Méthode de désagrégation statistico-dynamique adaptée aux forçages atmosphériques pour la modélisation de l'Océan Atlantique PDF Download
Are you looking for read ebook online? Search for your book and save it on your Kindle device, PC, phones or tablets. Download Méthode de désagrégation statistico-dynamique adaptée aux forçages atmosphériques pour la modélisation de l'Océan Atlantique PDF full book. Access full book title Méthode de désagrégation statistico-dynamique adaptée aux forçages atmosphériques pour la modélisation de l'Océan Atlantique by Marie Minvielle. Download full books in PDF and EPUB format.
Author: Marie Minvielle Publisher: ISBN: Category : Languages : en Pages : 175
Book Description
Ocean plays a main role in climate regulation. Its good representation by climate models is necessary to have a correct estimation of mean state and variability of the climate system. However, some biases are present in the representation of the ocean by climate models. This is partly due to their too low horizontal resolution and a poor representation of the atmospheric variables at the ocean surface. In this thesis, a statistico-dynamical downscaling method is carried out. The objective is to obtain a better estimation of oceanic mean state and variability over the Atlantic basin, by reducing the systematic biases of climate models. First, the method consists in reconstructing the surface atmospheric variables by statistical relationship with the large scale atmospheric dynamic, estimated by weather regimes. The resulting forcing set is used afterward to force an oceanic model at a higher resolution than the classical resolution used for the ocean in climate models. Once this method built and validated with the observations over the second half 20th century, it is applied to the large scale atmospheric dynamic from the historical simulation of the CNRM-CM3 climate model. The analysis of the reconstructed forcing and its associated oceanic simulation emphasizes the efficiency of the method. In spite of some weaknesses, the method fully accomplishes its role in correcting climate models biases.
Author: Marie Minvielle Publisher: ISBN: Category : Languages : en Pages : 175
Book Description
Ocean plays a main role in climate regulation. Its good representation by climate models is necessary to have a correct estimation of mean state and variability of the climate system. However, some biases are present in the representation of the ocean by climate models. This is partly due to their too low horizontal resolution and a poor representation of the atmospheric variables at the ocean surface. In this thesis, a statistico-dynamical downscaling method is carried out. The objective is to obtain a better estimation of oceanic mean state and variability over the Atlantic basin, by reducing the systematic biases of climate models. First, the method consists in reconstructing the surface atmospheric variables by statistical relationship with the large scale atmospheric dynamic, estimated by weather regimes. The resulting forcing set is used afterward to force an oceanic model at a higher resolution than the classical resolution used for the ocean in climate models. Once this method built and validated with the observations over the second half 20th century, it is applied to the large scale atmospheric dynamic from the historical simulation of the CNRM-CM3 climate model. The analysis of the reconstructed forcing and its associated oceanic simulation emphasizes the efficiency of the method. In spite of some weaknesses, the method fully accomplishes its role in correcting climate models biases.
Book Description
AFIN DE DETERMINER LES VALEURS DES PARAMETRES AJUSTABLES D'UN MODELE OCEANIQUE QUI CONDUISENT A LA MEILLEURE ADEQUATION POSSIBLE ENTRE MODELE ET OBSERVATIONS, UNE METHODE INVERSE ADAPTATIVE EST DEVELOPPEE ET EST APPLIQUEE A UN MODELE DE PREDICTION DE LA TEMPERATURE DE SURFACE (SST) DE L'OCEAN ATLANTIQUE TROPICAL. L'OPTIMISATION EST REALISEE EN MINIMISANT L'ECART ENTRE DONNEES OBSERVEES ET SIMULEES, QUI DEPEND D'ERREURS D'OBSERVATION ET DE MODELISATION. UNE PROCEDURE ADAPTATIVE EST MISE EN PLACE AU COURS DE LAQUELLE LE MODELE QUE L'ON OPTIMISE EST UTILISE AFIN DE CONSTRUIRE LE MODELE D'ERREURS D'OBSERVATION. L'OPTIMISATION EST APPLIQUEE AU CYCLE SAISONNIER MOYEN EN UTILISANT LES ANOMALIES DE SST POUR DIFFERENTES ANNEES ET DIFFERENTS FORCAGES ATMOSPHERIQUES PLAUSIBLES, COMME INFORMATION ADDITIONNELLE POUR CONSTRUIRE UNE ESTIMATION DE LA MATRICE DE COVARIANCE D'ERREURS D'OBSERVATION. EN UTILISANT UN MODELE IDEALISE POUR REPRESENTER LES ERREURS DE MODELISATION, LA PROCEDURE EST APPLIQUEE AU MODELE DE SST DE BLUMENTHAL ET CANE (1989) ET CONDUIT A DES VALEURS AFFINEES DES PLUSIEURS PARAMETRES DU MODELE ET DES FLUX DE CHALEUR. LA SIMULATION DE LA SST EN MOYENNE ANNUELLE EST AMELIOREE ALORS QUE LA SIMULATION DES FLUCTUATIONS SAISONNIERES ET DES ANOMALIES MENSUELLES NE L'EST PAS. L'ECART RESIDUEL ENTRE MODELE ET OBSERVATIONS DEMEURE TROP IMPORTANT POUR ETRE ATTRIBUE AUX ERREURS SUR LE FORCAGE ATMOSPHERIQUE ET SUR LES DONNEES OCEANIQUES UNIQUEMENT. IL EST LIE A LA GEOMETRIE SIMPLIFIEE DU MODELE AINSI QU'A SA MAUVAISE REPRESENTATION DU PHENOMENE DE REFROIDISSEMENT DES EAUX DE SURFACE PAR REMONTEE D'EAUX FROIDES (UPWELLING). L'EXISTENCE D'ERREURS DE MODELISATION PLUS IMPORTANTES QUE CELLES PRISES EN COMPTE DANS LE MODELE D'ERREUR THEORIQUE EST CONFIRME PAR UN TEST STATISTIQUE DE VALIDATION DES HYPOTHESES DU CALCUL INVERSE
Author: Marie Minvielle
Author: Marie Minvielle
Author: NATHALIE.. SENNECHAEL