Developing effective global strategies for climate mitigation requires an independent assessment of the greenhouse gas emission inventory at the urban scale. In the framework of the Dutch Ruisdael Observatory infrastructure project, we have enhanced the Dutch Atmospheric Large-Eddy-Simulation (DALES) model to simulate carbon dioxide (CO2) plume emission and three-dimensional dispersion within the turbulent boundary layer. The unique ability to explicitly resolve turbulent structures a the hectometer resolution (100 m) makes DALES particularly suitable for detailed realistic simulations of both singular high-emitting point sources and urban emissions, aligning with the goals of Ruisdael Observatory. The model setup involves a high-resolution simulation (100 m × 100 m) covering the main urban area of the Netherlands (51.5–52.5° N, 3.75–6.45° E). The model integrates meteorological forcing from the HARMONIE-AROME weather forecasting model, background CO2 levels from the CAMS reanalysis, and point source emissions and downscaled area emissions derived from the 1 km × 1 km emission inventory from the national registry. The latter are prepared using a sector-specific downscaling workflow, covering major emission categories. Biogenic CO2 exchanges from grasslands and forests are interactively included in the hectometer calculations within the heterogeneous land–surface model of DALES. Our evaluation strategy is twofold, comparing DALES simulations with (i) the state-of-the art LOTOS-EUROS model simulations and (ii) Ruisdael surface observations of the urban background in the Rotterdam area at Westmaas and Slufter and in situ rural Cabauw tower measurements. Our comprehensive statistical analysis confirmed the effectiveness of DALES at modeling the urban-scale CO2 emission distribution and plume dispersion under turbulent conditions but also revealed potential limitations and areas for further improvement. Thus, our new model framework provides valuable insights into the role of anthropogenic and biogenic contributions to local CO2 levels, as well as the transport and dispersion of CO2 emissions. This supports emission uncertainty reduction using atmospheric measurements and contributes to the development of effective regional climate mitigation strategies.

This paper presents the GPU porting through OpenACC directives of the Dutch Atmospheric Large-Eddy Simulation (DALES) application, a high-resolution atmospheric model. The code is written in Fortran 90 and features parallel (distributed) execution through spatial domain decomposition. We assess the performance of the GPU offloading, comparing the time-to-solution on regular and accelerated HPC nodes. A weak scaling analysis is conducted and portability across NVIDIA A100 and H100 hardware is discussed. Finally, we show how targeted kernels can benefit from further optimization with Kernel Tuner, a GPU kernels auto-tuning package.

Marine shallow cumulus clouds have long caused large uncertainty in climate projections. These clouds frequently organize into mesoscale (10-500 km) structures, through two processes that couple the clouds to shallow mesoscale circulations: (i) mesoscale moisture aggregation, and (ii) cold pools, driven locally from rain-evaporation. Since global climate models do not capture these mesoscale processes, while the degree of mesoscale organization is observed to correlate to shortwave cooling, it has been suggested that mesoscale processes modulate the cloud response to global warming. Here, we show that introducing mesoscale dynamics can indeed substantially alter top-of-the-atmosphere radiative budget, if the balance between the two circulations is upset. By homogenizing rain-evaporation across the horizontal domain, we suppress the cold-pool-driven circulations in a large ensemble of large-domain, large-eddy simulations. We find that cold pools reduce mesoscale ascent, thereby arresting a runaway self-aggregation of moisture into very moist regions. This reduces the net rainfall of the cumulus fields, moistens the boundary layer and thus reduces the emission of clear-sky longwave radiation to space, giving an ensemble-averaged warming of 1.88 W/m2. Our results highlight that the proper interplay between mesoscale processes is critical for capturing radiative budgets-especially in kilometer-scale climate models that only partially resolve aggregation and cold pools.

Marine shallow cumulus clouds have long caused large uncertainty in climate projections. Such “trade cumuli” frequently organize into mesoscale (10 to 500 km) structures, through two processes that couple the clouds to shallow mesoscale circulations: i) mesoscale moisture aggregation and ii) cold pools, driven by mesoscale rain evaporation beneath the mesoscale cloud structures. Since global climate models do not capture these mesoscale processes, while the degree of mesoscale organization is observed to correlate to shortwave cooling, it has been suggested that mesoscale processes modulate contemporary estimates of cloud response to global warming. Here, we show that introducing mesoscale dynamics can indeed substantially alter top-of-the-atmosphere radiative budget, if the balance between the two circulations is upset. By homogenizing mesoscale rain evaporation patterns, we suppress the formation and effects of cold pools in a large ensemble of large-domain, large-eddy simulations. The experiments reveal that cold pool dynamics reduce mesoscale ascent in the cloud systems, thereby arresting a runaway self-aggregation of moisture into very moist, cloudy regions that occurs without them. This reduces the net rainfall of the cumulus fields, moistens the cloud layer and thus reduces the emission of clear-sky longwave radiation to space, giving an ensemble-averaged warming of 1.88 W/m2. Our results highlight that the proper interplay between mesoscale processes is critical for capturing radiative budgets-especially in kilometer-scale climate models that only partially resolve shallow cumulus aggregation and cold pools.

