Premium
4D‐Var analysis of potential vorticity pseudo‐observations
Author(s) -
Guerin Roy,
Desroziers Gerald,
Arbogast Philippe
Publication year - 2006
Publication title -
quarterly journal of the royal meteorological society
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.744
H-Index - 143
eISSN - 1477-870X
pISSN - 0035-9009
DOI - 10.1256/qj.05.53
Subject(s) - potential vorticity , data assimilation , tropopause , hessian matrix , meteorology , climatology , middle latitudes , mathematics , operator (biology) , numerical weather prediction , cyclone (programming language) , diagonal , vorticity , environmental science , computer science , geology , vortex , physics , stratosphere , biochemistry , chemistry , geometry , repressor , field programmable gate array , transcription factor , computer hardware , gene
Abstract It is shown that a poor forecast of a midlatitude cyclone developing over the western Atlantic Ocean can be significantly improved by potential vorticity (PV) modifications in the vicinity of the dynamical tropopause in the initial conditions, in order to modify the position of the tropopause. Manually built PV corrections are introduced by using the ARPEGE‐IFS 4D‐Var assimilation system. For that purpose, a PV observation operator, its tangent‐linear and adjoint versions, based on a simplified form of Ertel PV, has been implemented. The impact of the PV pseudo‐observations in an optimal system (i.e. analysed using tuned observation‐error statistics) is positive; the maximum positive improvement of the forecast is obtained over western Europe when the cyclone reaches its maximum amplitude. The forecast is never degraded during the model run, whereas another run forced by the pseudo‐observations is characterized by the development of a forecast error pattern downstream to the main pattern of analysis increment. The conditioning of the minimization problem defined as the sparsity (for the non‐diagonal parts) of the Hessian matrix is weakly affected by the PV observation operator leading to quite good convergence. Copyright © 2006 Royal Meteorological Society.