Atmospheric Convection 

This session deals with atmospheric convection, being dry, shallow, or deep convection. Contributions on these aspects resulting from the use of large-eddy simulations, convection-permitting simulations, coarser-resolution simulations using parameterised convection and observations are welcome. Studies that investigate the organization of convection, being in idealized set-ups (radiative convective equilibrium and self-aggregation) or in observations, as well as studies that investigate the importance of organization for climate are particularly welcome. Besides this, studies that investigate general aspects of convection such as processes controlling the lifecycle of convection, interactions of convection with other physical processes and representation of convection in numerical weather prediction and climate models are also welcome.

Convener: Cathy Hohenegger | Co-conveners: Leo Donner, Adrian Tompkins, Holger Tost
vPICO presentations
| Wed, 28 Apr, 15:30–17:00 (CEST)

vPICO presentations: Wed, 28 Apr

Kieran Pope, Christopher Holloway, Thorwald Stein, Todd Jones, and Michael Whitall

Numerical models configured in radiative convective equilibrium (RCE) have shown that convection can self-aggregate, where initially randomly distributed convection becomes clustered despite homogeneous initial conditions and forcing. The degree of self-aggregation within a domain has important consequences for weather and climate, and varies considerably between models. Previous studies have shown that interactions between clouds and radiation are crucial drivers and maintainers of aggregation. This study investigates the direct radiative-convective feedbacks that are important to self-aggregation within elongated channel simulations of the UK Met Office Unified Model (UM) version 11.0. Our simulations are configured using three fixed sea surface temperatures (SSTs) following the radiative-convective equilibrium model intercomparison project (RCEMIP) protocol.

Our analysis builds on the vertically-integrated frozen moist static energy (FMSE) variance budget framework that assumes that aggregation increases as FMSE variance increases. The budget shows how interactions between FMSE and radiation, surface fluxes and advection contribute to increasing FMSE variance. This variance is highly sensitive to SST, however, by normalising FMSE between theoretical upper and lower limits based on SST, this sensitivity can be eliminated. This allows the variance of normalised FMSE to be a consistent aggregation metric across all SSTs. By deriving a new budget equation for normalised FMSE, we can see which interactions are important for aggregation and how these interactions are sensitive to SST. By defining cloud types based on the vertical distribution of condensed water, we study the importance of radiative interactions with each cloud type to aggregation, and how they change with SST.

We find that our simulations reach similar degrees of aggregation, despite the contributions of shortwave and longwave interactions decreasing with SST. Surface flux and advective feedbacks with FMSE become less negative with SST, accounting for the decreasing radiative feedback contribution. Longwave interactions with high-topped clouds are the main drivers and maintainers of aggregation, with their influence decreasing with SST as high clouds become less abundant. Longwave interactions in clear regions have significant positive effects in driving aggregation, however their contributions decrease once the convection becomes aggregated. Their longwave contributions to aggregation decrease with SST and can become negative at high SSTs once convection is aggregated. The shortwave interactions with water vapour are one of the key maintainers of aggregation, becoming more important as aggregation increases. Shortwave interactions are more important at cooler SSTs where there is a greater contrast in shortwave heating between moist and dry regions.

Results presented here are not necessarily representative of real world convection, these are merely results of this model configuration. Applying this technique to other models may highlight key differences in their cloud-radiative feedbacks and may help to explain differences in the degree of aggregation within numerical models.  

How to cite: Pope, K., Holloway, C., Stein, T., Jones, T., and Whitall, M.: Cloud-Radiation Interactions and their contributions to Convective Self-Aggregation, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-4731,, 2021.

Sophie Abramian, Caroline Muller, and Camille Risi

Investigating tropical squall lines with a cloud resolving model

Using a cloud resolving model, we attempt to clarify the physical processes responsible for the organization of deep clouds into squall lines in the tropics. To do so, we impose a vertical wind shear, and investigate the response of deep convection to different shear strengths in radiative convective equilibrium. As the magnitude of the shear increases, the convection becomes more and more organized into a line, perpendicular to the shear. It is due to the interaction of the low-level shear with the cold pools associated with convective downdrafts. Beyond a certain shear, called optimal shear, the line tends to orient at an angle to the shear. The existing literature suggests that this angle conserves the projection of the shear on the direction perpendicular to the squall line near the optimal value, a hypothesis that we further investigate here.

In this work, we propose a systematic method, based on image auto-correlation, to determine the angle of the squall line with respect to the shear. We highlight the existence of the sub-critical and super-critical regime, as predicted by earlier studies. In the sub-critical regime, squall lines are indeed perpendicular to the shear. Yet, angles of squall lines in the super-critical regime do not clearly correspond to the conservation of the projected component of the shear near the optimal value. In particular, squall lines often remain more perpendicular to the shear than expected.

We thus investigate the balance between shear and cold pool winds to explain this difference. Using statistical methods on extreme events, we find that this difference is due to an intensification of cold pool potential energy with shear. Cold pool intensification allows the squall line to better resist to the shear, and thus reduces its angle of orientation. This new feature leads us to conclude that two mechanisms maintain a squall line in wind shear : the orientation of clouds and the intensification of cold pools.