We investigate if mesoscale self-organisation of trade cumuli in 150 km-domain large-eddy simulations modifies the top-of-atmosphere radiation budget relative to 10 km-domain simulations, across 77 characteristic, idealized environments. In large domains, self-generated mesoscale circulations produce fewer, larger and deeper clouds, raising the cloud albedo. Yet they also precipitate more than small-domain cumuli, drying and warming the cloud layer, and reducing cloud cover. Consequently, large domains cool slightly less through the shortwave cloud-radiative effect, and slightly more through clear-sky outgoing longwave radiation, for a net cooling (−0.5 W (Formula presented.)). This cooling is generally smaller than the large-domain radiation's sensitivity to large-scale meteorological variability, which is similar in small-domain simulations and observations. Hence, mesoscale self-organisation would not alter weak trade-cumulus feedback estimates previously derived from small-domain simulations. We explain this with a symmetry hypothesis: ascending and descending branches of mesoscale circulations symmetrically increase and reduce cloudiness, weakly modifying the mean radiation budget.

The vertical profiles of the wind speed and direction in atmospheric boundary layers are strongly controlled by turbulence. Most global weather forecast and climate models parameterize the vertical transport of horizontal momentum by turbulent eddies by means of a downgradient eddy diffusion approach, in which the same stability-dependent eddy viscosity profile is applied to both horizontal wind components. In this study we diagnose eddy viscosity profiles from large-eddy simulations of five convective boundary layers with wind shear. Each simulation was forced by the same geostrophic wind of 7.5 (Formula presented.), but with different surface heat fluxes in the range between 0.03 and 0.18 (Formula presented.). We find that the eddy viscosity profiles for the two horizontal wind components differ significantly, in particular, we diagnose negative eddy viscosities, indicating vertical turbulent transport that is counter the mean gradient. This suggests that a purely downgradient diffusion approach for turbulent momentum fluxes is inadequate. A modified solution of the Ekman spiral demonstrates that different eddy viscosity profiles for the two horizontal wind components lead to a different wind profile. To improve parameterizations that apply a downgradient diffusion approach for momentum, correction terms to allow for non-local, boundary-layer scale transport should be incorporated.

The mesoscale self-organization of trade-cumulus cloud fields is a major cloud–climate uncertainty. Cold pools, i.e., pockets of cold, dense air resulting from rain evaporation, are a key mechanism in shaping these dynamics and are controlled by the large-scale forcing. We study the microphysical sensitivity of cloud-field self-organization through cold pools by varying the cloud droplet number concentration (Nc) from 20 to 1000 cm−3 in large-eddy simulations on large 154 km×154 km domains. We find that cold pools exhibit two distinct regimes of mesoscale self-organization. Under very low Nc conditions, cold pools transition from a stage in which they are small and randomly distributed to forming large, long-lived structures that perpetuate due to the collisions of cold pools at their fronts. Under high-Nc conditions, cold pools display strongly intermittent behavior and interact with clouds through small, short-lived structures. Thus, although Nc influences the number of cold pools and, in turn, mesoscale organization, cloud depth, and cloud albedo, we find its effect on cloud cover to be minimal. Comparing the microphysical sensitivity of cold-pool-mediated mesoscale dynamics to the external, large-scale forcing shows that Nc is as important as horizontal wind and large-scale subsidence for trade-cumulus albedo. Our results highlight that cold pools mediate the adjustments of trade-cumulus cloud fields to changes in Nc. Such mesoscale adjustments need to be considered if we are to better constrain the effective aerosol forcing and cloud feedback in the trade-wind regime.

