Consortium For Small-scale Modeling Research Articles

Machine learning approaches to predict the execution time of the meteorological simulation software COSMO

AbstractPredicting the execution time of weather forecast models is a complex task, since these models are usually performed on High Performance Computing systems that require large computing capabilities. Indeed, a reliable prediction can imply several benefits, by allowing for an improved planning of the model execution, a better allocation of available resources, and the identification of possible anomalies. However, to make such predictions is usually hard, since there is a scarcity of datasets that benchmark the existing meteorological simulation models. In this work, we focus on the runtime predictions of the execution of the COSMO (COnsortium for SMall-scale MOdeling) weather forecasting model used at the Hydro-Meteo-Climate Structure of the Regional Agency for the Environment and Energy Prevention Emilia-Romagna. We show how a plethora of Machine Learning approaches can obtain accurate runtime predictions of this complex model, by designing a new well-defined benchmark for this application task. Indeed, our contribution is twofold: 1) the creation of a large public dataset reporting the runtime of COSMO run under a variety of different configurations; 2) a comparative study of ML models, which greatly outperform the current state-of-practice used by the domain experts. This data collection represents an essential initial benchmark for this application field, and a useful resource for analyzing the model performance: better accuracy in runtime predictions could help facility owners to improve job scheduling and resource allocation of the entire system; while for a final user, a posteriori analysis could help to identify anomalous runs.

Future change of humid heat extremes and population exposure in Turkey

AbstractGlobal climate projections show that humid heat extremes will expand toward the higher latitudes, making the midlatitudes hotspots for these extremes. Therefore, a thorough explanation of their regional characteristics becomes crucial, given that the changes in these extremes can potentially render a large proportion of the global population at risk. Here, we perform the first analysis of historical and projected changes in the intensity and frequency of humid heat extremes and quantify the population exposure to these extremes in Turkey, using long‐term simulations from the non‐hydrostatic mesoscale model of Consortium for Small‐scale Modeling (COSMO‐CLM) under the RCP8.5 emission scenario. We portray not only the nationwide changes in the humid heat extremes and population exposure but also their regional aspects by exploiting the K‐means clustering algorithm. Our results suggest significant future increases in the intensity and frequency of these extremes over a wide geographical area, which includes the surroundings of Adana, Antalya, Izmir, Sakarya, Ordu and Diyarbakir, most of which are coastal locations. Over most of these regions, severe humid heat stress is expected to last nearly a month every year, with almost 56% of the land area is projected to experience local historical upper tail heat stress conditions for at least an additional 10 consecutive hours. Further, we explicate a significant rise in the number of people exposed to severe humid heat stress, concentrated along most coastal regions, by as much as 1.6 million person‐days. More than 20% of Turkey's population may confront severe humid heat stress for at least 1 h, with that percentage falling to 4.15% for at least five consecutive hours, which indicates that people will not only endure more intense humid heat stress but also be exposed to these conditions consecutively over a period of many hours.

Influence of air flow features on alpine wind energy potential

Wind energy is one of the potential options to fill the gap in renewable energy production in Switzerland during the winter season when the energy demand exceeds local production capacities. With likely further rising energy consumption in the future, the winter energy deficit may further increase. However, a reliable assessment of wind energy potential in complex terrain remains challenging. To obtain such information, numerical simulations are performed using a combination of the “Consortium for Small-scale Modeling” and “Weather Research and Forecasting” (COSMO-WRF) models initialized and driven by COSMO-1E model, which allows us to simulate the influence of topography at a horizontal resolution of 300 m. Two LiDAR measurement campaigns were conducted in the regions of Lukmanier Pass and Les Diablerets, Switzerland. Observational LiDAR data and measurements from nearby wind sensor networks are used to validate the COSMO-WRF simulations. The simulations show an improved representation of wind speed and direction near the ground compared to COSMO-1E. However, with increasing height and less effect of the terrain, COSMO-WRF tends to overestimate the wind speeds, following the bias that is already present in COSMO-1E. We investigate two characteristic mountain–terrain flow features, namely waves and Foehn. The effect of mountain-induced waves of the flow is investigated through an event that occurred in the area of Diablerets. One-year analysis for the frequency of conditions that are favorable for mountain wave formation is estimated. The Foehn impact on wind was observed in the Lukmanier domain. We attempt quantification of the probability of occurrence using the Foehnix model. The result shows a high probability of Foehn occurrence during the winter and early spring seasons. Our study highlights the importance of incorporating complex terrain-related meteorological events into the wind energy assessment. Furthermore, for an accurate assessment of wind speed in complex terrain, our study suggests the necessity to have a better representation of the topography compared to COSMO-1E.