How to cite: Abramian, S., Muller, C., and Risi, C.: Investigating tropical squall lines with a cloud resolving model, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-4434,, 2021.

Silas Boye Nissen and Jan O. Haerter

In radiative-convective equilibrium (RCE) simulations, convective self-aggregation (CSA) is the spontaneous organization into segregated cloudy and cloud-free regions. Evidence exists for how CSA is stabilized, but how it arises favorably on large domains is not settled. Using large-eddy simulations (LES), we link the spatial organization emerging from the interaction of cold pools (CPs) to CSA. We systematically weaken simulated rain evaporation to reduce maximal CP radii, Rmax, and find reducing Rmax causes CSA to occur earlier. We further identify a typical rain cell generation time and a minimum radius, Rmin, around a given rain cell, within which the formation of subsequent rain cells is suppressed. Incorporating Rmin and Rmax, we propose a toy model that captures how CSA arises earlier on large domains: when two CPs of radii ri,j ∈ [Rmin, Rmax] collide, they form a new convective event. These findings imply that CPs play a crucial role in RCE simulations by preventing the onset of CSA.

How to cite: Nissen, S. B. and Haerter, J. O.: How weakened cold pools open for convective self-aggregation, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-1520,, 2021.

Ronja Gronemeyer, Gorm Gruner Jensen, and Jan O. Haerter

Under radiative convective equilibrium (RCE), cloud populations simulated by large-eddy simulations (LES) can spontaneously segregate into cloudy and cloud-free subregions. This process is well-known as convective self-aggregation (CSA) [4]. But how initially randomly distributed raincells compete and merge until only one prevails, is not well-understood. We remove cold pools (CPs) in LES by suppressing the re-evaporation of rain, which leads to qualitatively different dynamics. This extreme case helps to understand the role of CPs in the formation of CSA and further has relevance when humidity is very high in the boundary layer, so very little rainfall evaporates.

When convection starts, patterns of high and low moisture develop, which increase in scale over time. In contrast to CSA with CPs, individual rain events and convection cells persist up to tens of hours in the course of this modified CSA [1]. For the long lasting individual rain clusters, we observe interesting oscillations in rain intensity and spatial extent. We define an algorithm, that tracks the tree-like merging behavior of initially many individual small raincells to a final, single, raincell of large area and precipitation yield. We conceptualize the LES behavior in a simple model, that assumes different rain events to compete for buoyancy. This hypothesis is justified when viewing rain events as linked to local maxima of relative humidity around cloud base. The clusters‘ dynamics seem to be dominated by merging with other events and ’stealing’ from smaller events, whereas splitting and emerging of new rain events seem neglectable after a build-up time. In each step, the conceptual model chooses two adjacent clusters. Initially, each cluster is attributed a ‘mass’ parameter of similar magnitude and a fraction (p) of the smaller ’mass’ (m2) is transferred to the bigger event (m1).

m1new = m1 + p(m1m2)
m2new = m2 - p(m1m2)

An event is removed, when its mass parameter is diminished to zero. In contrast to field based approaches [3], this approach implements discrete rich gets richer dynamics, to capture how individual cells grow. This conceptual model could be combined with existing models, where CP suppress the rain cells, but trigger new updrafts through the CP gust fronts [2]. Bringing together these two limits could further elucidate how CP dynamics can be made compatible with convective self-aggregation.

[1] Nadir Jeevanjee and David M Romps. Convective self-aggregation, cold pools, and domain size. Geophysical Research Letters, 40(5):994–998, 2013.

[2] Silas Boye Nissen and Jan O. Haerter. How weakened cold pools open for convective self-aggregation, 2020, arXiv:1911.12849v3.

[3] Julia M. Windmiller and George C. Craig. Universality in the spatial evolution of self-aggregation of tropical convection. Journal of the Atmospheric Sciences, 76(6):1677 – 1696, 01 Jun. 2019.

[4] Allison A Wing, Kerry Emanuel, Christopher E Holloway, and Caroline Muller. Convective self-aggregation in numerical simulations: A review. In Shallow Clouds, Water Vapor, Circulation, and Climate Sensitivity, pages 1–25. Springer, 2017.

How to cite: Gronemeyer, R., Jensen, G. G., and Haerter, J. O.: The winner takes it all - how long-lived raincells compete in cold pool-suppressed self-aggregation, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-6515,, 2021.

Romain Fiévet, Bettina Meyer, and Jan Olaf Haerter

Spontaneous aggregation of clouds is a puzzling phenomenon observed in field studies [Holloway et al. (2017)] and idealized simulations alike [Held et al. (1993), Bretherton et al. (2005)]. With its relevance to climate sensitivity and extreme events, aggregation continues to be heavily studied, [Wing et al., 2017 for a review], with radiative-convective feedbacks emerging as main drivers of simulated convective self-aggregation (CSA) [Mueller & Bony (2015)].

