Deep Convection Parameterization Research Articles

Autocorrelation—A Simple Diagnostic for Tropical Precipitation Variability in Global Kilometer‐Scale Climate Models

AbstractWe propose the lag‐1 autocorrelation of daily precipitation as a simple diagnostic of tropical precipitation variability in climate models. This metric generally has a relatively uniform distribution of positive values across the tropics. However, selected land regions are characterized by exceptionally low autocorrelation values. Low values correspond to the dominance of high frequency variance in precipitation, and specifically of high frequency convectively coupled equatorial waves. Consistent with previous work, we show that CMIP6 climate models overestimate the autocorrelation. Global kilometer‐scale models capture the observed autocorrelation when deep convection is explicitly simulated. When a deep convection parameterization is used, though, the autocorrelation increases over land and ocean, suggesting that land surface‐atmosphere interactions are not responsible for the changes in autocorrelation. Furthermore, the metric also tracks the accuracy of the representation of the relative importance of high frequency and low frequency convectively coupled equatorial waves in the models.

Improving Aerosol Radiative Forcing and Climate in E3SM: Impacts of New Cloud Microphysics and Improved Wet Removal Treatments

AbstractNumerous Earth system models exhibit excessive aerosol effective forcing at the top of the atmosphere (TOA), including the Department of Energy's Energy Exascale Earth System Model (E3SM). Here, in the context of the E3SM version 3 effort, the predicted particle property (P3) stratiform cloud microphysics scheme and an enhanced deep convection parameterization suite (ZM_plus) are implemented into E3SM. The ZM_plus includes a convective cloud microphysics scheme, a multi‐scale coherent structure parameterization for mesoscale convective systems, and a revised cloud base mass flux formulation considering impacts of the large‐scale environment. The P3 scheme improved cloud and radiation particularly over the Northern Hemisphere and the frequency of heavy precipitation over the tropics, and the ZM_plus improved clouds in the tropics. P3 decreases aerosol effective forcing by 0.15 W m−2, while the ZM_plus increases it by 0.27 W m−2, resulting from excessive direct (0.31 W m−2) and indirect forcing (−1.79 W m−2). The excessive aerosol forcings are due to aerosol overestimation associated with insufficient aerosol wet removal. By improving the physical treatments in the aerosol wet removal, we effectively mitigate anthropogenic aerosol overestimation and thus attenuate direct (0.09 W m−2) and indirect aerosol forcing (−1.52 W m−2). Adjustment to primary organic matter hygroscopicity reduces direct and indirect forcing to more reasonable values: −0.13 W m−2 and −1.31 W m−2, respectively. On climatology, improved aerosol treatments mitigate overestimation of aerosol optical depth.

Bimodality in simulated precipitation frequency distributions and its relationship with convective parameterizations

Bimodality in precipitation frequency distributions is often evident in atmospheric models, but rarely in observations. This study i) proposes a metric to objectively quantify the bimodality in precipitation distributions, ii) evaluates model simulations contributed to the Coupled Model Intercomparison Project (CMIP) phase 5 (CMIP5), phase 6 (CMIP6), and the DYnamics of the Atmospheric general circulation Modeled On Non-hydrostatic Domains (DYAMOND) project by comparing them to satellite-based and reanalysis precipitation products, and iii) investigates possible origins of bimodal precipitation distributions. Our results reveal that about 83% (20 out of 24) of CMIP5 and 70% (21 out of 30) of CMIP6 models used in this study exhibit bimodal distributions. The few DYAMOND models that use a deep convective parameterization also show bimodal distributions, while most DYAMOND models do not. Predictably, the bimodality originates from the separation of precipitation process between resolved grid-scale and parameterized subgrid-scale. However, in a larger number of models bimodality arises from the parameterized subgrid-scale convective precipitation alone.

Inclusion of the radiative effect of deep convective clouds in the Eta model simulations

AbstractConvective clouds play an important role in the local energy budget by directly interacting with solar and terrestrial radiation. However, radiation parameterization schemes of atmospheric models generally consider clouds produced from microphysics schemes or some other grid saturation criteria. Deep convective parameterization schemes tend to rain out the convective cloud before the radiation scheme perceives its water load. This may be a source of the positive bias of the incoming solar radiation at the surface. The objective of this work is to include the effects of deep convective clouds in the radiation scheme of the regional Eta model and to evaluate the impacts on the net radiative energy and other meteorological variables. The radiation scheme is the Rapid Radiative Transfer Model. The work is developed in four stages. The positive bias in the incoming solar radiation was diagnosed in the first stage. In the second stage, the parameters of the convective parameterization scheme were modified to increase convective precipitation. In the third stage, the parameters of the microphysics scheme were modified to increase explicit clouds. In the fourth and last stage, in addition to the previous modifications, the condensates from the convective parameterization were input into the radiation scheme. The runs were performed for a period of one summer rainy month with intense convective activity over South America. Including deep convective cloud condensates into the radiation scheme improved the cloud cover, the diurnal cycle of the surface net radiation, and the 2‐m temperature. However, the reduction of the net radiation at the surface caused the reduction of the available energy for convective instability and, consequently, the precipitation reduction. The results show the importance of including cumulus cloud water load in the radiative scheme for bias reduction in the radiative energy components.