Evaluation of COSMO-CLM Model Parameter Sensitivity in the Study of Extreme Events across the Eastern Region of India

The present study aims to identify the parameters from the Consortium for Small-scale Modelling in CLimate Mode (COSMO-CLM) regional climate model that strongly controls the prediction of extreme events over West Bengal and the adjoining areas observed between 2013 to 2018. Metrics, namely Performance Score (PS) screen out the most persuasive parameter on model output. Additionally, the Performance Index (PI) measure the reliability of the model and Skill Score (SS) establishes the model performance against the reference simulation leading to the optimization of the model for a given variable. In this study, parameter screening for four output variables such as 2m-temperature, surface latent heat flux, precipitation and cloud cover of COSMO-CLM is accomplished. For heat wave simulations, 2m-temperature and surface latent heat flux are explored whereas cloud cover and precipitation are examined for extreme rainfall events. A total of 25 adjustable parameters representing the following parameterization schemes: turbulence, land surface process, microphysics, convection, radiation and soil. Out of the six parameterization schemes, the scaling factor of the laminar boundary layer for heat (rlam_heat) and the ratio of laminar scaling factors for heat over sea and land (rat_sea) from the land surface process is sensitive to SLH, TP. The exponent to get the effective surface area (e_surf) from the land surface has a large impact on 2m-temperature. A few parameters from microphysics (cloud ice threshold for auto conversion), convection (mean entrainment rate for shallow convection) and radiation (parameter for computing the amount of cloud cover in saturated conditions) play a significant role in producing TP, and TCC fields. It is evident from the results that the parameter sensitivities on model performance depend on the choice of the meteorological field. Furthermore, in almost all input model parameters, the model performance reveals the opposite character in different domains for a given meteorological field.

Comparison of COSMO and ICON-LAM high-resolution numerical forecast for Romanian territory: Case studies and evaluation

This paper aims to offer the first detailed inter-comparison of the performance from the numerical weather prediction models Consortium for Small-scale Modeling (COSMO) and Icosahedral Nonhydrostatic General Circulation Model, Limited Area Mode (ICON-LAM), integrated for the Romanian territory at the horizontal resolution of 2.8 km. As the ICON numerical model is set to replace COSMO, a comparison is carried out between the operational configurations of the two models. First, two cases with severe weather (strong atmospheric instability, observed heavy precipitation, and strong wind) are analyzed: February 3-6, 2020 (heavy precipitation [snow] and very strong wind) and May 3-5, 2020 (heavy precipitation). This is followed by a statistical inter-comparison between the two models for the summer of 2020 (June, July, August). A statistical evaluation of the forecast quality from the two models is performed objectively, through statistical scores computed using surface observations from all available meteorological stations on the Romanian territory. Stratifications depending on station altitude and location are carried out, with emphasis on stations of particular interest, depending on the evolution of the synoptic situation. Following the statistical evaluation, results for the precipitation forecast do not show significant improvement in favor of either model. However, ICON-2.8 km mostly outperforms COSMO-2.8 km for surface parameters.

Water isotopic characterisation of the cloud–circulation coupling in the North Atlantic trades – Part 1: A process-oriented evaluation of COSMOiso simulations with EUREC4A observations