In state-of-the art cloud-resolving models, CSA finds itself consistently hampered by finer horizontal resolutions [Muller & Held (2012), Yanase et al. (2020)]. This feature was ascribed to the effect of cold pool (CP) gust fronts in opposing the positive moisture feedback underlying CSA [Jeevanjee & Romps (2013)]. In contrast, recent numerical experiments [Haerter et al. (2020)] with diurnally oscillating surface temperature highlights an orthogonal effect: stronger CPs, driven by small-scale density gradients, promote cloud field self-organization into mesoscale convective systems (MCS). Interestingly, this upscale growth, which we here term diurnal self-organisation (DSO), differs from classical CSA as it is driven by CPs rather than large-scale radiative imbalances. In stark contrast to CSA, strengthening CPs promotes this organization effect.

Hence, numerical simulations of CSA and DSO should go beyond the typical cloud-resolving paradigm and achieve cold pool-resolving capabilities. The current study systematically examines the impact of model resolution on CP effects. First, numerical convergence is probed in a 12km x 20km laterally periodic domain where a single CP propagates and self-collides at the domain's edges. As the spatial resolution is stepwise increased from 250 to 25m, it is shown that the initially coarsely resolved density current dissipates and collision and updraft effects are weak. As finer resolution is approached, we identify a cold pool resolving resolution D, which is deemed satisfactory for propagation and collision properties. Second, convergence for a (250km)2 domain under a diurnal radiative cycle is assessed at various spatial resolutions, including the scale D. This mesoscale configuration allows us to quantify the impact of resolution of cold pool dynamics on DSO.

Together, this work systematically lays out the numerical requirements to study mesoscale clustering by means of explicit numerical simulations.

How to cite: Fiévet, R., Meyer, B., and Haerter, J. O.: Defining a cold pool-resolving scale for numerical simulations of convective self-organisation, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-5874,, 2021.

Nicolas Rochetin, Cathy Hohenegger, Ludovic Touzé-Peiffer, and Najda Villefranque

In this paper, a conceptual model to dene density currents is proposed. Based on theory, observations and modelling studies, we dene convective density currents as 3D coherent structures with an anomalously cold core, an adjacent wind gust and two vertical layers: a well-mixed one near the surface and a stratied one above. With this definition, a methodology to identify and label individual density currents in convection permitting simulations is designed. The method is illustrated through its application to four distinct cloud scenes issued from a convetion-permitting simulation. From this methodology, new dynamic, thermodynamic and geometric features related to the density currents imprint on the Planetary Boundary Layer are revealed. The method is found to be i) robust in time, ii) relevant in distinct convective regimes, iii) relevant in land and oceanic situations and iv) adapted to both Cloud Resolving Models and Large Eddy Simulations. It also provides proxies such as the number, the spatial coverage, the mean radius and the mean velocity of density currents, from which a detailed analysis of their role in convection life-cycle and spatial organization could be performed in the near future.

How to cite: Rochetin, N., Hohenegger, C., Touzé-Peiffer, L., and Villefranque, N.: A physically-based robust definition of convectively generated density currents : detection and characterization in convection-permitting simulations, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-7674,, 2021.

Bastian Kirsch, Cathy Hohenegger, Daniel Klocke, and Felix Ament

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 explicitly resolve cold pools, however, they lack reference data for an adequate validation. Available operational networks are too coarse and, therefore, miss the horizontal structure and dynamics of cold pools.

The pioneering field experiment FESST@HH aims to shed light on this observational blind spot. During summer 2020 a dense network of 102 ground-based stations covering the greater area of Hamburg (Germany) realized meteorological measurements at sub-mesoscale resolution (Δx < 2 km, Δt ≤ 10 s), that provide novel insights into previously unobserved features of cold pools. Over three months more than 30 cold-pool events of different strength and size from various types of convection were detected. Analyses of prominent cases suggest a strong relationship between the local perturbations in air temperature and pressure within a cold pool, that allows inference about its vertical depth based on the hydrostatic assumption. Furthermore, temporary decoupling of horizontal variability in these signals reveal the presence of local non-hydrostatic pressure perturbations caused by convective downdrafts. The presented work will help to better understand the characteristics and life cycle of cold pools and to identify potential biases in convection-permitting simulations.

How to cite: Kirsch, B., Hohenegger, C., Klocke, D., and Ament, F.: Sub-mesoscale cold-pool observations during FESST@HH 2020, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-1333,, 2021.

Ákos János Varga and Hajnalka Breuer

Atmospheric convection is a major source of hazardous weather in many parts of the world, including the Pannonian Basin in Central Europe. It is therefore indispensable to explore the climatological distribution of convective environments and related phenomena, for which model data provide a spatially and temporally consistent alternative.

Several studies compared convective parameters derived from reanalysis datasets to radiosonde observations, but such evaluation of climate model output is less frequent.

This study uses sounding station measurements to verify convective environmental parameters derived from the ERA5 reanalysis and relatively coarse resolution WRF regional climate simulations for the 1985–2010 period over the wider Pannonian Basin region. Common parcel thermodynamic variables and environmental indices are calculated, such as CAPE, CIN, LI, TPW, lapse rate and wind shear. We carry out pointwise comparison between observed and modeled convective parameters in terms of basic statistical metrics and climatological means on a daily and monthly basis. Both pressure and model level data from ERA5 are included in the analysis.

