Displays

The phenomenon of convective aggregation in idealised radiative convective equilibrium simulations has the ability to change the mean state of its domain. When compared to non-aggregation conditions, these simulations usually have warmer drier mean atmospheres, with stronger precipitation in the convective areas. Many of these idealised experiments use a fixed sea surface temperature (SST), where higher temperatures generally increase the scale of aggregation. SST gradients have been shown to organise convection, yet there has been no work done to investigate the impact of heating perturbations in the air on the aggregation of convection. Here we investigate how strong diabatic heating of the atmospheric column affect the existence and properties of convective aggregation. These perturbations provide a link to studying the effect of large pollution plumes on convection, for example during the Indian monsoon season.

An aerosol model is used to insert plumes of strongly absorbing aerosols into aquaplanet, non-rotating, global RCE simulations. We study the sensitivity of the response to aerosol optical depth (AOD) and aerosol radiative properties under different SSTs.

Without any forcing, the simulations at low SST do not aggregate while at high SST they do. We also see that adding the forcing causes aggregation at both temperatures for a wide range of AODs. Detailed investigation shows that the diabatic heating source causes two circulations to develop, one with low-level convergence towards the plume and high-level divergence away from the plume. A secondary circulation works tangentially to the plume, again with low-level convergence and high-level divergence, driving the formation of several radial branches of aggregated convection. We argue that, as we see this aggregation for plumes with realistic AODs, this could be an analogue for real-world organisation during high pollution events. Future work will investigate the difference in mechanisms between forced and unforced convective aggregation as well as conducting similar experiments in smaller, cloud resolving domains.

How to cite: Dingley, B., Dagan, G., and Stier, P.: Forced Convective Aggregation, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-16505, https://doi.org/10.5194/egusphere-egu2020-16505, 2020.

The spontaneous aggregation of convective clouds over a moist portion of the domain is ubiquitous in cloud resolving model simulations. This phenomenon significantly reduces the domain mean total water vapor and enhances the outgoing long radiation. In this study we use the system of atmospheric modeling (SAM) in a radiative-convective equilibrium (RCE) setup in order to investigate the impact of an interactive sea surface temperature (SST) on the aggregation progress. We use a slab ocean (with depth of 5, 10 and 50 m) with constant target SST to which the domain mean SST is relaxed. Our results show that, consistent with previous studies, an interactive SST delays the aggregation with a larger impact for a shallower slab. This effect is enhanced for a smaller target SST.

The aggregation proceeds by the expansion of non-convective dry areas. Before aggregation, dry areas are associated with warmer surface due to enhanced short-wave radiation. During and after the aggregation, a single large dry patch develops and is associated with a colder surface. This cooling is due to a reduction in downwelling long-wave radiation and to enhanced latent heat flux due to drier boundary layer. The edge of the dry patch has warm SST anomaly forming a ring of warm water around it that favors divergence of low-level moist air from the dry patch and accelerates dry patch expansion. This is favored by a positive surface pressure anomaly (PSFC) in the dry patch.

Therefore, at first, the warm SST anomaly opposes the divergent flow from dry regions, opposing the aggregation. Then the cold SST anomaly that develops in dry regions increases the divergent flow and favors the dry patch expansion. For a small ocean slab, the warm SST anomaly that develops in the dry areas at early times inhibits the dry patch expansion and can significantly delay the beginning of aggregation.

How to cite: Shamekh, S., Muller, C., Duvel, J.-P., and D'Andrea, F.: Impact of an Interactive SST on the Convective Aggregation Process, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-14941, https://doi.org/10.5194/egusphere-egu2020-14941, 2020.

Convective self-aggregation is a modelling paradigm for thunderstorm organisation over a constant-temperature tropical sea surface. This setup can give rise to cloud clusters over timescales of weeks. In reality, sea surface temperatures do oscillate diurnally, affecting the atmospheric state. Over land, surface temperatures vary more strongly, and rain rate is significantly influenced. Here, we carry out a substantial suite of cloud-resolving numerical experiments, and find that even weak surface temperature oscillations enable qualitatively different dynamics to emerge: the spatial distribution of rainfall is only homogeneous during the first day. Already on the second day, the rain field is firmly structured. In later days, the clustering becomes stronger and alternates from day to day. We show that these features are robust to changes in resolution, domain size, and surface temperature, but can be removed by a reduction of the amplitude of oscillation, suggesting a transition to a clustered state. Maximal clustering occurs at a scale of l_max≈180 km, a scale we relate to the emergence of mesoscale convective systems. At l_max rainfall is strongly enhanced and far exceeds the rainfall expected at random. We explain the transition to clustering using simple conceptual modelling. Our results may help clarify how continental extremes build up and how cloud clustering over the tropical ocean could emerge much faster than through conventional self-aggregation alone.

How to cite: Haerter, J. O., Meyer, B., and Nissen, S. B.: Diurnal Self-Aggregation, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-11878, https://doi.org/10.5194/egusphere-egu2020-11878, 2020.

