Non-Hydrostatic Climate Modelling, Part II (NHCM-2)

Due to the increasing demand for local scale climate change information, regional climate models (RCMs) are increasingly operated with higher resolved resolutions. Modern RCMs are able to capture many regional climate processes and they cover the meso-ß scales (20 km to 200 km) sufficiently enough for applications in climate research. Based on the success in numerical weather prediction (NWP) and supported by the general progress in computing technology, RCMs are now starting to approach the meso- scales (2 km to 20 km). This jump in scales, however, is not straightforward. Relevant processes (e.g. deep convection) on former unresolved (parameterised) scales become resolved, and it is largely unclear how current RCMs (originally developed for larger scales) are able to capture climate processes and their interplay throughout the scales. In complex terrain, where mountains have substantial effects on weather and climate, this becomes even more important due to the influence of orography. In addition, model evaluation becomes challenging: observational data sufficiently covering the natural variability only exist in exceptional cases (e.g. in special observation campaigns) and shifts in time and/or space between modelled and observed quantities (double penalty problem) limit the application of traditional error statistics. In the previous project "Non-Hydrostatic Climate Modelling (NHCM-1)", funded by the Austrian Science Fund (FWF) (project number P19619-N10), first test simulations in climate mode on scales where deep convection becomes resolved (=3 km grid spacing) have been conducted and analysed in the European Alpine region. The project focused on the exploration of error ranges of near surface variables and on the detection/assessment of added value by implementing both, reference data from operational nowcasting systems and evaluation techniques from NWP avoiding the double penalty problem. Based on the scientific success of NHCM-1, this proposed follow-up project NHCM-2 brings together latest developments in climate research and NWP. It (1) investigates the ability of state-of-the-art non-hydrostatic RCMs operated at convection-resolving scales to capture important orographic-induced climate processes in the European Alpine region on regional (meso-ß) scales, (2) contributes to the improvement of RCMs with respect to these climate processes, (3) continues introducing advanced analysis tools and highly resolved gridded reference data from NWP in climate research, and (4) aims to enable next generation long-term climate simulations, i.e. convection-resolving climate simulations (CRCSs), in the Alpine region.

How well do current regional climate models capture the influences of the Alpine massif on atmospheric processes (shielding effect, orographic precipitation)? This urgent question is increasingly confronting climate research with new tasks; at long last, it is important to find out in what quality conclusions about the future climate can be made. With the development of climate models, which are now able to resolve individual thunderstorms due to increasingly smaller grid spacing, the requirements for model evaluation are also increasing and current high-performance computers are brought to the limits of their capacities. In the framework of the project "Non-Hydrostatic Climate Modelling, II" (NHCM-2), not only evaluation methods from weather prediction were adapted for climate research, but also new process-based methods for model evaluation were developed that allow to isolate recurring weather phenomena (e.g. fronts) in climate model data and compare them with observation data in a statistical manner. In this way it first has been shown that the reproducibility of precipitation in the Alpine region in climate models is significantly improved as model resolution increases, in particular due to the strong influence of the terrain. It has also been shown that climate models on convection-permitting scales (~3 km grid spacing) have significant (physical process-related) advantages over coarser resolved models (~12 km grid spacing), especially in weather situations with strong thunderstorms. The influences of the terrain on rainfall in the Eastern Alps (shielding effect, elevation dependency) are very well reproduced by these so-called "convection-permitting" models up to an altitude of ~1400 m, above the quality of the models remains speculative due to observation errors. In NHCM-2, the further development of these convection-permitting models not only laid the basis for next-generation climate projections, but also supported procurement processes to cover the enormous technical effort required.