Pregled bibliografske jedinice broj: 1025877
Matching initial and lateral perturbations in a convection permitting ensemble
Matching initial and lateral perturbations in a convection permitting ensemble // Joint Congress of the 6th International Conference on Meteorology and Climatology of the Mediterranean & Challenges in Meteorology 5
Zagreb, Hrvatska, 2017. str. 1-1 (predavanje, nije recenziran, sažetak, znanstveni)
CROSBI ID: 1025877 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
Matching initial and lateral perturbations in a convection permitting ensemble
Autori
Keresturi, Endi ; Wang, Yong ; Meier, Florian ; Weidle, Florian ; Wittmann, Christoph
Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, sažetak, znanstveni
Skup
Joint Congress of the 6th International Conference on Meteorology and Climatology of the Mediterranean & Challenges in Meteorology 5
Mjesto i datum
Zagreb, Hrvatska, 22.02.2017. - 24.02.2017
Vrsta sudjelovanja
Predavanje
Vrsta recenzije
Nije recenziran
Ključne riječi
blending, ensemble data assimilation, initial condition perturbations, Jk, limited-area ensemble
(Blending, ensemble data assimilation, initial condition perturbations, Jk, limited-area ensemble)
Sažetak
A new idea for dealing with the consistent initialconditions (ICs) and lateral boundary conditions(LBCs) perturbation is proposed and implementedin a model framework with AROME. The idea is asfollows. Ensemble 3DVAR is used for generating initialperturbations with randomly perturbed observationsand first guess from the short range ensemble forecast.To have the large scale information correctly in theperturbed ICs and consistent with the perturbedLBCs, we introduce a new term in the 3DVAR dataassimilation formulation ÂŰ Jk in addition to Jo(observation term) and Jb (background term). Thisis just the global gridded perturbed ICs as a newsource of information in 3DVAR which we assimilatetogether with the mentioned EDA. The system usedhere is AROME in an ensemble framework (2, 5L90)with 16 ensemble members (plus control) coupled toECMWF-EPS (18L91). For complete implementationof Jk, we need ECMWF-EPS forecasts and their errorcovariances contained in V-matrix. V-matrix is calcu-lated via ensemble method by interpolating ECMWFdata directly to AROME domain. However, thosenewly added wavelengths due to interpolation, whichcontain no useful information, will be discarded duringthe 3DVAR itself. I.e. they are not allowed to impactJk. Also, the scale selection (which scales are to be usedfrom the driving model) is discussed. V-matrix is keptunivariate for studies presented here. The system istested in several case studies. Local strong convection in summer and a snowfall event in winter over Southand Central Europe. Also, a long-term verificationstudy is performed. The tests are still ongoing, resultsand data are still being analysed. Final conclusions andresults, especially with respect to the Jk term, will bepresented at the conference.
Izvorni jezik
Engleski
Znanstvena područja
Geofizika