Collisions represent one of the most important processes through which cold pools—essential boundary layer features of precipitating systems—help to organize convection. For example, by colliding with one another, expanding cold pools can trigger new convective cells, a process that has been argued to be important to explain the deepening of convection and the maintenance of mesoscale convective systems for many hours. In spite of their role, collisions are an understudied process, and many aspects remain to be fully clarified. In order to quantify the importance of collisions on the life cycle of cold pools, we will present some results based on a combination of numerical simulations in radiative-convective equilibrium and a Lagrangian cold pool tracking algorithm. First, we will discuss how the Lagrangian algorithm can be used to estimate that the median time of the first collision for the simulated cold pools is under 10 minutes. We will then show that cold pools are significantly deformed by collisions and lose their circular shape already at the very early stages of their life cycle. Finally, we will present results suggesting that cold pools appear to be clustered, and we will provide some estimates of the associated temporal and spatial scales.

How to cite: Torri, G. and Kuang, Z.: A Lagrangian perspective on cold pool collisions, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-6219, https://doi.org/10.5194/egusphere-egu2020-6219, 2020.

Cold pools are essential for organizing convection and play a particular role in convective initiation in the afternoon and evening. Both aspects are deficient in current convection-permitting models and a better representation of cold pools is likely necessary to overcome these deficiencies. In a recent investigation, we identified several sensitivities of cold pool driven convective initiation to model resolution within hectometer simulations. In particular, a causal graph analysis has revealed that the dominant impact of model resolution on convective initiation is due to too weak gust front vertical velocities. This implies that cold pool gust fronts in km-scale models are too weak to trigger sufficient new convection.

To address this deficiency, we develop a parameterization for the convection-permitting COSMO model to improve the representation of cold pool gust fronts. We use the potential temperature gradient to identify cold pool gust fronts and enhance vertical wind tendencies within these gust front regions.  Also, we perturb horizontal wind tendencies to yield 3d non-divergent perturbations.  This parameterization strengthens gust front circulations and thereby enhances cold pool driven convective initiation. Consequently, precipitation is amplified and becomes more organized in the afternoon and evening. This improves the diurnal cycle of precipitation and also has some positive impact on the spatial distribution as quantified by the fraction skill score. Furthermore, cold pools themselves are strengthened, which can further enhance the gust front circulations, giving rise to a feedback loop. 

How to cite: Hirt, M. and Craig, G.: Cold pool driven convective initiation: How can we improve its representation in km-scale models?, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-5260, https://doi.org/10.5194/egusphere-egu2020-5260, 2020.

Cold pools are areas of cool downdraft air that form through evaporation underneath precipitating clouds and spread on the surface as density currents. Their importance for the development and maintenance of convection is long known. Modern Large-Eddy simulations with a grid spacing of 1 km or less are able to explicitly resolve cold pools, however, they lack reference data for an adequate validation. Available point measurements from operational networks are too coarse and, therefore, miss the horizontal structure and dynamics of cold pools.

The upcoming measurement campaign FESSTVaL (Field Experiment on Sub-mesocale Spatio-Temporal Variability in Lindenberg) aims to test novel measurement strategies for the observation of previously unresolved sub-mesoscale boundary layer structures and phenomena, such as cold pools. The key component of the experiment during this summer will be a dense network of ground-based measurements within 15 km around the Meteorological Observatory Lindenberg near Berlin. The network of 100 low-cost APOLLO (Autonomous cold POoL LOgger) stations allows to record air pressure and temperature with a spatial and temporal resolution of 100 m and 1 s, respectively. We present first results from a test campaign during last summer that successfully demonstrated the ability of the proposed network stations to observe cold pool dynamics on the sub-mesoscale.

How to cite: Kirsch, B., Ament, F., Hohenegger, C., and Klocke, D.: Sub-mesoscale observations of cold pools during FESSTVaL, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-9779, https://doi.org/10.5194/egusphere-egu2020-9779, 2020.

Idealized simulations have shown decades ago that shallow clouds generate internal gravity waves, which under certain atmospheric background conditions become trapped inside the troposphere and influence the development of clouds. These feedbacks, which occur at horizontal scales of up to several tens of km are neither resolved, nor parameterized in traditional global climate models (GCMs), while the newest generation of GCMs is starting to resolve them. The interactions between the convective boundary layer and trapped waves have almost exclusively been studied in highly idealized frameworks and it remains unclear to what degree this coupling affects the organization of clouds and convection in the real atmosphere. Here, the coupling between clouds and trapped waves is examined in storm-resolving simulations that span the entirety of the tropical Atlantic and are initialized and forced by meteorological analyses. The coupling between clouds and trapped waves is sufficiently strong to be detected in these simulations of full complexity.  Stronger upper-tropospheric westerly winds are associated with a stronger cloud-wave coupling. In the simulations this results in a highly-organized scattered cloud field with cloud spacings of about 19 km, matching the dominant trapped wavelength. Based on the large-scale atmospheric state wave theory can reliably predict the regions and times where cloud-wave feedbacks become relevant to convective organization. Theory, the simulations and satellite imagery imply a seasonal cycle in the trapping of gravity waves. 

How to cite: Stephan, C.: Seasonal modulation of trapped gravity waves and their imprints on trade wind clouds, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-2612, https://doi.org/10.5194/egusphere-egu2020-2612, 2020.

The thermal convection structures (TCS) and their characteristics manifestations in the atmospheric boundary layer were investigated using the data from acoustic Doppler sodar LATAN-3M. A longwave LATAN-3M sodar with a vertical resolution of 20 m in 2007 and 10 m in 2016, 2018, 2019, a pulse emission interval of 5 s in 2007 and 3 s in 2016, 2018, 2019, an altitude range of 400–600 m in 2007 and 350 m in 2016, 2018, 2019, and a basic carrier frequency of 2 kHz in 2007 and 3 kHz in 2016, 2018, 2019 had measured the profiles of the wind velocity components which were used for calculating the scale of TCS. Experimental data were being obtained during the field campaigns organized by the A.M. Obukhov Institute of Atmospheric Physics RAS in Rostov region and over semi-arid zones of the Caspian lowland in the eastern part of Kalmykia Republic, Russia.