Abstract. Naturally available, stable, and heavy water molecules such as HDO and H218O have a lower saturation vapour pressure than the most abundant light water molecule H216O; therefore, these heavy water molecules preferentially condense and rain out during cloud formation. Stable water isotope observations thus have the potential to provide information on cloud processes in the trade-wind region, in particular when combined with high-resolution model simulations. In order to evaluate this potential, nested COSMOiso (isotope-enabled Consortium for Small Scale Modelling; Steppeler et al., 2003; Pfahl et al., 2012) simulations with explicit convection and horizontal grid spacings of 10, 5, and 1 km were carried out in this study over the tropical Atlantic for the time period of the EUREC4A (Elucidating the role of clouds-circulation coupling in climate; Stevens et al., 2021) field experiment. The comparison to airborne in situ and remote sensing observations shows that the three simulations are able to distinguish between different mesoscale cloud organisation patterns as well as between periods with comparatively high and low rain rates. Cloud fraction and liquid water content show a better agreement with aircraft observations with higher spatial resolution, because they show strong spatial variations on the scale of a few kilometres. A low-level cold-dry bias, including too depleted vapour in the subcloud and cloud layer and too enriched vapour in the free troposphere, is found in all three simulations. Furthermore, the simulated secondary isotope variable d-excess in vapour is overestimated compared to observations. Special attention is given to the cloud base level, which is the formation altitude of shallow cumulus clouds. The temporal variability of the simulated isotope variables at cloud base agrees reasonably well with observations, with correlations of the flight-to-flight data as high as 0.7 for δ2H and d-excess. A close examination of isotopic characteristics under precipitating clouds, non-precipitating clouds, clear-sky and dry-warm patches at the altitude of cloud base shows that these different environments are represented faithfully in the model with similar frequencies of occurrence, isotope signals, and specific-humidity anomalies as found in the observations. Furthermore, it is shown that the δ2H of cloud base vapour at the hourly timescale is mainly controlled by mesoscale transport and not by local microphysical processes, while the d-excess is mainly controlled by large-scale drivers. Overall, this evaluation of COSMOiso, including the isotopic characterisation of different cloud base environments, suggests that the simulations can be used for investigating the role of atmospheric circulations on different scales for controlling the formation of shallow cumulus clouds in the trade-wind region, as will be done in part 2 of this study.

Identification of high-wind features within extratropical cyclones using a probabilistic random forest – Part 2: Climatology over Europe

Abstract. Strong winds associated with extratropical cyclones are one of the most dangerous natural hazards in Europe. These high winds are mostly associated with five mesoscale features: the warm (conveyor belt) jet (WJ); the cold (conveyor belt) jet (CJ); cold frontal convection (CFC); strong cold-sector (CS) winds; and, in some cases, the sting jet (SJ). The timing within the cyclone's life cycle, the location relative to the cyclone core and further characteristics differ between these features and, hence, likely also their associated forecast errors. In Part 1 of this study (Eisenstein et al., 2022a), we introduced the objective and flexible identification tool RAMEFI (RAndom-forest-based MEsoscale wind Feature Identification), which distinguishes between the WJ, CFC and CS as well as CJ and SJ combined. RAMEFI is based on a probabilistic random forest trained on station observations of 12 storm cases over Europe. Being independent of spatial distribution, RAMEFI can also be applied to gridded data. Here, we use RAMEFI to compile a climatology over 19 extended winter seasons (October–March 2000–2019) based on high-resolution regional reanalyses of the German Consortium for Small-scale Modelling (COSMO) model over Europe. This allows the first ever long-term objective statistical analysis of the mesoscale wind features, including their occurrence frequency, geographical distribution and characteristics. For western and central Europe, we demonstrate that the CS is prominent in most winter storms, while CFC is the least common cause of high winds, both in terms of frequency and affected area. However, probably due to convective momentum transport, CFC is on average the cause of the highest gusts after the CJ and has the highest gust factor. As expected, CFC high-wind areas show high levels of humidity and overcast conditions. In contrast, the CS is characterised by sunnier conditions interspersed by patchy cumulus clouds, leading to a broader cloud cover distribution than for other features. The WJ produces the weakest winds on average but affects a larger area than CJ. Central Europe is more strongly affected by WJ and CFC winds, while the CJ usually occurs farther north over the North and Baltic seas, northern Germany, Denmark and southern Scandinavia. System-relative composites show that the WJ and CFC tend to occur earlier in the cyclone life cycle than the CJ and CS. Consistently, the CS is the most common cause of high winds over eastern Europe, where cyclones tend to occlude, represented by a narrowing warm sector and weakening cold front. The WJ mostly occurs within the south-eastern quadrant of a cyclone bordered by the narrow CFC in the west. However, the location of CFC varies greatly between cases. The CS occurs in the south-western quadrant, while the CJ appears closer to the cyclone centre, sometimes stretching into the south-eastern quadrant. This objective climatology largely confirms previous, more subjective investigations but puts these into climatological context. It allows a more detailed analysis of feature properties and provides a solid foundation for model assessment and forecast evaluation in future studies.

