Estimation of synoptic and mesoscale forecast error covariances in a limited area model
Berre, L.
Année de publication
2000
Statistical and balance features of forecast errors are generally <br>incorporated in the background constraint of variational data <br>assimilation. Forecast error covariances are here estimated with a <br>spectral approach and from a set of forecast differences; <br>autocovariances are calculated with a nonseparable scheme, and multiple <br>linear regressions are used in the formulation of cross covariances. <br>Such an approach was first developed for global models; it is here <br>adapted to ALADIN, a bi-Fourier high-resolution limited-area model, and <br>extended to a multivariate study of humidity forecast errors. Results <br>for autocovariances confirm the importance of nonseparability, in terms <br>of both vertical variability of horizontal correlations and dependence <br>of vertical correlations with horizontal scale; high-resolution spatial <br>correlations are obtained, which should enable a high-resolution <br>analysis. Moreover nonnegligible relationships are found between <br>forecast errors of humidity and those of mass and wind fields.</div>
Accès à la notice sur le site du portail documentaire de Météo-France