The wind was weak and the convection was well-developed in the case studies over July of years 2007, 2016, 2018, 2019. A moving rectangular filter was used for averaging the original data of the horizontal and vertical wind-velocity components. The averaging interval had been empirically chosen and, in this case, amounted to 10 min. At such values, the spatiotemporal velocity-field structure was adequately reproduced.

The original method of acoustic sounding data treatment for extracting TCS has been developed and put to an evaluation test. The episodes of the vertical velocities above limit values at which TCS aroused hypothetically were considered. As the threshold, a few alternatives were used: 0.3 m/s, 0.6 m/s and 1.2 m/s. The duration of vertical velocity excess over the threshold, the maximum velocity within this interval and the horizontal scale were calculated. It is assumed that TCS move forward with some averaged velocity during any relatively small time step. In this case, the spatial distribution of velocity field and its time variations have been reproduced suitably.

The statistical distribution was close to Rayleigh distribution:

p(U) = (2U/U₀²)*exp ((U_m²-U ²)/U₀²),

where U₀² = (<U ²>-U_m²), <U ²> is the root-mean-square vertical velocity of TCS, and U_m – the threshold for vertical velocity. This closeness can facilitate the understanding of the processes in the so-called “grey-zone” of numerical simulation and be implemented in the parameterization, forecast of TCS. Note that Rayleigh distribution is applied to the statistics of the intense moist convective vortices and also of the height of the ocean waves.

This work was supported by Russian Foundation for Basic Research (projects No.19-05-50110, No.19-05-01008, No.17-05-41121), and by fundamental research program of Russian Academy of Science (program No.1).

How to cite: Vazaeva, N., Chkhetiani, O., Kurgansky, M., Kallistratova, M., Lyulyukin, V., and Zaytseva, D.: Statistical Characteristics of Thermal Convection Structures based on Acoustic Sounding Data, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-1061, https://doi.org/10.5194/egusphere-egu2020-1061, 2020.

Thermal convection is of major importance in various astro- and geophysical systems, exemplary are buoyancy driven flows in the atmosphere or in the stellar interior. It has been studied for decades in an idealized model system - the Rayleigh-Bénard convection (RBC) - which consists of a horizontal fluid layer heated at the bottom and cooled at the top. Within the Oberbeck-Boussinesq approximation this system is controlled by two parameters only. These are the Rayleigh number (Ra), which represents the thermal driving and the Prandtl number (Pr) that relates the momentum and thermal diffusivities of the fluid. Convection flows in geo- and astrophysics are often influenced by Coriolis forces due to the rotation of the planet or the star. In RBC-system, Coriolis forces are introduced by rotating the convection cell around its vertical axis. The rotation is expressed by an additional dimensionless control parameter, i.e., the inverse Rossby number 1/Ro. We study experimentally the influence of rotation on the heat transport and the temperature field at very large Ra in the High Pressure Convection Facility (HPCF) in Göttingen. The facility consists of a cylindrical cell of 1.10m diameter and 2.20m height that is filled with pressurized sulfur hexafluoride (SF₆) at up to 19bar. The height of the cell and the large density of SF₆ enable us to reach very large Ra (up to 8×10¹⁴) at 0.74<Pr<0.96. The cell is mounted on a rotating table and connected to the non-rotating world via water feed-throughs and slip rings. With these, the signals of more than 100 thermistors close to the sidewalls are collected.
We find a monotonic decrease of the heat transport with increasing rotation rate. Furthermore, we measure quantities of the flow close to the lateral side walls of the convection cylinder. For large rotation rates we analyze this as part of the recently proposed “Boundary Zonal Flow” (BZF), where the vertical heat transport is enhanced and warm (cold) up (down) flow self-organizes in a periodic manner. In the experiment we observe the BZF most notably in the probability density function of the temperature, which develops a bimodal Gaussian distribution. We also find that the periodic warm-cold structure drifts in anti-cyclonic direction and thus form traveling waves of the temperature field.

How to cite: Wedi, M., van Gils, D., Ahlers, G., Bodenschatz, E., and Weiss, S.: Experimental investigation on turbulent rotating thermal convection at large Rayleigh numbers, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-9511, https://doi.org/10.5194/egusphere-egu2020-9511, 2020.

How to cite: Garrett, T., Krueger, S., Glenn, I., and Ferlay, N.: Thermodynamic constraints on the size distributions of tropical convective clouds. , EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-6465, https://doi.org/10.5194/egusphere-egu2020-6465, 2020.

The disparity in the increase in atmospheric water vapour and in the global energy budget with global warming is expected to lead to a greater contribution of precipitation from deep convective clouds (DCCs) to total precipitation. How this increase occurs is uncertain however; while many climate models predict that the intensity of precipitation in individual storms will increase while occurring at the same frequency, satellite observations of tropical cloud clusters have shown that the frequency of organised deep convective precipitation events is increasing. By studying the interactions between deep convective precipitation and the energy and water budgets, we aim to achieve a better understanding of how these budgets affect the intensity, frequency and organisation of deep convective clouds and the cloud feedbacks on subsequent convection.