In line with previous research, the ERA5 model level dataset reasonably represents the climatological distribution of convection-related variables. Preliminary results suggest that the WRF regional climate model is also quite skillful in reproducing convective environments, but large biases exist compared to the observations and reanalysis.

The research was supported by the Hungarian National Research, Development and Innovation Office, Grant No. FK132014.

How to cite: Varga, Á. J. and Breuer, H.: Verification of climatological ERA5 and WRF convective environments using radiosonde data, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-2550,, 2021.

Ting-Chen Chen, Man-Kong Yau, and Daniel J. Kirshbaum

     Slantwise convection and the associated release of conditional symmetric instability (CSI) have been recognized as important baroclinic processes. Recent climatological studies have highlighted its significant association with midlatitude cyclone activities, raising questions about whether large-scale models can resolve slantwise convection and whether it should be parameterized.

     To address this issue, the present study simulates isolated free moist slantwise convection in an initially statically stable environment using the 2D idealized, non-hydrostatic Weather Research and Forecasting (WRF) Model. We first examined the sensitivity of the slantwise convection to the cross-band grid spacing (Δy; varied from 40 to 1 km) and found that experiments with ∆y> 5 km fail to capture the band dynamics and larger-scale feedbacks robustly and thus require parameterization. As most of the current convective parameterization schemes target upright convection in a local column, we implemented an additional 2D slantwise convective parameterization scheme and evaluated its impact for coarse-grid runs.

     The slantwise convective parameterization scheme operates along a sloped trajectory on a horizontally-variant cross section perpendicular to the local thermal wind, adjusting the environment toward a natural state to CSI within a given time scale. With the addition of the slantwise convective parameterization scheme, significant improvements are found in precipitation and the strength of the slantwise updraft, bringing the coarser-grid (∆y=40 km) simulation closer to the finer-grid (converged) results than its counterpart with only the upright convection scheme. After testing the slantwise convective parameterization scheme under idealized frameworks, we will further apply it to regional models to evaluate its benefit to the weather forecasting in real cases.

How to cite: Chen, T.-C., Yau, M.-K., and Kirshbaum, D. J.: The parameterization of slantwise convection in a numerical model, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-9136,, 2021.

Convective organization discussion
Blaž Gasparini, Adam Sokol, Casey Wall, Dennis Hartmann, and Peter Blossey

Geostationary satellite observations of tropical maritime convection indicate an afternoon maximum in anvil cloud fraction that cannot be explained by the diurnal cycle of deep convection peaking in the night. This implies that the daytime anvils must be more widespread and/or long lived compared with the anvils that are formed during the night.

We study the decay of anvil clouds in an idealized cloud resolving modelling setup in which a cloud is initialized in the middle of the model domain to identify what causes differences in the evolution depending on the time of the day in which the cloud is detrained from a deep convective core. We show that daytime anvils are both longer lived and more widespread. The main reason for their longevity is the heating due to absorption of shortwave radiation, which leads to a mesoscale ascent within the cloud, helping to loft and spread the cloud further than the nighttime anvils. The nighttime anvil cloud top is dominated by longwave radiative cooling, which drives a circulation that erodes the cloud top by entrainment of drier environmental air and leads to a cloud descent and shorter lifetime. 

Additional simulations in radiative convective equilibrium setup with a realistic diurnal cycle of insolation confirm the crucial role of shortwave heating in increasing the daytime anvil cloud top and anvil longevity. In addition, the mesoscale ascent also modifies daytime anvil properties, leading to an increased ice water content, higher ice crystal number concentration and larger ice crystal radius near cloud top.

How to cite: Gasparini, B., Sokol, A., Wall, C., Hartmann, D., and Blossey, P.: A modelling perspective on anvil evolution differences between day and night, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-785,, 2021.

Gorm Gruner Jensen, Romain Fiévet, and Jan O. Haerter

Convective self-aggregation (CSA) is an established modelling paradigm for large-scale thunderstorm clusters, as they form in mesoscale convective systems, the Madden-JulianOscillation or tropical cyclo-genesis [1]. The onset of CSA is characterized by the spontaneous formation of persistently dry patches with suppressed deep convective rainfall. Recently another type of spatio-temporal pattern formation was observed in simulations where the diurnal cycle was mimicked by a sinusoidally varying surface temperature [2]. This diurnal aggregation (DA) is characterized by clusters of intense rain that correlate negatively from one day to the next. 

Here we demonstrate that the diurnal cycle can also act as a trigger of persistently dry patches resembling the early stages of CSA. When the surface temperature is held constant, CSA has been shown to occur within simulations of coarse horizontal model resolution, but not when the resolution was increased [3]. We show that, when a temporally periodic surface temperature forcing is imposed, persistently convection free patches occur even faster when the spatial resolution is increased. The failure to achieve CSA at high horizontal resolution has so far been attributed to the more pronounced cold pool effects at such resolution. In our simulations these cold pools in fact play a key role in promoting CSA. Our results have implications for the origin of persistent convective organization over continents and the sea — and point a path towards achieving such clustering under realistic conditions.