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

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

Cloud Phase Simulation at High Latitudes in EAMv2: Evaluation Using CALIPSO Observations and Comparison With EAMv1

AbstractThis study performs a comprehensive evaluation of the simulated cloud phase in the U.S. Department of Energy, Energy Exascale Earth System Model, atmosphere model version 2 (EAMv2), and version 1 (EAMv1). Enabled by the CALIPSO (Cloud‐Aerosol Lidar and Infrared Pathfinder Satellite Observation) simulator, EAMv2 and EAMv1 predicted cloud phase is compared against the GCM‐Oriented CALIPSO Cloud Product (CALIPSO‐GOCCP) at high latitudes where mixed‐phase clouds are prevalent. Our results indicate that the underestimation of cloud ice in simulated high‐latitude mixed‐phase clouds in EAMv1 has been significantly reduced in EAMv2. The increased ice clouds in the Arctic mainly result from the modification on the WBF (Wegner‐Bergeron‐Findeisen) process in EAMv2. The impact of the modified WBF process is moderately compensated by the low limit of cloud droplet number concentration in cloud microphysics and the new dCAPE_ULL trigger used in deep convection in EAMv2. Moreover, it is found that the new trigger largely contributes to the better cloud phase simulation over the Norwegian Sea and Barents Sea in the Arctic and the Southern Ocean where large errors are found in EAMv1. However, errors in simulated cloud phase in EAMv1, such as the overestimation of supercooled liquid clouds near the surface in both hemispheres and the underestimation of ice clouds over Antarctica, persist in EAMv2. This study highlights the impact of deep convection parameterizations, which has received little attention, on high‐latitude mixed‐phase clouds, and the importance of continuous improvement of cloud microphysics in climate models for accurately representing mixed‐phase clouds.

Using Radiative Convective Equilibrium to Explore Clouds and Climate in the Community Atmosphere Model

AbstractCharacteristics of, and fundamental differences between, the radiative‐convective equilibrium (RCE) climate states following the Radiative‐Convective Equilibrium Model Intercomparison Project (RCEMIP) protocols in the Community Atmosphere Model version 5 (CAM5) and version 6 (CAM6) are presented. This paper explores the characteristics of clouds, moisture, precipitation and circulation in the RCE state, as well as the tropical response to surface warming, in CAM5 and CAM6 with different parameterizations. Overall, CAM5 simulates higher precipitation rates that result in larger global average precipitation, despite lower outgoing longwave radiation compared to CAM6. Differences in the structure of clouds, particularly the amount and vertical location of cloud liquid, exist between the CAM versions and can, in part, be related to distinct representations of shallow convection and boundary layer processes. Both CAM5 and CAM6 simulate similar peaks in cloud fraction, relative humidity, and cloud ice, linked to the usage of a similar deep convection parameterization. These anvil clouds rise and decrease in extent in response to surface warming. More generally, extreme precipitation, aggregation of convection, and climate sensitivity increase with warming in both CAM5 and CAM6. This analysis provides a benchmark for future studies that explore clouds, convection, and climate in CAM with the RCEMIP protocols now available in the Community Earth System Model. These results are discussed within the context of realistic climate simulations using CAM5 and CAM6, highlighting the usefulness of a hierarchical modeling approach to understanding model and parameterization sensitivities to inform model development efforts.

The interaction of deep convection with the general circulation in Titan's atmosphere. Part 1: Cloud resolving simulations

The deep convective cloud–environment feedback loop is likely important to Titan's global methane, energy, and momentum cycles, just as it is for Earth's global water, energy, and momentum budgets. General circulation models of Titan's atmosphere are unable to explicitly simulate deep convection and must instead parameterize the impact of this important subgrid-scale phenomenon on the model-resolved atmospheric state. The goal of this study is to better quantify through cloud resolving modeling the effects of deep convective methane storms on their environment and to feed that information forward to improve parameterizations in global models. Dozens of atmospheric profiles unstable with respect to deep moist convection are extracted from the global Titan Atmospheric Model (TAM) and used to initialize the cloud-resolving Titan Regional Atmospheric Modeling System (TRAMS). Mean profiles of heating/cooling and moistening/drying of the large-scale environment in TRAMS indicate that Titan's deep convection forces the environment in a manner analogous to Earth: Large-scale subsidence of the environmental air warms and dries the environment, but clouds can also moisten the environment through the detrainment and evaporation of condensate near cloud top. Relative humidity profiles and characteristic convective time scales are derived to guide the tuning of the deep convective parameterization implemented in TAM, as described in a companion paper. The triggering of convection, the dry convective mixing of the planetary boundary layer, and the entrainment of environmental air into rising air parcels are found to be critical to determining whether a deep convective cloud will form. Only profiles with relatively large convective available potential energy (CAPE) and well mixed planetary boundary layers with high relative humidity were found to produce storms. Environments with low-level thermal inversions and planetary boundary layers with low relative humidity or rapidly decreasing moisture with height failed to generate deep convection in TRAMS despite positive CAPE.