The new generation of geostationary imaging satellites provides greatly improved observations of dynamic processes. Using optical-flow techniques, we show how a semi-Lagrangian perspective can be applied to GOES advanced baseline imager observations in the thermal IR spectrum, and how this perspective can improve our observations of the dynamics of DCCs. In this new perspective we are able to robustly track DCCs over their entire lifecycle. As a result, the interactions between energy budgets, organisation and growing convection can be linked to subsequent precipitation and radiative feedbacks over the entire lifetime of the DCC.

In a case study over the continental US, we observe a suppression of convective strength in the days following large, organised convective storms. Compared to similar DCCs prior to the large organised events, the subsequent DCCs develop more slowly and, despite having a similar maximum anvil cloud extent, have a shorter overall lifetime. Furthermore, the later anvil clouds and convective cores have warmer cloud top brightness temperatures by 10 K and up to 20 K respectively. We hope to gain a greater understanding of whether these changes are due to large scale dynamics associated with the large, organised convection or can instead be attributed to local cloud and thermodynamic feedbacks.

How to cite: Jones, W., Heikenfeld, M., Christensen, M., and Stier, P.: A semi-Lagrangian perspective of the lifecycle and interactions of deep convective clouds in geostationary satellite observations, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-21644, https://doi.org/10.5194/egusphere-egu2020-21644, 2020.

Deep convection plays an important role in driving the large-scale circulation and the complex interaction between moist convection and the large-scale circulation regulates the thermodynamic structure of the tropical atmosphere. 

The convectional thoughts of the thermodynamic structure of the tropical atmosphere are that the horizontal temperature in the free troposphere is homogeneous, which is referred to as weak temperature gradient (WTG), while the vertical structure follows a moist-adiabatic lapse rate. However, it is not known how accurate WTG holds and which moist- adiabatic process the tropical atmosphere indeed experiences. This study centers around the horizontal and vertical structure of the tropical atmosphere and uses the global storm resolving simulations (GSRMs) from ICON at 2.5 km to investigate them

The virtual effect or the vapor buoyancy effect arises from that the molecular weight of water vapor is much smaller than that of dry air. With the same pressure and temperature, this virtual effect makes moist air lighter than dry air. As the horizontal buoyancy differences are eliminated by convection gravity waves, virtual temperature, a temperature variable including the moisture conditions, is expected to be homogeneous. Then, to obtain a homogeneous virtual temperature horizontally, the absolute temperature has to change to accommodate the horizontal moisture difference. The model results show that virtual temperature is relatively homogeneous at mid- and lower troposphere. Therefore, the virtual effect plays a very important role in the horizontal temperature structure, making the absolute temperature colder in moist regions and warmer in dry regions. However, in the upper troposphere, both the absolute temperature and the virtual temperature are not homogeneous, and vary as a function of moisture, indicating a weakening influence of convection gravity waves there.

We use saturation equivalent potential temperature (theta-es) to explore the vertical structure of the tropical atmosphere. Theta-es is expected to be conserved above the lifting condensation level (LCL) if calculated following the exact moist-adiabatic process that tropical atmosphere undergoes. The pseudo-adiabat and the reversible-adiabat with the effect of condensate loading are compared. To minimize the horizontal differences in theta-es due to moisture, we also define theta-es to account for the virtual effect and the condensate loading effect. The model results suggest that the actual moist-adiabatic process that tropical atmosphere experiences is between the pseudo-adiabat and the reversible-adiabat with the effect of condensate loading assuming air parcels originating from 972 hPa. 

The above results are broadly consistent with the results from ERA5 reanalysis.

How to cite: Bao, J. and Stevens, B.: The elements of the thermodynamic structure of the tropical atmosphere, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-21657, https://doi.org/10.5194/egusphere-egu2020-21657, 2020.

To study the influence of convective momentum transport (CMT) on wind, boundary layer and cloud evolution in a marine cold air outbreak (CAO) we use Large-Eddy Simulations subjected to different baroclinicity (wind shear) but similar surface forcing. The simulated domain is large enough ( ≈100 × 100 km²) to develop typical mesoscale cellular convective structures.  We find that a maximum friction induced by momentum transport (MT) locates in the cloud layer for an increase of geostrophic wind with height (forward shear, FW) and near the surface for a decrease of wind with height (backward shear, BW). Although the total MT always acts as a friction, the interaction of friction-induced cross-isobaric flow with the Coriolis force can develop super-geostrophic winds near the surface (FW) or in the cloud layer (BW). The contribution of convection to MT is evaluated by decomposing the momentum flux by column water vapor and eddy size, revealing that CMT acts to accelerate sub-cloud layer winds under FW shear and that mesoscale circulations contribute significantly to MT for this horizontal resolution (250 m), even if small scale eddies are non-negligible and likely more important as resolution increases. Under FW shear, a deeper boundary layer and faster cloud transition are simulated, because MT acts to increase surface fluxes and wind shear enhances turbulent mixing across cloud tops. Our results show that the coupling between winds and convection is crucial for a range of problems, from CAO lifetime and cloud transitions to ocean heat loss and near-surface wind variability.

How to cite: Saggiorato, B., Nuijens, L., Siebesma, A. P., de Roode, S., Sandu, I., and Papritz, L.: The influence of convective momentum transport and vertical wind shear on the evolution of a cold air outbreak, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-7076, https://doi.org/10.5194/egusphere-egu2020-7076, 2020.