[1]  Christopher S Bretherton, Peter N Blossey, and Marat Khairoutdinov.  An energy-balance analysisof deep convective self-aggregation above uniform SST.Journal of the Atmospheric Sciences, 62(12):4273–4292, 2005.
[2]  J. O. Haerter, B. Meyer, and S. B. Nissen.  Diurnal self-aggregation.npj Climate and AtmosphericScience, 3:30, 2020.
[3]  Caroline  Muller  and  Sandrine  Bony.   What  favors  convective  aggregation  and  why?GeophysicalResearch Letters, 42(13):5626–5634, 2015.  doi:

How to cite: Jensen, G. G., Fiévet, R., and Haerter, J. O.: The diurnal cycle can trigger convective self-aggregation , EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-16096,, 2021.

Jan O. Haerter, Gorm Gruner Jensen, and Romain Fiévet

Convective self-aggregation is a well-studied atmospheric state, obtained in typically multi-week idealized numerical experiments, where boundary conditions are constant and spatially homogeneous. As radiative convective equilibrium is approached, the atmosphere develops a heavily precipitating moist patch, which is surrounded by subsiding, cloud-free regions. It was recently shown that a homogeneous, but temporally oscillating surface temperature can quickly lead to the emergence of so-called mesoscale convective systems (MCS, diameters of >100 km) - on temporal scales of only a few days. Furthermore, the patterns formed by these MCS remind of checkerboards, and alternate from day to day [1]. 

We here extend this finding further, to add realism to the otherwise preserved idealization: Mimicking a form of “miniature tropics” we retain a laterally periodic domain (Lx, Ly), but impose spatial variation in mean surface temperature along one dimension - reminiscent of a meridional reduction in mean surface temperature, when moving poleward from the equator. By making the wavelength of spatial variation commensurate with domain size, we retail double-periodic lateral boundary conditions. When the diurnal cycle is set to zero, the system quickly organizes to a forcefully aggregated caricature of the actual tropics - with heavy convection near the equator and pronounced subsidence and enhanced long-wave cooling in the subtropics. When the diurnal cycle is increased, bi-diurnal temporal oscillations appear, which lead to a single precipitation peak centered on the equator on one day, but a bimodal meridional pattern with precipitation away from the equator on the next.

Our findings, obtained for a still idealized numerical experiment, may have implications for “edge intensifications” suggested from observations and numerical modeling of tropical precipitation patterns near the ITCZ [2,3].

[1] Haerter, J.O., Meyer, B. & Nissen, S.B. Diurnal self-aggregation. npj Clim Atmos Sci 3, 30 (2020).

[2] Mapes, B. E., E.-S. Chung, W. M. Hannah, H. Masunaga, A. J. Wimmers and C. S. Velden, 2018: The meandering margin of the meteorological moist Tropics, Geophys. Res. Lett., 45, 1177-1184. doi:10.1002/2017GL076440

[3] Windmiller, J. M., & Hohenegger, C. 2019: Convection on the edge. J. Adv. Model. Earth Syst., 11, 3959-3972, 10.1029/2019MS001820

How to cite: Haerter, J. O., Jensen, G. G., and Fiévet, R.: Miniature tropics and bi-diurnal oscillations, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-3401,, 2021.

Chimene Laure Daleu

A series of high‐resolution three‐dimensional simulations of the diurnal cycle of deep convection over land are performed using the new Met Office NERC cloud‐resolving model. This study features scattered convection. A memory function is defined to identify the effects of previous convection in modifying current convection. It is based on the probability of finding rain at time t0 and at an earlier time t0−Δt compared to the expected probability given no memory. The memory is examined as a function of the lag time Δt. It is strongest at gray‐zone scales of 4–10 km, there is a change of behavior for spatial scales between 10 and 15 km, and it is reduced substantially for spatial scales larger than 25 km. At gray‐zone scales, there is a first phase of the memory function which represents the persistence of convection and it is maintained for about an hour. There is a second phase which represents the suppression of convection in regions which were raining 1 to 3 hr previously, and subsequently a third phase which represents a secondary enhancement of precipitation. The second and third phases of the memory function develop earlier for weaker forcing. When thermodynamic fluctuations resulting from the previous day are allowed to influence the development of convection on the next day, there are fewer rainfall events with relatively large sizes, which are more intense, and thus decay and recover more slowly, in comparison to the simulations where feedback from previous days is removed. Further sensitivity experiments reveal that convective memory attributed to these thermodynamic fluctuations resides in the lower troposphere.

How to cite: Daleu, C. L.: Memory Properties in Cloud‐Resolving Simulations of the Diurnal Cycle of Deep Convection., EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-4958,, 2021.

