Chemical Data Assimilation with CMAQ: Continuous vs. Discrete Advection Adjoints

Authors:
Tianyi Gou;Kumaresh Singh;Adrian Sandu
Affiliations:
Department of Computer Science, Virginia Tech., Blacksburg, USA 24060;Department of Computer Science, Virginia Tech., Blacksburg, USA 24060;Department of Computer Science, Virginia Tech., Blacksburg, USA 24060
Venue:
ICCS 2009 Proceedings of the 9th International Conference on Computational Science
Year:
2009

Citing 3
Cited 0

On the limited memory BFGS method for large scale optimization

Mathematical Programming: Series A and B
Recipes for adjoint code construction

ACM Transactions on Mathematical Software (TOMS)
Adjoint sensitivity analysis of regional air quality models

Journal of Computational Physics

Quantified Score

Hi-index	0.00

Visualization

Abstract

The Community Multiscale Air Quality (CMAQ) system is the Environmental Protection Agency's main modeling tool for atmospheric pollution studies. CMAQ-ADJ, the adjoint model of CMAQ, offers new capabilities such as receptor-oriented sensitivity analysis and chemical data assimilation. This paper presents the construction of discrete advection adjoints in CMAQ. The new adjoints are thoroughly validated against finite differences. We assess the performance of discrete and continuous advection adjoints in CMAQ on sensitivity analysis and 4D-Var data assimilation applications. The results show that discrete adjoint sensitivities better agree with finite difference value than their continuous counterparts. However, continuous adjoints result in a faster convergence of the numerical optimization in 4D-Var data assimilation. Similar conclusions apply to modified discrete adjoints.