How to cite: Zeman, C., Ban, N., Wedi, N., and Schär, C.: Validity of the hydrostatic approximation at convection-resolving scales: Diagnostic analysis and model intercomparison, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-9736, https://doi.org/10.5194/egusphere-egu2020-9736, 2020.

The importance of convective organization on the global circulation has been recognized for a long time, but parameterizations of the associated processes are missing in global climate models. Contemporary convective parameterizations commonly use a convective plume model (or a spectrum of plumes). This is perhaps appropriate for unorganized convection but the assumption of a gap between the small cumulus scale and the large-scale motion fails to recognize mesoscale dynamics manifested in mesoscale convective systems (MCSs) and multi-scale cloud systems associated with the MJO. Organized convection is abundant in environments featuring vertical wind shear, and significantly modulates the life cycle of moist convection, the transport of heat and momentum, and accounts for a large percentage of precipitation in the tropics. Mesoscale convective organization is typically associated with counter-gradient momentum transport, and distinct heating profiles between the convective and stratiform regions.

Moncrieff, Liu and Bogenschutz (2017) recently developed a dynamical based parameterization of organized moisture convection, referred to as multiscale coherent structure parameterization (MCSP), for global climate models. A prototype version of MCSP has been implemented in the NCAR Community Earth System Model (CESM) and the Energy Exascale Earth System Model (E3SM), positively affecting the distribution of tropical precipitation, convectively coupled tropical waves, and the Madden-Julian oscillation. We will show the further development of the MCSP and its impact on the simulation of mean precipitation and variability in the two global climate models.

How to cite: Chen, C.-C., Liu, C., Moncrieff, M., and Richter, Y.: Evaluation of Organized Convection Parameterization in Global Climate Models, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-3758, https://doi.org/10.5194/egusphere-egu2020-3758, 2020.

Convective self-aggregation in radiative convective equilibrium has been studied due to its similarities to organized convection in the tropics. As tropical atmospheric phenomena are embedded in a large-scale flow, we impose a background wind to the model setup using convection-permitting simulation to analyze the interaction of convective self-aggregation with the background wind. The simulations show that when imposing a background wind, the convective cluster propagates in the direction of the imposed wind but slows down compared to what pure advection would suggest, and eventually becomes stationary. The dynamic process dominates slowing down the propagation speed of the cluster because the surface momentum flux acts as a drag on the near-surface wind, terminating the propagation. The thermodynamic process through the wind-induce surface feedback contributes to only 6% of the propagation speed of the convective cluster and is strongly modified by the dynamic process.

How to cite: Jung, H., Naumann, A. K., and Stevens, B.: Role of dynamic and thermodynamic processes for the propagation of organized convection in a large-scale flow, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-5075, https://doi.org/10.5194/egusphere-egu2020-5075, 2020.

In this study we are trying to understand (limits of) predictability related to (organised) convection and its upscale error growth.

For that purpose we aim to analyse the impact of three convection driving and amplifying processes, namely latent heat release, redistribution of moist static energy and convective momentum transport on the development of the convective cells. Furthermore, we plan to investigate uncertainties in these processes on downward propagation of the flow and ensemble spread.

The first results to be presented regard an idealised and strongly organised case of splitting convective storms modeled at different resolutions and with some small adaptations in the model convective cloud resolving model CM1. Currently processed resolution experiments show that both the actual divergence field and the processes supected to underlie it exhibit some sensitivity to model resolution on the subkilometre scale (100-1000 m). We can also show that the upper tropospheric divergence can be directly related to the latent heat release, as it is located vertically above the major latent heat releases. Nevertheless, neither the vertical redistribution of moist static energy nor the convective momentum transport are negligible and all three impact the divergent outflow of the convective storm.

How to cite: Groot, E. and Tost, H.: Sensitivity analysis of divergence and underlying processes around organised convection with a cloud resolving model, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-5630, https://doi.org/10.5194/egusphere-egu2020-5630, 2020.

Shallow convection over an unfrozen lake (Lake Ontario) during a cold-air outbreak is simulated using the Weather Research and Forecasting model (WRF) with two horizontal grid spacings, 148 m and 1.33 km. The dynamics and microphysics of the simulated convective snow band are compared to radar and aircraft observations. The dynamical and microphysical changes that occur when going from 1.33-km to 148-m grid spacing are explored. Improved representation of the convective dynamics at higher resolution leads to a better representation of the microphysics of the snowband compared to radar and aircraft observations. Stronger updrafts in the high-resolution grid lead to larger ice nucleation rates and produce ice particles that are more heavily rimed and thus faster falling. These changes to the ice particle properties in the high resolution grid limit aggregation rates and result in more realistic radar reflectivity patterns. Graupel, observed at the surface, is produced in the strongest convective updrafts, but only at the higher resolution. Ultimately, the quantitative precipitation forecast is improved at a higher grid resolution. Additionally, the duration of heavy precipitation just onshore, where convection collapses, is better predicted.

How to cite: Jensen, A., Geerts, B., and Bergmaier, P.: Sensitivity of convective cell dynamics and microphysics to model resolution for lake-effect shallow convection, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-6313, https://doi.org/10.5194/egusphere-egu2020-6313, 2020.