Diurnal cycle discussion
Basile Poujol, Andreas Prein, Caroline Muller, and Maria Molina

Organized convective systems produce heavier downpours and can become more intense with climate change. While organized convection is well studied in the tropics and mid-latitudes, few studies have focused on the physics and climate change impacts of pan-Arctic convective systems, where they can produce flash flooding, landslides, or ignite wildfires.

We use a convection-permitting model to simulate Alaska’s climate under current and end of the century high emission scenario conditions. We apply a precipitation tracking algorithm to identify intense, organized convective systems, which are projected to triple in frequency and extend to the northernmost regions of Alaska under future climate conditions. The present study assesses the reasons for this rapid increase in organized convection by investigating dynamic and thermodynamic changes within future storms and their environments, in light of canonical existing theories for mid-latitude and tropical deep convection.


In a future climate, more moisture originates from Arctic marine basins and relative humidity over continental Alaska is projected to increase due to sea ice loss, which is in sharp contrast to lower-latitude land regions that are expected to become drier. This increase in relative humidity favors the onset of organized convection through more unstable thermodynamic environments, increased low-level buoyancy, and weaker downdrafts.

Our confidence in these results is increased by showing that these changes can be analytically derived from basic physical laws. This suggests that organized thunderstorms might become more frequent in other pan-Arctic continental regions highlighting the uniqueness and vulnerability of these regions to climate change.

How to cite: Poujol, B., Prein, A., Muller, C., and Molina, M.: Dynamic and Thermodynamic Impacts of Climate Change on Organized Convection in Alaska, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-5942,, 2021.

Falco Bentvelsen, Geert Lenderink, and Pier Siebesma

We investigate the hypothesis that invigoration of convective updrafts under warming conditions contributes to the stronger than Clausius-Clapeyron (CC) scaling. Focus is on a mid-latitude case of extreme precipitation, based on idealised forcing conditions derived for the Netherlands, with strong surface forcing as well as strong forcing from large-scale rising motion associated with the passage of a synoptic scale low pressure or frontal system. Various Large Eddy Simulations (LES) of this composite case have been performed on a 192x192 km domain. By systematically perturbing the atmospheric temperature profile, a large response of cloud dynamics to warming with larger and more vigorous cloud structures in the warmer runs has been found.*

Here, we study these cloud dynamics further by investigating the vertical wind velocity in the cloud (cores). Updrafts play a key role in rain formation by transporting moisture upward in the clouds. We will demonstrate how the distributions of these vertical velocities near the surface and at different levels in the clouds respond to warming in this mid-latitude setting and how they relate to cloud properties as cell size and buoyancy.


*Lochbihler, K., Lenderink, G., and Siebesma, A. P. (2019). Response of extreme precipitating cellstructures to atmospheric warming. Journal of Geophysical Research: Atmospheres

How to cite: Bentvelsen, F., Lenderink, G., and Siebesma, P.: The role of updrafts in the scaling of extreme precipitation in mid-latitudes, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-13571,, 2021.

Nicolas Da Silva, Sara Shamekh, Caroline Muller, and Benjamin Fildier

Convective organisation has been associated with extreme precipitation in the tropics. Here we investigate the impact of convective self-aggregation on extreme rainfall rates. We find that convective self-aggregation significantly increases precipitation extremes, for 3-hourly accumulations but also for instantaneous rates (+ 30 %). We show that this latter enhanced instantaneous precipitation is mainly due to the local increase in relative humidity which drives larger accretion efficiency and lower re-evaporation and thus a higher precipitation efficiency.

An in-depth analysis based on an adapted scaling of precipitation extremes, reveals that the dynamic contribution decreases (- 25 %) while the thermodynamic is slightly enhanced (+ 5 %) with convective aggregation, leading to lower condensation rates (- 20 %). When the atmosphere is more organized into a moist convecting region, and a dry convection-free region, deep convective updrafts are surrounded by a warmer environment which reduces convective instability and thus the dynamic contribution. The moister boundary-layer explains the positive thermodynamic contribution. The microphysic contribution is increased by + 50 % with aggregation. The latter is partly due to reduced evaporation of rain falling through a moister near-cloud environment (+ 30 %), but also to the associated larger accretion efficiency (+ 20 %).

Thus, the change of convective organization regimes in a warming climate could lead to a much more different evolution of tropical precipitation extremes than expected from thermodynamical considerations. Improved fundamental understanding of convective organization and its sensitivity to warming, as well as its impact on precipitation extremes, is hence crucial to achieve accurate rainfall projections in a warming climate.

How to cite: Da Silva, N., Shamekh, S., Muller, C., and Fildier, B.: How does convective self-aggregation affect precipitation extremes?, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-5216,, 2021.

