETKF was found to be a useful tool for quickly estimating the error covariance expected after assimilating measurements into the hydrodynamic model. It, thus, provided a means of quantifying the ‘usefulness’ (in terms of error variance) of possible sampling schemes. [Copyright &y& Elsevier]