A hybrid Bayesian Kalman filter and applications to numerical wind speed modeling. Issue 167 (August 2017)

• Record Type: Journal Article
• Authors: Galanis, George; Papageorgiou, Evgenia; Liakatas, Aristotelis
• Is Part Of: Journal of wind engineering and industrial aerodynamics. Issue 167(2017)
• Page Start: 1
• Page End: 22
• Publication Date: 2017-08
• Subjects: Bayesian inference -- Kalman filtering -- Numerical atmospheric modeling; Wind-pressure -- Periodicals; Buildings -- Aerodynamics -- Periodicals; Pression du vent -- Périodiques; Constructions -- Aérodynamique -- Périodiques; Buildings -- Aerodynamics; Wind-pressure; Periodicals
• DOI: 10.1016/j.jweia.2017.04.007
• Languages: English
• ISSNs: 0167-6105

Abstract

Abstract: A new hybrid optimization technique for numerical environmental simulation models is proposed and tested in this work. Bayesian modeling is utilized in conjunction with a nonlinear Kalman filter towards a novel post process algorithm applied to numerical wind speed simulations. The new model is tested on idealized data as well as on numerical model forecasts leading to promising results and supporting both the reduction of systematic biases but also the significant limitation of the error variability and the associated forecast uncertainty, a point where classical Kalman filters usually fail to contribute. Highlights: A hybrid optimization system is proposed for numerical models predictions. Non linear Kalman filters are combined with Bayesian modeling targeting to the reduction of biases and error variability. The proposed system is evaluated over idealized data and numerical wind speed forecasts with very promising results.