Recent observations of the trade-wind regions highlight the covariability between cold-pool properties and mesoscale cloud organization. Given the covariability of organization with cloud cover and albedo, this suggests a potential impact of cold pools on the cloud radiative effect (CRE). To explore this, we use an ensemble of 103 large-domain, high-resolution, large-eddy simulations and investigate how the variability in cold pools is determined by large-scale external cloud-controlling factors and shaped by processes within the mesoscale. It is demonstrated that the size and frequency of occurrence of cold pools are strongly influenced by the near-surface horizontal wind speed and large-scale subsidence. The temporal evolution of cold pools is strongly correlated with the diurnality in radiation. Even without external variability, we find a strong intermittent behavior in the evolution of cold pools, governed by a complex interplay between cold pools and clouds which expresses itself in the form of shallow squall lines. These squall lines result from precipitating downdrafts, cold pool outflows and the resulting gust fronts, reinforcing parent clouds. Cold pools influence the CRE of trade cumuli, but only when they exist during the day. This emphasizes the importance of the synchronization between cold-pool events and the diurnal cycle of insolation for the dependence of the CRE on cold pools.

The mesoscale self-organization of trade-cumulus cloud fields is a major cloud-climate uncertainty. Cold pools, i.e. pockets of cold, dense air resulting from rain evaporation, are a key mechanism in shaping these dynamics and are controlled by the large-scale forcing. We study the microphysical sensitivity of cloud-field self-organization through cold pools by varying cloud-droplet number concentration Nc from 20 to 1000 /cm3 in large-eddy simulations on large 154×154 km2-domains. We find that cold pools exhibit two distinct regimes of mesoscale self-organization. In very low-Nc conditions, cold pools transition from a stage where they are small and randomly distributed to forming large, long-lived structures that perpetuate due to the collisions of cold pools at their fronts. Under high-Nc conditions, cold pools display strongly intermittent behaviour and interact with clouds through small, short-lived structures. While Nc thus influences the number of cold pools and, in turn, mesoscale organization, cloud depth, and cloud albedo, we find its effect on cloud cover to be minimal. Comparing the microphysical sensitivity of cold-pool-mediated mesoscale dynamics to the external, large-scale forcing shows that Nc is as important as horizontal wind and large-scale subsidence for trade-cumulus albedo. Our results highlight that cold pools mediate adjustments of trade-cumulus cloud fields to changes in Nc. Such mesoscale adjustments need to be considered if we are to better constrain the effective aerosol forcing and cloud feedback in the trade-wind regime.

Recent observations of the trades highlight the covariability between cold pool (CP) properties and cloud cover, suggesting a potential impact of CPs on the cloud radiative effect (CRE). To explore this, we use an ensemble of 103 large-domain, high-resolution, large-eddy simulations (Cloud Botany). We investigate the extent to which the variability in CPs is driven by external conditions or convective self-organization. Our findings show that CPs are notably controlled by large-scale conditions, specifically (horizontal) wind speed and subsidence. The temporal evolution of CPs is tightly related to the diurnality in radiation. To understand the extent to which CPs vary with self-organization, we switch off the diurnality in radiation. Despite the absence of the diurnal cycle, CP time series still exhibit fluctuations. These fluctuations result from the recharge-discharge of thermodynamic and dynamic properties of the sub-cloud layer owing to CP-cloud interactions. Our results demonstrate that circulations induced by CPs reinforce the parent clouds, resulting in deepening and scale growth, followed by mesoscale arcs enclosing clear-sky areas. Finally, we show that CPs influence CRE, but only when they exist during the day. Our findings emphasize the importance of the relationship between the timescales of self-organization and the diurnal cycle of external conditions, greatly influencing the CRE dependency on self-organizing CPs.

Open boundary conditions were developed for atmospheric large-eddy simulation (LES) models and implemented into the Dutch Atmospheric Large-Eddy Simulation model. The implementation was tested in a “Big Brother”-like setup, in which the simulation with open boundary conditions was forced by an identical control simulation with periodic boundary conditions. The results show that the open boundary implementation has minimal influence on the solution. Both the mean state and the turbulent structures are close to the control simulation, and disturbances at the in- and outflow boundaries are negligible. To emulate a setup in which the LES is coupled to a coarser model, the influence of coarse boundary input was tested by smoothing the output of the periodic control simulation both temporally and spatially before feeding it as input to the simulation with open boundary conditions. When smoothing is applied over larger spatial and longer temporal scales, disturbances start to form at the inflow boundary and an area exists where turbulence needs to develop. Adding synthetic turbulence to the smoothed input reduces the size of this area and the magnitude of the disturbances.