Mergers of cells in a severe convective weather on 22 July 2008 are simulated and analyzed by Mesoscale Model 5 (referred to as MM5）and radar network data. Observation results show that, the horizontal scale of the echo above 30 dBZ, which represent the small cells, is about 10 km, and the small cells that the echo centers are 20km apart merge into a larger cell at dozens of km of horizontal scale.. Mergers begin from the peripheral radar echo, and then strong central radar echo merges at the low level, at last, the acreage of strong radar echo increases after the merger. The contrast between the observations and the simulation results shows that they are consistent. Analysis on the simulation results of two kinds of cell mergers at different development stages based on the third network model output shows that, while the cell pairs are with almost the same intensity, cells would develop after merger; while one of the cell pairs is in stronger development however the other one weaker, the stronger cell would keep on development and the weaker cell would die out. During the merger, a new cloud water center appears in the low convergence region between the cell pairs, and would replace the two cloud water centers of the former cells, or the new cloud water center would merger with one of the old cloud water centers while the other old cloud water center disappears. The analysis of the simulation results also shows that, the cell merger would lead to the cloud top lifting and the increase in the radar echo, content of cloud water and ice, surface rainfall.

How to cite: Zhai, J. and Huang, Y.: Numerical Simulations of Convective Cloud Merging Processes at Different Development Stages, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-6683, https://doi.org/10.5194/egusphere-egu2020-6683, 2020.

The tropospheric lapse rate in the tropics follows a moist adiabat quite closely and is mainly set by surface temperature and humidity in the convecting regions. Therefore, warming or biases at the surface are transferred via the moist adiabat to the upper troposphere. However, climate models show large discrepancies in the upper troposphere and recent observed upper tropospheric warming is around 0.5K weaker than predicted by the moist adiabat theory. Here we use the control simulations of the CMIP5 ensemble to show that large differences in the upper troposphere exist in the mean state that are unrelated to inter-model differences in the lower troposphere. In fact, CMIP5 models diverge (positively and negatively) from the moist pseudoadiabat by up to 2K at 300hPa. Precipitation weighted SSTs have recently been used to resolve the discrepancy between models and observations in upper tropospheric warming, but we show that they are not able to explain the differences in the mean state. While it is difficult to exactly depict the reasons for the inter-model spread, we demonstrate how the upper tropospheric lapse rate can deviate from the moist adiabat for the same lower tropospheric state with AMIP experiments. For this we use the ICON-A model, in which we tune convective and microphysical parameters. An improved understanding of the effect of different parameterisations on the models' lapse rates may help to better understand differences in the response to global warming.

How to cite: Keil, P., Schmidt, H., and Stevens, B.: Lapse rate deviations from the moist adiabat in the tropical upper troposphere in climate models, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-9467, https://doi.org/10.5194/egusphere-egu2020-9467, 2020.

Cold pool generated by convective clouds is an evaporatively cooled dry region which spreads out near the surface. Studying the cold pool characteristics enhances our understanding about convective clouds such as shallow-to-deep transition of convective clouds, long-lived squall line, and triggering secondary convection. In this study, cold pools over Germany are detected and characterized using phase 0 results of DYAMOND (stands for DYnamics of the Atmospheric general circulation Modeled On Non-hydrostatic Domains) intercomparison project. We aim to understand how the cold pool characteristics over Germany depend on topographic height, accompanying cloud size, and model.

Nine model results of the DYAMOND collection are remapped into 0.1˚ × 0.1˚ regular grid system. Cold pool cluster is defined as a cluster with an area larger than ~64 km² (4 grids), with the perturbation virtual (density) potential temperature below 2 K and the maximum precipitation rate greater than 1 mm h^–1. Detected cold pools are re-categorized by the topographic height to decompose cold pools related to orographic precipitation and by the accompanying cloud size to decompose cold pools related to large cloud system.

During simulated period (40 days from 1 August 2016), model averaged total detected cold pool number is 5.59 h^–1. Although more number of cold pool clusters are detected over low topographic area (1.34 h^–1 and 4.25 h^–1 over high and low area, respectively), area weighted cold pool cluster number is 3.82 times larger over high topographic area (17.55 h^–1 and 4.60 h^–1 over high and low area, respectively). Most of cold pool clusters are accompanied by larger clouds than themselves (78 %) and 9 % of cold pools are detected outside of cloud cover. Except for the cold pools accompanied by clouds of synoptic low pressure system, most of cold pools are detected in the daytime. Cold pool clusters over high topographic area are larger, more non-circular shaped, colder, and with lower wind speed than those over low topographic area. Cold pool clusters accompanied by small clouds are colder, drier, with higher wind speed, and with stronger precipitation than those accompanied by large clouds. In this study, relationship between cold pool characteristic parameters in each category is also investigated. To understand how cold pool feature varies from model to model, the cold pool characteristic parameters in each DYAMOND model result are compared and analyzed.

How to cite: Seo, J. M. and Hohenegger, C.: Detection and Characterization of Cold Pools over Germany in the DYAMOND Simulations, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-9711, https://doi.org/10.5194/egusphere-egu2020-9711, 2020.

Cold pools are produced by rain evaporation from
convective thunderstorms and play an important role 
in many atmospheric phenomena (e.g. transition to deep convection and convective self-aggregation). From observational
and numerical studies, it has been found that intersecting cold pools
increase the likelihood of triggering convection.
We test this hypothesis by combining observational
radar data from Darwin (Australia) with a simple conceptual model.