Caroline Muller and Takayabu Yukari

In this work, we review recent important advances in our understanding of the response of precipitation extremes to warming from theory and from idealized cloud-resolving simulations. A theoretical scaling for precipitation extremes has been proposed and refined in the past decades, allowing to address separately the contributions from the thermodynamics, the dynamics and the microphysics. Theoretical constraints, as well as remaining uncertainties, associated with each of these three contributions to precipitation extremes, will be discussed. Notably, although to leading order precipitation extremes seem to follow the thermodynamic theoretical expectation in idealized simulations, considerable uncertainty remains regarding the response of the dynamics and of the microphysics to warming, and considerable departure from this theoretical expectation is found in observations and in more realistic simulations. We also emphasize key outstanding questions, in particular the response of mesoscale convective organization to warming. Observations suggest that extreme rainfall often comes from organized system in very moist environments. Improved understanding of the physical processes behind convective organization is needed in order to achieve accurate extreme rainfall prediction in our current, and in a warming climate.

How to cite: Muller, C. and Yukari, T.: Response of precipitation extremes to warming: what have we learned from theory and idealized cloud-resolving simulations, and what remains to be learned?, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-4914,, 2021.

Precipitation change discussion
Tom Dror, Mickael D. Chekroun, Orit Altaratz, and Ilan Koren

Warm convective clouds play a key role in the Earth’s radiative and water budgets. Nonetheless, they still comprise the largest source of uncertainty in climate model’s prediction of cloud feedback and climate sensitivity. The latter might be affected by the variety of patterns that warm convective clouds form on the mesoscale, an effect which is largely uninvestigated, and even more so over land. A large subset of continental shallow convective cumulus (Cu) fields was shown to have unique spatial properties and to form mostly over forests and vegetated areas thus referred to as green Cu. Green Cu fields form organized mesoscale patterns, yet the underlying mechanisms, as well as the time variability of these patterns, are still lacking understanding.  In this work, we characterize the organization of green Cu in space and time, by using data-driven organization metrics, and by decomposing the high-resolution GOES–16 data using an Empirical Orthogonal Function (EOF) analysis. We extract and quantify modes of organization present in a green Cu field, during the course of a day. The EOF decomposition shows the field's key organization features such as cloud streets, and it also reveals hidden ones, as the propagation of gravity waves (GW), and the development of a highly ordered grid of clouds that extends over hundreds of kilometers, over a time span that scales as the field's lifetime. We then use cloud fields that were reconstructed from different subgroups of modes to quantify the cloud street's wavelength and aspect ratio, as well as the GW dominant period.

How to cite: Dror, T., Chekroun, M. D., Altaratz, O., and Koren, I.: EOF Analysis of Green Cumulus Mesoscale Organization, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-10884,, 2021.

Thirza van Laar and Roel Neggers

The spatial variability of Trade wind cumulus cloud fields has been found to be of great importance for understanding their role in Earth's climate system. In this study the focus is on the spacing between individual cumulus clouds. The main objective is to establish how inter-cloud spacing depends on cloud size, information that is crucial for understanding cloud-radiation interaction and spatial organization, and for informing grey zone parametrizations. To this end, a large-domain high resolution ICON LES simulation of marine shallow cumulus cloud fields is used. The domain is located at the subtropical Atlantic and the simulations are performed for the time of the recent NARVAL South campaign (December 2013). The simulations are compared with MODIS satellite imagery and research flight measurements, showing a good agreement between observations and the simulation, on both cloud size statistics and the vertical structure of the boundary layer.  

To determine the size and locations of the clouds, a cluster analysis was applied to the data. The inter-cloud, or nearest neighbor spacing (NNS), is addressed using four different expressions, classic definitions but also novel ones. For all definitions the spacing increases with cloud size, but the dependence is strongly influenced by the used definition. The classic definition, the spacing between clouds of any size, shows a well-defined logarithmic dependence on cloud size. The logarithmic relation can be explained by the abundance of closely-spaced small clouds. The small distances between these clouds form an upper bound for the NNS for the larger cloud sizes. In contrast, the spacing between clouds of a similar size increases exponentially with size. We argue that the exponential size-dependence reflects the mesoscale dynamics that drive the horizontal size of large convective cells, of which the cumulus clouds are the visible parts.  

How to cite: van Laar, T. and Neggers, R.: On the size dependence of cumulus cloud spacing, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-8377,, 2021.

Chris Holloway, Jian-Feng Gu, Bob Plant, and Todd Jones

The normalized distributions of thermodynamic and dynamical variables both within and outside shallow clouds are investigated through a composite algorithm using large eddy simulation of the BOMEX case. The normalized magnitude is maximum near cloud center and decreases outwards. While relative humidity (RH) and cloud liquid water (q) decrease smoothly to match the environment, the vertical velocity, virtual potential temperature (θ) and potential temperature (θ) perturbations have more complicated behaviour towards the cloud boundary. Below the inversion layer, θv becomes negative before the vertical velocity has turned from updraft to subsiding shell outside the cloud, indicating the presence of a transition zone where the updraft is negatively buoyant. Due to the downdraft outside the cloud and the enhanced horizontal turbulent mixing across the edge, the normalized turbulence kinetic energy (TKE) and horizontal turbulence kinetic energy (HTKE) decrease more slowly from the cloud center outwards than the thermodynamic variables. The distributions all present asymmetric structures in response to the vertical wind shear, with more negatively buoyant air, stronger downdrafts and larger TKE on the downshear side. We discuss several implications of the distributions for theoretical models and parameterizations. Positive buoyancy near cloud base is mostly due to the virtual effect of water vapor, emphasising the role of moisture in triggering. The mean vertical velocity is found to be approximately half the maximum vertical velocity within each cloud, providing a constraint on some models. Finally, products of normalized distributions for different variables are shown to be able to well represent the vertical heat and moisture fluxes, but they underestimate fluxes in the inversion layer because they do not capture cloud top downdrafts.

