II: Scenarios & Global Climate Model Selection


February 6, 2024


An overview of the II proposed climate scenarios, future periods and selection of global climate models (GCMs) for the II simulations. This report describes these choices and the methodology used to select the GCMs, future periods and scenarios.

1 Introduction

The main aim of II is to close the gap between regional climate model information and local impacts in order to provide climate services to support climate adaptation and mitigation. To achieve this ambition the project will simulate future climates in different scenarios and translate this information into local impacts. The local climate can be simulated using regional climate models (RCMs). Regional climate models downscale (“zoom in on”) global climate model (GCMs) on a smaller domain (e.g. Belgium). RCMs increase the resolution of the climate simulations which results in a more detailed representation of the climate, see Figure 1.

The \(6^{th}\) IPCC assessment report (AR6) makes use of 60 different GCMs from the latest Coupled Model Intercomparison Project Phase 6 (CMIP6) and 24 RCMs from the Coordinated Regional Climate Downscaling Experiment (CORDEX) (Gutiérrez & Treguier, 2021). Due to their higher resolution, RCMs better represent extreme precipitation events at sub-daily time scales (Prein et al., 2015; Termonia, Van Schaeybroeck, et al., 2018) and are therefore well suited to study extreme events which are expected to increase in intensity and frequency in the future (Masson-Delmotte et al., 2021). In the II project we will use 3 RCMs at 4km resolution over Belgium: ALARO, COSMO-CLM and MAR, and each RCM will downscale (zoom) multiple GCMs.

All three RCMs were used in the first project and have been improved since then. MAR (Modèle Atmosphérique Régional) (Wyard et al., 2017) is a hydrostatic RCM developed by the Liège University. The latest version of MAR (v3.14) includes a new convection scheme (Doutreloup et al., 2019), parametrization of the urban heat island, and improved simulation of energy fluxes and vegetation seasonality. COSMO-CLM is a non-hydrostatic RCM that is maintained by the Climate Limited-area Modeling (CLM) community. This community continuously improves the model and the latest version (COSMO-CLM 6.0), which uses the TERRA-URB (Trusilova et al., 2015) land-surface scheme, will be used in the II project. Finally, ALARO is an RCM developed by the international ALADIN consortium (Termonia, Fischer, et al., 2018). The latest version (ALARO-1) will be used together with a surface scheme SURFEX (Hamdi et al., 2014; Masson et al., 2013) which better simulates urban areas.

High-resolution RCM-simulations are computationally expensive and therefore, combining all RCMs and GCMs is not feasible. Additionally, the computational cost also limits the number of future time periods one can reasonably simulate. Therefore, within II, we are forced to select 7 GCMs that will each by downscaled by at least one of the 3 RCMs for the current climate and two 20-year future periods.

RCM-simulations are not only computationally expensive, they are also time consuming to run. Once these simulations have started, decisions on the scenarios and the selected GCMs to downscale (zoom) can not be changed. Therefore, it is important that the chosen scenarios and selected GCMs are in line with the project goals and the stakeholders needs. This report describes these choices using scientific arguments and stakeholder priorities.

In a later stage, the II project will help translate these RCM simulations to local impacts. This will be done by using the RCM simulations as input for impact models. Two hydrological impact models (SCHEME and WOLF) will be used to simulate flooding in river basins. The urban model URBCLIM will make 100m resolution simulation that can be used alongside the urban observation campaign to evaluate the micro-scale (e.g. buildings and trees) impact of cities on the local climate. Major Belgian cities will also be simulated at a 1km resolution using the urban model SURFEX. Additionally, the project aims to co-create “tales of future weather” together with core stakeholders which will enable stakeholder-relevant climate impacts to be investigated.