Exploring hail and lightning diagnostics over the Alpine-Adriatic region in a km-scale climate model

Abstract. The north and south of the Alps, as well as the eastern shores of the Adriatic Sea, are hot spots of severe convective storms, including hail and lightning associated with deep convection. With advancements in computing power, it has become feasible to simulate deep convection explicitly in climate models by decreasing the horizontal grid spacing to less than 4 km. These kilometer-scale models improve the representation of orography and reduce uncertainties associated with the use of deep convection parameterizations. In this study, we perform km-scale simulations for eight observed cases of severe convective storms (seven with and one without observed hail) over the Alpine-Adriatic region. The simulations are performed with the climate version of the regional model Consortium for Small-scale Modeling (COSMO) that runs on graphics processing units (GPUs) at a horizontal grid spacing of 2.2 km. To analyze hail and lightning we have explored the hail growth model (HAILCAST) and lightning potential index (LPI) diagnostics integrated with the COSMO-crCLIM model. Comparison with available high-resolution observations reveals good performance of the model in simulating total precipitation, hail, and lightning. By performing a detailed analysis of three of the case studies, we identified the importance of significant meteorological factors for heavy thunderstorms that were reproduced by the model. Among these are the moist unstable boundary layer and dry mid-level air, the topographic barrier, as well as an approaching upper-level trough and cold front. Although COSMO HAILCAST tends to underestimate the hail size on the ground, the results indicate that both HAILCAST and LPI are promising candidates for future climate research.

The Sensitivity of the Icosahedral Non-Hydrostatic Numerical Weather Prediction Model over Greece in Reference to Observations as a Basis towards Model Tuning

The ICON (icosahedral non-hydrostatic) numerical weather prediction model (NWP)’s sensitivity is evaluated for the geographical area of Greece. As ICON model has recently been endorsed operationally by the Members of the COSMO (Consortium for Small-scale Modeling); this attempt is in line with the further understanding of the model features, especially in the considered domain, characterized by a complex orography as well as an almost equally partitioned land–sea surface area. An extraordinary number of 24 model parameters have been tested for the whole 2020 year in reference to 88 Greek meteorological stations, with regard to the standard synoptic meteorological variables of 2 m temperature, 2 m minimum and maximum temperatures, dew-point temperature, 10 m wind intensity and 12 h accumulated precipitation. For these variables, the model sensitivity is given in terms of the annual average of all stations for the fifth lead day of the model runs when the sensitivity is expected to reach its peak. It was found that there is a considerable impact regarding the minimum and maximum values for many of the examined parameters in reference to their default values, and consideration is given to a heuristic recommendation on the selection of the most sensitive parameters.

Irradiance and cloud optical properties from solar photovoltaic systems

Abstract. Solar photovoltaic power output is modulated by atmospheric aerosols and clouds and thus contains valuable information on the optical properties of the atmosphere. As a ground-based data source with high spatiotemporal resolution it has great potential to complement other ground-based solar irradiance measurements as well as those of weather models and satellites, thus leading to an improved characterisation of global horizontal irradiance. In this work several algorithms are presented that can retrieve global tilted and horizontal irradiance and atmospheric optical properties from solar photovoltaic data and/or pyranometer measurements. The method is tested on data from two measurement campaigns that took place in the Allgäu region in Germany in autumn 2018 and summer 2019, and the results are compared with local pyranometer measurements as well as satellite and weather model data. Using power data measured at 1 Hz and averaged to 1 min resolution along with a non-linear photovoltaic module temperature model, global horizontal irradiance is extracted with a mean bias error compared to concurrent pyranometer measurements of 5.79 W m−2 (7.35 W m−2) under clear (cloudy) skies, averaged over the two campaigns, whereas for the retrieval using coarser 15 min power data with a linear temperature model the mean bias error is 5.88 and 41.87 W m−2 under clear and cloudy skies, respectively. During completely overcast periods the cloud optical depth is extracted from photovoltaic power using a lookup table method based on a 1D radiative transfer simulation, and the results are compared to both satellite retrievals and data from the Consortium for Small-scale Modelling (COSMO) weather model. Potential applications of this approach for extracting cloud optical properties are discussed, as well as certain limitations, such as the representation of 3D radiative effects that occur under broken-cloud conditions. In principle this method could provide an unprecedented amount of ground-based data on both irradiance and optical properties of the atmosphere, as long as the required photovoltaic power data are available and properly pre-screened to remove unwanted artefacts in the signal. Possible solutions to this problem are discussed in the context of future work.