We identify precipitation objects in the radar data. It is assumed that each rain event produces a cold pool
that is initialized at the center of the precipitation cell. Cold pools are simulated with a stochastic surface growth model.
The spatial coordinate of each collision event is recorded. 
Collectively these points take the shape of a Voronoi diagram. 
According to our hypothesis, the probability of new rain events should decay with spatial distance to the Voronoi.

Our preliminary results suggest that rain events cluster in the
vicinity of the Voronoi with a higher frequency that one would expect if cold pool collisions did not stimulate convection. 
To conclude, our findings suggest that dynamic collisions between cold pools increase the likelihood of convection in the surrounding area.
This work allows us to study the effect of cold pools from radar data, despite cold pools being invisible to the radar images,
using a simple object-based model of convective cold pools. 

How to cite: Halvorsen, G., Meyer, B., and Härter, J.: Cold pool collisions as a triggering mechanism of convection, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-12367, https://doi.org/10.5194/egusphere-egu2020-12367, 2020.

We present results of radiative convective equilibrium runs using the WRF model coupled to an interactive slab ocean model, for which a relaxation term removes energy to constrain the domain mean sea surface temperature to a target value over a given timescale. By using a short adjustment timescale of one minute, drift in the mean temperature is constrained and the impact of the slab ocean is only through the spatial heterogeneity in sea  surface flux. We show how thin slabs slow the onset of organization, and conduct sensitivity experiments to determine the relative contributions of the radiative, sensible and latent surface fluxes, with surface fluxes key. Once clustering starts, the surface feedback acts to aid organization onset due to the drying atmosphere, although the speed of clustering onset is not significantly changed, indicating that it could be determined by a water vapour diffusive timescale as suggested by Windmiller and Craig.  An additional set of experiments that permit the mean surface temperature to undergo a diurnal adjustment show how diurnal variations in SST oppose the atmospheric radiative forcing and also act to prevent clustering onset. We show the mechanism for this acts through the reduction of the diurnal variation of convective mass flux and the distance between updraft towers.

How to cite: Semie, A. G. and Tompkins, A. M.: Impact of interactive sea surface temperature on convective clustering in radiative convective equilibrium, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-14171, https://doi.org/10.5194/egusphere-egu2020-14171, 2020.

Thermally-driven circulations in mountainous terrain can play an essential role in the initiation of deep moist convection: They advect moisture at low levels and provide the necessary trigger mechanism to lift air parcels above the level of free convection.
Current limited-area numerical weather prediction models with a horizontal grid spacing of around 1 km may adequately resolve the larger-scale thermal circulations, namely, valley winds and plain-to-mountain winds, but not the small-scale slope winds. In addition, the planetary boundary-layer parametrizations typically employed in these models are based on the assumption of horizontally homogeneous and flat terrain and assume none of the turbulent boundary-layer eddies are explicitly resolved.
In this contribution, we investigate the problems that arise due to these deficiencies in the given context using idealized numerical simulations with the WRF model. We compare simulations at different horizontal resolutions in the turbulence gray zone with LES simulations. Previous idealized modeling studies have shown that simulations at kilometer-scale resolution may produce stronger moisture convergence due to thermally-driven circulations and thus earlier and more vigorous convection over the mountain ridges compared with an LES model.
We focus on strongly-inhibited initial conditions that lead to deep moist convection with a kilometer-scale but not with an LES model and investigate the reasons for the different convective behavior. The benefits of scale-adaptive boundary-layer schemes for the studied process are evaluated.

How to cite: Göbel, M., Serafin, S., and Rotach, M.: Model resolution dependence of convection initiation by orographically-induced thermal circulations, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-18167, https://doi.org/10.5194/egusphere-egu2020-18167, 2020.

When precipitation evaporates in a sub-saturated boundary layer, it cools the air and produces dense downdrafts, which flow towards the surface and can spread horizontally as a gust front. These spreading “cold pools” (CPs) can trigger convection and thus new precipitation events due to dynamical and thermodynamical lifting mechanisms. Due to their role in the local organization of convection, CP properties are currently being studied with the use of high-resolution numerical simulations. Measurement campaigns have been conducted over the ocean to validate the models. However, fewer studies have specifically targeted cold pools over land.

We use the observational network of the Netherlands (meteorological stations and radar) to study CPs developing from summer convection and their role in triggering new convective events over land. Detailed information about CP gust fronts in terms of temperature, wind speed, heat fluxes, moisture and pressure at high vertical resolution is obtained from time series, measured at the 213-meter Cabauw tower. We aim to create an algorithm that detects the passage of a CP from the tower time series to automatize the finding of CPs from a point measurement. To confirm the results, we have access to temperature time series from a spatially dense crowdsourcing weather station network (WOW-NL).

The properties of the detected CPs are further studied with imagery from the Herwijnen Doppler radar, situated in proximity to the Cabauw tower. We can see clear signatures of spreading CPs in reflectivity plots, probably caused by the upwelling of dust and insects in the gust front. We currently explore how this can serve as a direct way of visualizing the dynamics of CPs and their collisions.

With enough observations of CPs, we expect to learn more about the CP spreading velocity and lifetime in dependence of precipitation intensity of the generating precipitation cell and eventually triggered cell. This link will help gain more insight into the role of CPs in organizing convection over land.

How to cite: Kruse, I. L., Haerter, J. O., and Meyer, B.: Characterizing cold pool interactions over land with observational data from the Netherlands, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-18494, https://doi.org/10.5194/egusphere-egu2020-18494, 2020.

