Optimal interpolation data assimilation of surface currents by utilizing pseudo measurement with Monte Carlo simulation
Lei Ren, Stephen Nash, Michael Hartnett
1Department of Civil Engineering, National Univerisity of Ireland, Galway, University Road, Galway, Ireland
2Ryan Institute, Galway, Ireland
Optimal Interpolation (OI) data assimilation is a technique to combine available observations with background states to improve prediction states. In this research, pseudo measurement of surface currents generated by adding noise with Monte Carlo simulation is used to update the background states with optimal interpolation. The core of Optimal Interpolation data assimilation is the definition of background error covariance, which determines to what extent the model background states will be corrected to match the observations. The background error covariance is computed before the data assimilation process. The model background errors are calculated from the mean over a short time interval ten minutes. A series of sensitivity tests with Optimal Interpolation are done by calculating Root Mean Square Error (RMSE) to decide the appropriate parameters. The improvement of Optimal Interpolation at reference points is measured in Taylor diagrams, and the surface current maps of test domain show the effectiveness of Optimal Interpolation.