Simulating Hail and Lightning Over the Alpine Adriatic Region—A Model Intercomparison Study

AbstractHail is a significant convective weather hazard, often causing considerable crop and property damage across the world. Although extremely damaging, hail still remains a challenging phenomenon to model and forecast, given the limited computational resolution and the gaps in understanding the processes involved in hail formation. Here, eight hailstorms occurring over the Alpine‐Adriatic region are analyzed using simulations with the Weather Research and Forecasting (WRF) and the Consortium for Small Scale Modeling (COSMO) models, with embedded HAILCAST and Lightning Potential Index (LPI) diagnostics at kilometer‐scale grid spacing (∼2.2 km). In addition, a systematic model intercomparison study is performed to investigate the ability of the different modeling systems in reproducing such convective extremes, and to further assess the uncertainties associated with simulations of such localized phenomena. The results are verified by hailpad observations over Croatia, radar estimates of hail over Switzerland, and lightning measurements from the LINET network. The analysis reveals that both HAILCAST and LPI are able to reproduce the affected area and intensities of hail and lightning. Moreover, hail and lightning fields produced by both models are similar, although a slight tendency of WRF to produce smaller hail swaths with larger hailstones and higher LPI compared to COSMO is visible. It is found that these differences can be explained by systematic differences in vertical profiles of microphysical properties and updraft strength between the models. Overall, results are promising and indicate that both HAILCAST and LPI could be valuable tools for real‐time forecasting and climatological assessment of hail and lightning in current and changing climate.

N°368 – Non-neurologic diseases mimicking stiff person syndrome

In climate modelling, a critical issue is the accurate representation of urban areas in space and variation over time. Urban features may be parameterised through urban canopy parameters (UCPs), including the impervious surface area (ISA). One of the aims of the Consortium for Small-scale Modelling (COSMO) and the CLM-Community is to improve urban numerical weather prediction (NWP) and climate simulations by implementing the urban canopy scheme TERRA_URB in the ICOsahedral Nonhydrostatic (ICON) modelling framework. To harmonise and update the datasets adopted for the UCPs, the Consortium has launched the Priority Project “City Induced Temperature change Through A'dvanced modelling” (PP CITTÀ). Within this framework, the study explores recent and detailed land cover datasets to update ISA and harmonise it with datasets adopted for the UCPs. We identify and compare some relevant datasets at the pan-European level. The results show the benefits obtained by correlating the UCPs from urban Local Climate Zones taken from the ECOCLIMAP-Second Generation with the ISA depending on the imperviousness degree provided by the Copernicus Land Monitoring Service. These results will foster an updating of NWP and climate models with urban surface parameterisations to match urban features while avoiding inconsistencies consistently.

Survey of land cover datasets for updating the imperviousness field in urban parameterisation scheme TERRA_URB for climate and weather applications

In climate modelling, a critical issue is the accurate representation of urban areas in space and variation over time. Urban features may be parameterised through urban canopy parameters (UCPs), including the impervious surface area (ISA). One of the aims of the Consortium for Small-scale Modelling (COSMO) and the CLM-Community is to improve urban numerical weather prediction (NWP) and climate simulations by implementing the urban canopy scheme TERRA_URB in the ICOsahedral Nonhydrostatic (ICON) modelling framework. To harmonise and update the datasets adopted for the UCPs, the Consortium has launched the Priority Project “City Induced Temperature change Through A'dvanced modelling” (PP CITTÀ). Within this framework, the study explores recent and detailed land cover datasets to update ISA and harmonise it with datasets adopted for the UCPs. We identify and compare some relevant datasets at the pan-European level. The results show the benefits obtained by correlating the UCPs from urban Local Climate Zones taken from the ECOCLIMAP-Second Generation with the ISA depending on the imperviousness degree provided by the Copernicus Land Monitoring Service. These results will foster an updating of NWP and climate models with urban surface parameterisations to match urban features while avoiding inconsistencies consistently.

How well does a convection-permitting regional climate model represent the reverse orographic effect of extreme hourly precipitation?