Self-aggregation of convective cloud activity has attracted a lot of attention due to its role in the emergence of large scale weather phenomena such as the formation of hurricanes. Simulations with uniform boundary conditions show self-aggregation ocurring on the time-scale of a few weeks [1]. Recently, numerical experiments demonstrated that spatial clustering can form withing a few days, when the system is driven by diurnal temperature oscillations [2]. These simulations indicate that there may be a discontinuous phase transition between the clustered and the non-clustered state, i.e. that a threshold for the amplitudes of diurnal temperature oscillations exist, below which clustering does not emerge. A conceptual model has been proposed, suggesting that the phase transition might give rise to hysteresis in the sense that clustering emerging at high temperature amplitudes might persist if the amplitude is subsequently reduced to a level below the critical threshold.

Here we test the hysteresis–hypothesis explicitly by performing cloud-resolving simulations with a high-amplitude transient period followed by a period with a low diurnal temperature amplitude. In reality, diurnal temperature oscillations have significantly larger amplitudes over land than over the ocean. The existence of hysteresis effects in the convective cloud clustering could have profound implications: clusters formed over land, where diurnal temperature variations are large, could persist over the ocean when transported there by large-scale wind advection. Once present, the clusters could even intensify over the ocean—with possible implications for cyclogenesis.

[1] CJ Muller and IM Held, Journal of the Atmospheric Sciences, 69(8):2551–2565, 2012.
[2] arXiv:2001.04740 [physics.ao-ph]

How to cite: Jensen, G. G. and Haerter, J.: Hysteresis in self-organized mesoscale convective systems driven by diurnal temperature oscillations, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-20870, https://doi.org/10.5194/egusphere-egu2020-20870, 2020.

Deep convective rain events in the tropics represent an essential ingredient of Earth's climate system, acting as a driver of large-scale circulation which distributes energy and moisture. Understanding how they organize remains a challenge. Observational studies indicate that tropical convection may be understood as an instance of self-organized criticality (SOC) [1]:
(i) Rain rate vs. column water vapor follows a clear "pickup curve": Essentially no rain is observed in dry areas, while at column moistures above a critical value the rain rate increases sharply.
(ii) Rain events and clusters, defined as groups of contiguous rainy points in space and/or time, have size distributions well described by power laws.
The first result indicates that the atmosphere undergoes a phase transition, separating a non-raining "inactive" phase from a rainy "active" one. The second result suggests that the system is found close to the critical transition point, where "scale-free" power law distributions are expected. Indeed, observations find typical moisture values to be close to the critical moisture value. 

SOC theory would suggest that the observational results are an emergent phenomenon, caused by simple local interactions that carry over to larger scales. However, to our knowledge, no simple SOC model linking moisture and rainfall has been suggested that explains how criticality arises from convective processes while also predicting the observed rain cluster sizes. A more complete theory, especially on spatial aspects, is lacking.
We therefore present a simple spatiotemporal model of the atmospheric water budget, exploring whether a fuller picture including spatial information can be developed. Each site of the model represents an atmospheric column, where water can enter through surface evaporation, leave as surface rain, or get redistributed among neighboring sites due to convective in- and outflows. We analyse a cloud resolving model simulation in radiative-convective equilibrium, by grouping grid points into three categories: rainy points (convectively active), neighbors of rainy points and others (convectively passive). Tendencies, evolutions and transitions are examined to identify local "rules" to inform the interactions and parameters in our model.
Hence, in this project we use a simple model approach to find whether local convection mechanisms of water rearrangements can explain why tropical rainfall seems to show critical characteristics. In addition, this might aid development of convective parameterizations for climate models: Including a few key convective scale interactions, suggested by our model, might help to better capture important effects of subgrid correlations in a simple way.
 
[1] Peters, O., and J. D. Neelin (2006), Critical phenomena in atmospheric precipitation, Nat. Phys., 2(6), 393-396, doi:10.1038/nphys314.

How to cite: Svendsen, M.: Criticality in Tropical Rainfall: A Simple Water Budget Model, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-22156, https://doi.org/10.5194/egusphere-egu2020-22156, 2020.

Squall lines are the prominent feature over Singapore region creating strongly localized rain events due to vigorous localized convective activity. These convective systems have relatively small spatial and temporal scales compared to other atmospheric features like monsoons, thus the prediction of these features lack accuracy. The SINGV numerical weather prediction model is able to provide improved weather forecasts over Singapore region, however, challenges still exist in predicting the thunderstorm/squall line events in onset, location, intensity and lead time. A few real-time case studies of squall lines indicate that SINGV could not capture these features appropriately, while WRF did a better forecasting. To understand the issues with SINGV model, idealized simulations replicating the Weismann & Klemp ‘82 case are conducted keeping similar physics in both the models. Preliminary results indicate that both models behave differently: WRF displays organized convection whereas in SINGV the storm splits at the early stages. Cross-sectional details along the propagating squall line suggest that the updrafts and downdrafts, at the storm development stages, are moderately higher in SINGV compared to WRF. It is speculated that these stronger updrafts in SINGV carry anomalously large amount of liquid water to the upper troposphere where these are converted into rain, which in turn result in stronger downdrafts facilitating the splitting of initial storm. Further analysis is required to conclude our speculation.

How to cite: Rajagopalan, R., Dipankar, A., and Huang, X.-Y.: Numerical Simulation of Squall line in idealized SINGV and WRF Models, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-15974, https://doi.org/10.5194/egusphere-egu2020-15974, 2020.