Automatic differentiation in Odyssée

Rostaing, Nicole; Dalmas, Stéphane; Galligo, André

doi:10.3402/tellusa.v45i5.15060

Cited by 55 publications

(25 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this study the latter approach is adopted, which comprises the coding and implementation of the adjoint operators of T h , T z , A and D z given in (9). The adjoint chemistry was coded by hand, while for the advection and diffusion routines the AMC adjoint model compiler (Giering and Kaminski, 1998) and O∂yssée differentiation system (Rostaing et al, 1993;Faure and Papegay, 1998) has been used. The correctness of the adjoint code was checked by the method proposed by Chao and Chang (1992).…”

Section: The Eurad Forward Model Ctm2 and Its Adjointmentioning

confidence: 99%

Emission rate and chemical state estimation by 4-dimensional variational inversion

et al. 2007

View full text Add to dashboard Cite

Abstract. This study aims to assess the potential and limits of an advanced inversion method to estimate pollutant precursor sources mainly from observations. Ozone, sulphur dioxide, and partly nitrogen oxides observations are taken to infer source strength estimates. As methodology, the fourdimensional variational data assimilation technique has been generalised and employed to include emission rate optimisation, in addition to chemical state estimates as usual objective of data assimilation. To this end, the optimisation space of the variational assimilation system has been complemented by emission rate correction factors of 19 emitted species at each emitting grid point, involving the University of Cologne mesoscale EURAD model. For validation, predictive skills were assessed for an August 1997 ozone episode, comparing forecast performances of pure initial value optimisation, pure emission rate optimisation, and joint emission rate/initial value optimisation.Validation procedures rest on both measurements withheld from data assimilation and prediction skill evaluation of forecasts after the inversion procedures. Results show that excellent improvements can be claimed for sulphur dioxide forecasts, after emission rate optimisation. Significant improvements can be claimed for ozone forecasts after initial value and joint emission rate/initial value optimisation of precursor constituents. The additional benefits applying joint emission rate/initial value optimisation are moderate, and very useful in typical cases, where upwind emission rate optimisation is essential. In consequence of the coarse horizontal model grid resolution of 54 km, applied in this study, comparisons indicate that the inversion improvements can rest on assimilating ozone observations only, as the inclusion of NO x observations does not provide additional forecast skill. Emission estimates were found to be largely independent from initial Correspondence to: H. Elbern (he@eurad.uni-koeln.de) guesses from emission inventories, demonstrating the potential of the 4D-var method to infer emission rate improvements. The study also points to the need for improved horizontal model resolution to more efficient use of NO x observations.

show abstract

Section: The Eurad Forward Model Ctm2 and Its Adjointmentioning

confidence: 99%

Emission rate and chemical state estimation by 4-dimensional variational inversion

et al. 2007

View full text Add to dashboard Cite

show abstract

“…However, generation of adjoint codes by hand is time-consuming and error-prone. Recently several automatic differentiation tools such as Odyssee (Rostaing et al 1993), ADIFOR (Bischof et al 1996) and TAMC (Giering 1999) are developed and applied to atmospheric models.…”

Section: Adjoint Methodsmentioning

confidence: 99%

Recent Progress of Data Assimilation Methods in Meteorology

Tsuyuki

Miyoshi

2007

Journal of the Meteorological Society of Japan

View full text Add to dashboard Cite

Data assimilation is a methodology for estimating accurately the state of a time-evolving complex system like the atmosphere from observational data and a numerical model of the system. It has become an indispensable tool for meteorological researches as well as for numerical weather prediction, as represented by extensive use of reanalysis datasets for research purposes. New advances of data assimilation methods emerged from the 1980s. This review paper presents the theoretical background and implementation of two advanced data assimilation methods: four-dimensional variational assimilation (4D-Var) and ensemble Kalman filtering (EnKF), which currently draw much attention in the meteorological community. Recent research results in Japan on those methods are reviewed, especially on mesoscale applications of 4D-Var and tests of the local ensemble transform Kalman filter (LETKF). Comparison of 4D-Var and EnKF is also briefly discussed. An outline of the mesoscale 4D-Var system of the Japan Meteorological Agency, which is the first operational 4D-Var for a mesoscale model, is given in Appendix.

show abstract

“…A number of AD software tools are available, including ADIFOR [16], Odyssee [110], and TAMC [50] for FORTRAN 77, and ADIC [17] and ADOL-C [55] for C/C++.…”

Section: Automatic Differentiationmentioning

confidence: 99%

What Color Is Your Jacobian? Graph Coloring for Computing Derivatives

Gebremedhin¹,

Manne²,

Pothen³

2005

SIAM Rev.

232

174

View full text Add to dashboard Cite

Abstract. Graph coloring has been employed since the 1980s to efficiently compute sparse Jacobian and Hessian matrices using either finite differences or automatic differentiation. Several coloring problems occur in this context, depending on whether the matrix is a Jacobian or a Hessian, and on the specifics of the computational techniques employed. We consider eight variant vertex coloring problems here. This article begins with a gentle introduction to the problem of computing a sparse Jacobian, followed by an overview of the historical development of the research area. Then we present a unifying framework for the graph models of the variant matrix estimation problems. The framework is based upon the viewpoint that a partition of a matrix into structurally orthogonal groups of columns corresponds to distance-2 coloring an appropriate graph representation. The unified framework helps integrate earlier work and leads to fresh insights; enables the design of more efficient algorithms for many problems; leads to new algorithms for others; and eases the task of building graph models for new problems. We report computational results on two of the coloring problems to support our claims. Most of the methods for these problems treat a column or a row of a matrix as an atomic entity, and partition the columns or rows (or both). A brief review of methods that do not fit these criteria is provided. We also discuss results in discrete mathematics and theoretical computer science that intersect with the topics considered here.

show abstract

Automatic differentiation in Odyssée

Cited by 55 publications

References 6 publications

Emission rate and chemical state estimation by 4-dimensional variational inversion

Emission rate and chemical state estimation by 4-dimensional variational inversion

Recent Progress of Data Assimilation Methods in Meteorology

What Color Is Your Jacobian? Graph Coloring for Computing Derivatives

Contact Info

Product

Resources

About