Abstract. Estimating future short-duration extreme precipitation in mountainous regions is fundamental for risk management. High-resolution convection-permitting models (CPMs) represent the state of the art for these projections, as they resolve convective processes that are key to short-duration extremes. Recent observational studies reported a decrease in the intensity of extreme hourly precipitation with elevation. This “reverse orographic effect” could be related to processes which are subgrid even for CPMs. To quantify the reliability of future projections of extreme short-duration precipitation in mountainous regions, it is thus crucial to understand to what extent CPMs can reproduce this effect. Due to the computational demands however, CPM simulations are still too short for analyzing extremes using conventional methods. We use a non-asymptotic statistical approach (Simplified Metastatistical Extreme Value: SMEV) for the analysis of extremes from short time periods, such as the ones of CPM simulations. We analyze an ERA-Interim-driven Consortium for Small-Scale Modeling (COSMO-crCLIM, convection-resolving Climate Modelling) simulation (2000–2009; 2.2 km resolution), and we use hourly precipitation from 174 rain gauges in an orographically complex area in northeastern Italy as a benchmark. We investigate the ability of the model to simulate the orographic effect on short-duration precipitation extremes, as compared to observational data. We focus on extremes as high as the 20-year return levels. While overall good agreement is reported at daily and hourly duration, the CPM tends to increasingly overestimate hourly extremes with increasing elevation, implying that the reverse orographic effect is not fully captured. These findings suggest that CPM bias-correction approaches should account for orography. SMEV's capability of estimating reliable rare extremes from short periods promises further applications on short-time-period CPM projections and model ensembles.

Heavy snowfall event over the Swiss Alps: did wind shear impact secondary ice production?

Abstract. The change in wind direction and speed with height, referred to as vertical wind shear, causes enhanced turbulence in the atmosphere. As a result, there are enhanced interactions between ice particles that break up during collisions in clouds which could cause heavy snowfall. For example, intense dual-polarization Doppler signatures in conjunction with strong vertical wind shear were observed by an X-band weather radar during a wintertime high-intensity precipitation event over the Swiss Alps. An enhancement of differential phase shift (Kdp>1∘ km−1) around −15 ∘C suggested that a large population of oblate ice particles was present in the atmosphere. Here, we show that ice–graupel collisions are a likely origin of this population, probably enhanced by turbulence. We perform sensitivity simulations that include ice–graupel collisions of a cold frontal passage to investigate whether these simulations can capture the event better and whether the vertical wind shear had an impact on the secondary ice production (SIP) rate. The simulations are conducted with the Consortium for Small-scale Modeling (COSMO), at a 1 km horizontal grid spacing in the Davos region in Switzerland. The rime-splintering simulations could not reproduce the high ice crystal number concentrations, produced too large ice particles and therefore overestimated the radar reflectivity. The collisional-breakup simulations reproduced both the measured horizontal reflectivity and the ground-based observations of hydrometeor number concentration more accurately (∼20 L−1). During 14:30–15:45 UTC the vertical wind shear strengthened by 60 % within the region favorable for SIP. Calculation of the mutual information between the SIP rate and vertical wind shear and updraft velocity suggests that the SIP rate is best predicted by the vertical wind shear rather than the updraft velocity. The ice–graupel simulations were insensitive to the parameters in the model that control the size threshold for the conversion from ice to graupel and snow to graupel.

Identification of high-wind features within extratropical cyclones using a probabilistic random forest – Part 1: Method and case studies

