Resolution of sharp fronts in the presence of model error in variational data assimilation

Melina A. Freitag, Nancy Nichols, Chris BuddQuarterly 2012 · DOI

Abstract

Abstract We show that the four‐dimensional variational data assimilation method (4DVar) can be interpreted as a form of Tikhonov regularization, a very familiar method for solving ill‐posed inverse problems. It is known from image restoration problems that L 1 ‐norm penalty regularization recovers sharp edges in the image more accurately than Tikhonov, or L 2 ‐norm, penalty regularization. We apply this idea from stationary inverse problems to 4DVar, a dynamical inverse problem, and give examples for an L 1 ‐norm penalty approach and a mixed total variation (TV) L 1 – L 2 ‐norm penalty approach. For problems with model error where sharp fronts are present and the background and observation error covariances are known, the mixed TV L 1 – L 2 ‐norm penalty performs better than either the L 1 ‐norm method or the strong‐constraint 4DVar ( L 2 ‐norm) method. A strength of the mixed TV L 1 – L 2 ‐norm regularization is that in the case where a simplified form of the background error covariance matrix is used it produces a much more accurate analysis than 4DVar. The method thus has the potential in numerical weather prediction to overcome operational problems with poorly tuned background error covariance matrices. Copyright © 2012 Royal Meteorological Society
BibTeX
@inproceedings{freitag2012resolution,
  author = {Melina A. Freitag and Nancy Nichols and Chris Budd},
  title = {{Resolution of sharp fronts in the presence of model error in variational data assimilation}},
  booktitle = {Quarterly},
  year = {2012},
  doi = {10.1002/qj.2002},
}

Loading related content...

Figure