Premium
Land surface initialization strategy for a global reforecast dataset
Author(s) -
Boisserie M.,
Decharme B.,
Descamps L.,
Arbogast P.
Publication year - 2016
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 - 10.1002/qj.2688
Subject(s) - initialization , ensemble forecasting , environmental science , meteorology , range (aeronautics) , computer science , interim , climatology , set (abstract data type) , global forecast system , numerical weather prediction , geography , geology , materials science , archaeology , composite material , programming language
A 32‐year global ensemble reforecast dataset has recently been developed at Météo‐France that is approximatively consistent with the operational global ensemble forecast system PEARP. Unlike ECMWF or NCEP, Météo‐France does not possess a reanalysis of its own operational forecast system. Therefore, the initial atmospheric state and boundary conditions of the reforecasts are from the ECMWF ERA‐Interim reanalysis. This article presents a study of the sensitivity of the reforecasts to the method of land‐surface initialization. To this end, two sets of short‐range hindcasts using different land‐surface initialization approaches are compared. The first set is initialized from interpolated ERA‐Interim land‐surface fields based on a transfer function. The second set is initialized from offline simulations of the Météo‐France land‐surface model (SURFEX) driven by the 3‐hourly near‐surface atmospheric fields of the ERA‐Interim reanalysis. Each set is run from 1800 UTC initial conditions and up to +108 h. Because better results overall are found using offline SURFEX simulations, this latter approach was chosen to perform an ensemble reforecast dataset. Then, this ensemble reforecast database will be used to build a climatology of the operational ensemble prediction system of Météo‐France, which will, in turn, help to better estimate systematic forecast errors and, more importantly, improve the forecasting of rare extreme weather events.