Multiscale Performance of the ALARO-0 Model for Simulating Extreme Summer Precipitation Climatology in Belgium

Abstract

Daily summer precipitation over Belgium from the Aire Limitee Adaptation Dynamique Developpement International (ALADIN) model and a version of the model that has been updated with physical parameterizations, the so-called ALARO-0 model [ALADIN and AROME (Application de la Recherche a l'Operationnel a Meso-Echelle) combined model, first baseline version released in 1998], are compared with respect to station observations for the period 1961-90. The 40-yr European Centre for Medium-Range Weather Forecasts Re-Analysis (ERA-40) is dynamically downscaled using both models on a horizontal resolution of 40 km, followed by a one-way nesting on high spatial resolutions of 10 and 4 km. This setup allows us to explore the relative importance of spatial resolution versus parameterization formulation on the model skill to correctly simulate extreme daily precipitation. Model performances are assessed through standard statistical errors and density, frequency, and quantile distributions as well as extreme value analysis, using the peak-over-threshold method and generalized Pareto distribution. The 40-km simulations of ALADIN and ALARO-0 show similar results, both reproducing the observations reasonably well. For the high-resolution simulations, ALARO-0 at both 10 and 4 km is in better agreement with the observations than ALADIN. The ALADIN model consistently produces too high precipitation rates. The findings demonstrate that the new parameterizations within the ALARO-0 model are responsible for a correct simulation of extreme summer precipitation at various horizontal resolutions. Moreover, this study shows that ALARO-0 is a good candidate model for regional climate modeling.

Local impact analysis of climate change on precipitation extremes: are high-resolution climate models needed for realistic simulations?

“…Hereafter, ALARO is used as shorthand name for the ALARO-0 model described in De Troch et al (2013). The ALADIN model is the limited area model (LAM) version of the Action de Recherche Petite Echelle Grande Echelle Integrated Forecast System (ARPEGE-IFS).…”

“…The physics parameterization package of the ALARO model was designed specifically for running at resolutions between 3 and 8 km. The specific characteristics of the Modular Multiscale Microphysics and Transport (3MT) convection scheme used in the ALARO model lead to a good multiscale performance, particularly in convection-permitting resolutions (De Troch et al, 2013). The ALARO simulations for the present climate conditions over Belgium were performed for the periods 1961-1990 and 1981-2010 at resolutions ranging from 40 km down to 4 km, both using a set of simulations forced with ERA-40 or ERAInterim reanalysis as well as with the CNRM-CM3 GCM for the historical control run (Table 1).…”

“…Generally, it can be inferred from the results that, at the synoptic (daily) scale, the projections by the ALARO model are consistent with those from the driving GCMs. De Troch et al (2013) pointed out that an increase in spatial resolution in the ALARO model is not as important as the parameterization scheme used for extreme precipitation modeling at the daily scale. Figure 8 shows the change factors for daily and 3-hourly precipitation computed using the CCLM EC-EARTH model with different spatial resolutions and the driving EC-EARTH GCM for the winter and summer seasons.…”

Study of the Jacobian of an extended Kalman filter for soil analysis in SURFEXv5

“…In the ALARO model, ALADIN is updated with the ALARO-0 physics package. This parameterisation has been designed to run at resolutions from the mesoscale to the cloud-resolving scales in a scaleaware manner, based on the modelling approach of the Modular Multiscale Microphysics and Transport (3MT) cloud and precipitation scheme of Gerard and Geleyn (2005), Gerard (2007), and Gerard et al (2009), and has been validated up to a spatial resolution of 4 km for NWP (Gerard et al, 2009;De Meutter et al, 2015) and climate (Hamdi et al, , 2014bDe Troch et al, 2013). The ALARO-0 physics package is coupled to the dynamics via a physics-dynamics interface based on a flux-conservative formulation of the equations proposed by Catry et al (2007).…”

Methodologies to characterize uncertainties in regional reanalyses

“…These three groups (MO, SMHI, HErZ) develop RRAs based on their operational NWP models, which are the Unified Model (Davies et al, 2006), the HARMONIE model (Bubnova et al, 1995;De Troch et al, 2013;Gerard et al, 2009), and the COnsortium for Small-scale MOdelling (COSMO) model (Schättler et al, 2014), respectively. They plan to produce a deterministic run and a set of ensembles of reanalyses with up to 20 members each.…”