Abstract. Strong winds associated with extratropical cyclones are one of the most dangerous natural hazards in Europe. These high winds are mostly associated with five mesoscale dynamical features: the warm (conveyor belt) jet (WJ); the cold (conveyor belt) jet (CJ); cold frontal convection (CFC); strong cold-sector winds (CS); and, at least in some storms, the sting jet (SJ). The timing within the cyclone's life cycle, the location relative to the cyclone core and some further characteristics differ between these features and, hence, likely also the associated forecast errors. Here, we present a novel objective identification approach for these high-wind features using a probabilistic random forest (RF) based on each feature’s most important characteristics in near-surface wind, rainfall, pressure and temperature evolution. As the CJ and SJ are difficult to distinguish in near-surface observations alone, these two features are considered together here. A strength of the identification method is that it works flexibly and is independent of local characteristics and horizontal gradients; thus, it can be applied to irregularly spaced surface observations and to gridded analyses and forecasts of different resolution in a consistent way. As a reference for the RF, we subjectively identify the four storm features (WJ, CS, CFC, and CJ and SJ) in 12 winter storm cases between 2015 and 2020 in both hourly surface observations and high-resolution reanalyses of the German Consortium for Small-scale Modeling (COSMO) model over Europe, using an interactive data analysis and visualisation tool. The RF is then trained on station observations only. The RF learns physically consistent relations and reveals the mean sea level pressure (tendency), potential temperature, precipitation amount and wind direction to be most important for the distinction between the features. From the RF, we get probabilities of each feature occurring at the single stations, which can be interpolated into areal information using Kriging. The results show a reliable identification for all features, especially for the WJ and CFC. We find difficulties in the distinction of the CJ and CS in extreme cases, as the features have rather similar meteorological characteristics. Mostly consistent results in observations and reanalysis data suggest that the novel approach can be applied to other data sets without the need for adaptation. Our new software RAMEFI (RAndom-forest-based MEsoscale wind Feature Identification) is made publicly available for straightforward use by the atmospheric community and enables a wide range of applications, such as working towards a climatology of these features for multi-decadal time periods (see Part 2 of this paper; Eisenstein et al., 2022d), analysing forecast errors in high-resolution COSMO ensemble forecasts and developing feature-dependent post-processing procedures.

Do the large-eddy simulations yield deeper atmospheric boundary layers in comparison to the RANS model simulations?

In this article, we address the behaviour of a RANS and an LES model in the simulation of the diurnally evolving atmospheric boundary layer (ABL) over Thiruvananthapuram, a coastal station in India, for contrasting sky conditions during the passage of a tropical cyclone Ockhi. Based on the prevailing meteorological conditions over this experimental site, 5 December 2017 and 1 December 2017 are chosen as representative days for the clear-sky and cloudy conditions, respectively. The Consortium for small-scale modelling (COSMO), a regional numerical weather prediction model is chosen to represent the RANS technique, and a mesoscale-to-LES nested configuration of the Parallelized LES Model (PALM) is taken to represent the LES technique. The LES experiments are performed at 30 m and 100 m grid spacings. The main objective of these experiments is to assess the individual model biases of nested LES in conjunction with a RANS model with a special emphasis on the diurnal evolution of ABL heights. The ABL heights inferred from the LES experiments were found to be deeper than that in the RANS experiments. The decay of ABL during the evening hours was rapid in the LES experiments, whereas it was more smooth in the RANS experiment. The study also evaluates the sensitivity of the PALM to the grid spacing in a mesoscale-to-LES nesting configuration.

Columnar and surface urban aerosol in the Moscow megacity according to measurements and simulations with the COSMO-ART model

Abstract. Urban aerosol pollution was analyzed over the Moscow megacity region using the COSMO-ART (COSMO – COnsortium for Small-scale MOdelling, ART – Aerosols and Reactive Trace gases) online coupled mesoscale model system and intensive measurement campaigns at the Moscow State University Meteorological Observatory (MSU MO, 55.707∘ N, 37.522∘ E) during the April–May period in 2018 and 2019. We analyzed mass concentrations of particulate matter with diameters smaller than 10 µm (PM10), black carbon (BC) and aerosol gas precursors (NOx, SO2, CHx) as well as columnar aerosol parameters for fine and coarse modes together with different meteorological parameters, including an index characterizing the intensity of particle dispersion (IPD). Both model and experimental datasets have shown a statistically significant linear correlation of BC with NO2 and PM10 mass concentrations, which indicates mostly common sources of emissions of these substances. There was a pronounced increase in the BC/PM10 ratio from 0.7 % to 5.9 %, with the decrease in the IPD index related to the amplification of the atmospheric stratification. We also found an inverse dependence between the BC/PM10 ratio and columnar single-scattering albedo (SSA) for the intense air mixing conditions. This dependence together with the obtained negative correlation between wind speed and BC/PM10 may serve as an indicator of changes in the absorbing properties of the atmosphere due to meteorological factors. On average, the relatively low BC / PM10 ratio (for urban regions) of 4.7 % is the cause of the observed relatively high SSA = 0.94 in Moscow. Using long-term parallel aerosol optical depth (AOD) measurements over the 2006–2020 period at the MSU MO and under upwind clean background conditions at Zvenigorod Scientific Station (ZSS) of the IAP RAS (55.7∘ N, 36.8∘ E), we estimated the urban component of AOD (AODurb) and some other parameters as the differences at these sites. The annual mean AODurb at 550 nm was about 0.021 with more than 85 % of the fine aerosol mode. The comparisons between AODurb obtained from the model and measurements during this experiment have revealed a similar level of aerosol pollution of about AODurb=0.015–0.019, which comprised 15 %–19 % of the total AOD at 550 nm. The urban component of PM10 (PM10urb) was about 16 µg m−3 according to the measurements and 6 µg m−3 according to the COSMO-ART simulations. We obtained a pronounced diurnal cycle of PM10urb and urban BC (BCurb) as well as their strong correlation with the IPDs. With the IPD index change from 3 to 1 at night, there was about a 4 times increase in PM10urb (up to 30–40 µg m−3) and a 3 times increase in BCurb (up to 3–3.5 µg m−3). At the same time, no pronounced daily cycle was found for the columnar urban aerosol component (AODurb), although there was a slight increase in model AODurb at night.