Recent observations of the trade-wind regions highlight the covariability between cold-pool properties and mesoscale cloud organization. Given the covariability of organization with cloud cover and albedo, this suggests a potential impact of cold pools on the cloud radiative effect (CRE). To explore this, we use an ensemble of 103 large-domain, high-resolution, large-eddy simulations and investigate how the variability in cold pools is determined by large-scale external cloud-controlling factors (CCFs) and shaped by processes within the mesoscale. It is demonstrated that the size and frequency of occurrence of cold pools are strongly influenced by the near-surface horizontal wind speed and large-scale subsidence. The temporal evolution of cold pools is strongly correlated with the diurnality in radiation. Even without external variability, we find a strong intermittent behaviour in the evolution of cold pools, governed by a complex interplay between cold pools and clouds which expresses itself in the form of shallow squall lines. These squall lines result from precipitating downdrafts, cold pool outflows and the resulting gust fronts, reinforcing parent clouds. Cold pools influence the CRE of trade cumuli, but only when they exist during the day. This emphasizes the importance of the synchronization between cold-pool events and the diurnal cycle of insolation for the dependence of the CRE on cold pools.

Small shallow cumulus clouds (less-than 1 km) over the tropical oceans appear to possess the ability to self-organize into mesoscale (10–100 km) patterns. To better understand the processes leading to such self-organized convection, we present Cloud Botany, an ensemble of 103 large-eddy simulations on domains of 150 km, produced by the Dutch Atmospheric Large Eddy Simulation model on supercomputer Fugaku. Each simulation is run in an idealized, fixed, larger-scale environment, controlled by six free parameters. We vary these over characteristic ranges for the winter trades, including parameter combinations observed during the EUREC4A (Elucidating the role of clouds–circulation coupling in climate) field campaign. In contrast to simulation setups striving for maximum realism, Cloud Botany provides a platform for studying idealized, and therefore more clearly interpretable causal relationships between conditions in the larger-scale environment and patterns in mesoscale, self-organized shallow convection. We find that any simulation that supports cumulus clouds eventually develops mesoscale patterns in their cloud fields. We also find a rich variety in these patterns as our control parameters change, including cold pools lined by cloudy arcs, bands of cross-wind clouds and aggregated patches, sometimes topped by thin anvils. Many of these features are similar to cloud patterns found in nature. The published data set consists of raw simulation output on full 3D grids and 2D cross-sections, as well as post-processed quantities aggregated over the vertical (2D), horizontal (1D) and all spatial dimensions (time-series). The data set is directly accessible from Python through the use of the EUREC4A intake catalog.

HPCI User Report for the project Organization of shallow cumulus clouds on the Fugaku supercomputer.

A new generation of operational atmospheric models operating at horizontal resolutions in the range 200 m ∼ 2 km is becoming increasingly popular for operational use in numerical weather prediction and climate applications. Such grid spacings are becoming sufficiently fine to resolve a fraction of the turbulent transports. Here we analyze Large-eddy simulation results of a convective boundary layer obtained by coarsening horizontal grid spacings up to 800 m. The aim is to explore the dependency of the mean state and turbulent fluxes on the grid resolution. Both isotropic and anisotropic eddy diffusion approaches are evaluated, where in the latter case the horizontal and vertical eddy diffusivities differ in accord with their horizontal and vertical grid spacings. For coarsening horizontal grid sizes entrainment at the top of the boundary layer tends to get slightly enhanced for isotropic diffusion, whereas for the anisotropic diffusion approach the vertically well-mixed boundary-layer structure becomes severely degraded. An analysis of the energy spectrum shows that anisotropic diffusion causes relatively more dissipation of variance at smaller length scales. This leads, in turn, to a shift of spectral energy toward larger length scales that also becomes apparent from a rather different kind of spatial organization of convection. The present study therefore suggests that details with regards to the representation of processes at small scales might impact the organization at length scales much larger than the smallest scales that can be resolved by the model.