How to cite: Holloway, C., Gu, J.-F., Plant, B., and Jones, T.: Composited structure of shallow cumulus clouds, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-3180,, 2021.

Jian-Feng Gu, Robert Plant, Christopher Holloway, and Mark Muetzelfeldt

This study takes the first step to bridge the gap between the pressure drag of a shallow cloud ensemble and that of an individual cloud composed of rising thermals. It is found that the pressure drag for a cloud ensemble is primarily controlled by the dynamical component. The dominance of dynamical pressure drag and its increased magnitude with height are independent of cloud lifetime and are common features of individual clouds except that the total drag of a single cloud over life cycle presents vertical oscillations. These oscillations are associated with successive rising thermals but are further complicated by the evaporation-driven downdrafts outside the cloud. The horizontal vorticity associated with the vortical structure is amplified as the thermals rise to higher altitudes due to continuous baroclinic vorticity generation. This leads to the increased magnitude of local minima of dynamical pressure perturbation with height and consequently to increased dynamical pressure drag.

How to cite: Gu, J.-F., Plant, R., Holloway, C., and Muetzelfeldt, M.: Pressure drag for shallow cumulus clouds: from thermals to the cloud ensemble, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-1678,, 2021.

Maike Ahlgrimm, Daniel Klocke, Alberto de Lozar, Ekaterina Machulskaya, Mirjana Sakradzija, and Axel Seifert

The Icosahedral Model (ICON) of the German Weather Service (Deutscher Wetterdienst, DWD) is used for numerical weather prediction at global and regional scales. In the limited area mode, resolution is typically on the order of a few kilometers horizontal grid spacing. Deep convective transport is partially resolved at these scales, but shallow convection remains poorly represented without a parameterization.

A stochastic shallow convection scheme was developed in collaboration with the Max Planck Institute for Meteorology, and is now being implemented in ICON with a view towards operational use. The scheme is scale-adaptive and renders resolution-dependent tuning of the convection parameterization unnecessary. Mass flux limiters essential for the stable operation of the unaltered convection scheme can be removed when the stochastic perturbations are introduced.

Alongside the original, explicit stochastic scheme an approximation using stochastic differential equations (SDE) has been developed. The advantage of the SDE version is a lower computational and memory cost, and the ability to save and restart the model‘s stochastic cloud state easily.

Equivalence of the two versions can be demonstrated by running one version interactively, the other passively (“piggy-backing”). While the SDE approximation is computationally more efficient, the explicit version of the scheme can be easily extended to keep track of additional properties of the shallow cloud ensemble. For example, the convective updraft core fraction can be calculated for use in the diagnostic subgrid cloud scheme. Or knowledge of individual clouds’ depth can be used to derive a more realistic lateral detrainment profile than is currently in use.

We demonstrate the performance of the scheme and illustrate options and applications in single column mode, case studies and month-long hindcasts.

How to cite: Ahlgrimm, M., Klocke, D., de Lozar, A., Machulskaya, E., Sakradzija, M., and Seifert, A.: Options and extensions for the stochastic shallow convection scheme in ICON, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-11431,, 2021.

Roel Neggers and Philipp Griewank

Understanding the coupling between convective clouds and the general circulation, as well as addressing the grey zone problem in convective parameterization, requires insight into the genesis and maintenance of spatial patterns in cumulus cloud populations. In this study a simple toy model for recreating populations of interacting convective objects as distributed over a two-dimensional Eulerian grid is formulated to this purpose. Key elements at the foundation of the model include i) a fully discrete formulation for capturing discrete behavior in convective properties at small population sample sizes, ii) object age-dependence for representing life-cycle effects, and iii) a prognostic number budget allowing for object interactions and co-existence of multiple species. A primary goal is to optimize the computational efficiency of this system. To this purpose the object birth rate is represented stochastically through a spatially-aware Bernoulli process. The same binomial stochastic operator is applied to horizontal advection of objects, conserving discreteness in object number. The applicability to atmospheric convection as well as behavior implied by the formulation is assessed. Various simple applications of the BiOMi model (Binomial Objects on Microgrids) are explored, suggesting that important convective behavior can be captured at low computational cost. This includes i) subsampling effects and associated powerlaw scaling in the convective grey zone, ii) stochastic predator-prey behavior, iii) the down-scale turbulent energy cascade, and iv) simple forms of spatial organization and convective memory. Consequences and opportunities for convective parameterization in next-generation weather and climate models are discussed.

How to cite: Neggers, R. and Griewank, P.: BiOMi: A binomial stochastic framework on microgrids for efficiently modeling discrete statistics of convective populations, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-1506,, 2021.

Shallow convection discussion