Performance evaluation of a high‐resolution regional model over West Africa for operational use: A case study of August 2017

AbstractThe frequency of flash floods resulting from heavy rainfall over West Africa has increased in recent years with serious socio‐economic consequences. Therefore, the need to utilize numerical weather prediction models to forecast heavy rainfall events reliably is also rising at many operational meteorological centres in West Africa. This paper evaluates the performance of the Consortium for Small‐scale Modelling (COSMO) model of the German Meteorological Services (DWD) in predicting rainfall over West Africa for high‐impact rainfall events that occurred between 19 and 26 August 2017. The paper aims to investigate the synoptic forcings modulating daily rainfall variability during that period. Results show that COSMO simulates adequately the spatio‐temporal variability of rainfall distribution over West Africa, though with inherent biases. COSMO displays a decreasing skill in producing spatial rainfall distribution as rainfall amounts tend to 30 mm and above. Additionally, areas of heavy rainfall, mostly about 100–300 km southwest of the core of the Africa Easterly Jet (AEJ), often coincide with areas of decreasing mean sea level pressure of at least 0.6 hPa and areas of increasing convective available potential energy of at least 500 J/kg. Although not in all cases, the trough of the Africa Easterly Wave (AEW) is always located to the east of these areas. We show that not every storm, especially east of the prime meridian, is associated with an AEW trough. COSMO is able to reproduce the atmospheric dynamics modulating the daily rainfall variability, in addition to capturing the daily propagation of the AEW trough, and the core of the AEJ. However, the reproducibility skill of the model in predicting atmospheric dynamics may not transform into the predictive skill of the model in producing rainfall. Nevertheless, operational forecasters may be able to determine likely areas of heavy rainfall by estimating the position of the AEJ core based on the position of areas of the least falling pressure from COSMO. Finally, the incorporation of the fractions skill score metric based on the neighbourhood approach could also assist operational forecasters to decide at which scale a severe weather alert can be issued.

Combined effects of soil moisture and microphysical perturbations on convective clouds and precipitation for a locally forced case over Central Europe

AbstractTo study the combined impact of soil moisture and microphysical perturbations on convective clouds and precipitation over Central Europe, an ensemble of five dozen real‐world weather prediction forecasts was conducted with the COnsortium for Small‐scale MOdeling (COSMO) model at convection‐permitting resolution for a case with weak large‐scale forcing (6 June 2016). We find a large sensitivity of precipitation, ranging from +10% to 23% in 12‐hr precipitation totals. While the homogeneous soil‐moisture bias of 25% primarily controls the timing of convection initiation and the amount of surface rainfall, the number of cloud condensation nuclei and width of the cloud droplet size distribution mainly control the number, size, and lifetime of convective clouds. In moisture‐limited conditions, mainly positive couplings are acting. Drier soils, cleaner air, and a broader cloud droplet size distribution result in less rainfall. Wetter soils and more polluted conditions lead to fewer, but larger, cloud clusters. Since microphysical process rates depend systematically on the sign of the perturbations, but rainfall does not, there are compensating effects at work that buffer microphysical perturbations directly and impact the cloud condensate amount and the rainfall at the ground.