A new generation of operational atmospheric models operating at horizontal resolutions in the range 200 m up to ~ 2 km is becoming increasingly popular for operational use in numerical weather prediction and climate applications. Such grid spacings are becoming sufficiently fine to resolve a fraction of the turbulent transports. Here we analyze LES results of a convective boundary layer obtained by coarsening horizontal grid spacings up to 800 m. The aim is to explore the dependency of the mean state and turbulent fluxes on the grid resolution. Both isotropic and anisotropic eddy diffusion approaches are evaluated, where in the latter case the horizontal and vertical eddy diffusivities differ in accord with their horizontal and vertical grid spacings. For coarsening horizontal grid sizes entrainment at the top of the boundary layer tends to get slightly enhanced for isotropic diffusion. An analysis of the energy spectrum shows that anisotropic diffusion causes relatively more dissipation of variance at smaller length scales. This leads, in turn, to a shift of spectral energy towards larger length scales. This can also be clearly seen from the different kinds of spatial organization. The present study therefore suggests that details with regards to the representation of processes at small scales might impact the organization at length scales much larger than the smallest scales that can be resolved by the model.

In atmospheric modeling, superparameterization (SP) has gained popularity as a technique to improve cloud and convection representations in large-scale models by coupling them locally to cloud-resolving models. We show how the different representations of cloud water in the local and the global models in SP lead to a suppression of cloud advection and ultimately to a systematic underrepresentation of the cloud amount in the large-scale model. We demonstrate this phenomenon in a regional SP experiment with the global model OpenIFS coupled to the local model Dutch Atmospheric Large Eddy Simulation, as well as in an idealized setup, where the large-scale model is replaced by a simple advection scheme. As a starting point for mitigating the problem of suppressed cloud advection, we propose a scheme where the spatial variability of the local model's total water content is enhanced in order to match the global model's cloud condensate amount. The proposed scheme enhances the cloud condensate amount in the test cases, however a large discrepancy remains, caused by rapid dissipation of the clouds added by the proposed scheme.

In order to eliminate climate uncertainty w.r.t. cloud and convection parametrizations, superpramaterization (SP) [1] has emerged as one of the possible ways forward. We have implemented (regional) superparametrization of the ECMWF weather model OpenIFS [2] by cloud-resolving, three-dimensional large-eddy simulations. This setup, described in [3], contains a two-way coupling between a global meteorological model that resolves large-scale dynamics, with many local instances of the Dutch Atmospheric Large Eddy Simulation (DALES) [4], resolving cloud and boundary layer physics. The model is currently prohibitively expensive to run over climate or even seasonal time scales, and a global SP requires the allocation of millions of cores. In this paper, we study the performance and scaling behavior of the LES models and the coupling code and present our implemented optimizations. We mimic the observed load imbalance with a simple performance model and present strategies to improve hardware utilization in order to assess the feasibility of a world-covering superparametrization. We conclude that (quasi-)dynamical load-balancing can significantly reduce the runtime for such large-scale systems with wide variability in LES time-stepping speeds.

In this study, we investigate uncertainties in a large eddy simulation of the atmosphere, employing modern uncertainty quantification methods that have hardly been used yet in this context. When analysing the uncertainty of model results, one can distinguish between uncertainty related to physical parameters whose values are not exactly known, and uncertainty related to modelling choices such as the selection of numerical discretization methods, of the spatial domain size and resolution, and the use of different model formulations. While the former kind is commonly studied e.g. with forward uncertainty propagation, we explore the use of such techniques to also assess the latter kind. From a climate modelling perspective, uncertainties in the convective response and cloud formation are of particular interest, since these affect the cloud-climate feedback, one of the dominant sources of uncertainty in current climate models. Therefore we analyse the DALES model in the RICO case, a well-studied convection benchmark. We use the VECMA toolkit for uncertainty propagation, assessing uncertainties stemming from physical parameters as well as from modelling choices. We find substantial uncertainties due to small random initial state perturbations, and that the choice of advection scheme is the most influential of the modelling choices we assessed. This article is part of the theme issue 'Reliability and reproducibility in computational science: implementing verification, validation and uncertainty quantification in silico'.

We present the VECMA toolkit (VECMAtk), a flexible software environment for single and multiscale simulations that introduces directly applicable and reusable procedures for verification, validation (V&V), sensitivity analysis (SA) and uncertainty quantication (UQ). It enables users to verify key aspects of their applications, systematically compare and validate the simulation outputs against observational or benchmark data, and run simulations conveniently on any platform from the desktop to current multi-petascale computers. In this sequel to our paper on VECMAtk which we presented last year [1] we focus on a range of functional and performance improvements that we have introduced, cover newly introduced components, and applications examples from seven different domains such as conflict modelling and environmental sciences. We also present several implemented patterns for UQ/SA and V&V, and guide the reader through one example concerning COVID-19 modelling in detail. This article is part of the theme issue 'Reliability and reproducibility in computational science: implementing verification, validation and uncertainty quantification in silico'.