Evaluation and Comparison of Two Deep Convection Parameterization Schemes at Convection-Permitting Resolution

AbstractCoarse-grained results from a large-eddy simulation (LES) using the Weather Research and Forecasting (WRF) Model were compared in this study with the WRF simulations at a typical convection-permitting horizontal grid spacing of 3 km for an idealized case of deep moist convection. The purpose of this comparison is to identify major differences at the subgrid process level between two widely used deep convection parameterization schemes in the WRF Model. It is shown that there are considerable differences in subgrid process representations between the two schemes due to different parameterization formulations and underlying assumptions. The two schemes not only differ in trigger function, subgrid cloud model, and closure assumptions but also disagree with the coarse-grained LES results in terms of vertical mass flux profiles. Thus, it is difficult to discern which scheme is more advantageous over the other at the subgrid process level. The conclusions from this study highlight the importance of establishing benchmarks using observations and LES to develop and evaluate convection parameterization schemes suitable for models at convection-permitting resolution.

Evaluation framework for sub-daily rainfall extremes simulated by regional climate models

AbstractSub-daily precipitation extremes are high-impact events that can result in flash floods, sewer system overload, or landslides. Several studies have reported an intensification of projected short-duration extreme rainfall in a warmer future climate. Traditionally, regional climate models (RCMs) are run at a coarse resolution using deep-convection parameterization for these extreme events. As computational resources are continuously ramping up, these models are run at convection-permitting resolution, thereby partly resolving the small-scale precipitation events explicitly. To date, a comprehensive evaluation of convection-permitting models is still missing. We propose an evaluation strategy for simulated sub-daily rainfall extremes that summarizes the overall RCM performance. More specifically, the following metrics are addressed: the seasonal/diurnal cycle, temperature and humidity dependency, temporal scaling and spatio-temporal clustering. The aim of this paper is: (i) to provide a statistical modeling framework for some of the metrics, based on extreme value analysis, (ii) to apply the evaluation metrics to a micro-ensemble of convection-permitting RCM simulations over Belgium, against high-frequency observations, and (iii) to investigate the added value of convection-permitting scales with respect to coarser 12-km resolution. We find that convection-permitting models improved precipitation extremes on shorter time scales (i.e, hourly or two-hourly), but not on 6h-24h time scales. Some metrics such as the diurnal cycle or the Clausius-Clapeyron rate are improved by convection-permitting models, whereas the seasonal cycle appears robust across spatial scales. On the other hand, the spatial dependence is poorly represented at both convection-permitting scales and coarser scales. Our framework provides perspectives for improving high-resolution atmospheric numerical modeling and datasets for hydrological applications.

Convection‐permitting modeling with regional climate models: Latest developments and next steps

AbstractApproximately 10 years ago, convection‐permitting regional climate models (CPRCMs) emerged as a promising computationally affordable tool to produce fine resolution (1–4 km) decadal‐long climate simulations with explicitly resolved deep convection. This explicit representation is expected to reduce climate projection uncertainty related to deep convection parameterizations found in most climate models. A recent surge in CPRCM decadal simulations over larger domains, sometimes covering continents, has led to important insights into CPRCM advantages and limitations. Furthermore, new observational gridded datasets with fine spatial and temporal (~1 km; ~1 h) resolutions have leveraged additional knowledge through evaluations of the added value of CPRCMs. With an improved coordination in the frame of ongoing international initiatives, the production of ensembles of CPRCM simulations is expected to provide more robust climate projections and a better identification of their associated uncertainties. This review paper presents an overview of the methodology to produce CPRCM simulations and the latest research on the related added value in current and future climates. Impact studies that are already taking advantage of these new CPRCM simulations are highlighted. This review paper ends by proposing next steps that could be accomplished to continue exploiting the full potential of CPRCMs.This article is categorized under: Climate Models and Modeling > Earth System Models

Model intercomparison of COSMO 5.0 and IFS 45r1 at kilometer-scale grid spacing

Abstract. The increase in computing power and recent model developments allow for the use of global kilometer-scale weather and climate models for routine forecasts. At these scales, deep convective processes can be partially resolved explicitly by the model dynamics. Next to horizontal resolution, other aspects such as the applied numerical methods, the use of the hydrostatic approximation, and time step size are factors that might influence a model's ability to resolve deep convective processes. In order to improve our understanding of the role of these factors, a model intercomparison between the nonhydrostatic COSMO model and the hydrostatic Integrated Forecast System (IFS) from ECMWF has been conducted. Both models have been run with different spatial and temporal resolutions in order to simulate 2 summer days over Europe with strong convection. The results are analyzed with a focus on vertical wind speed and precipitation. Results show that even at around 3 km horizontal grid spacing the effect of the hydrostatic approximation seems to be negligible. However, time step proves to be an important factor for deep convective processes, with a reduced time step generally allowing for higher updraft velocities and thus more energy in vertical velocity spectra, in particular for shorter wavelengths. A shorter time step is also causing an earlier onset and peak of the diurnal cycle. Furthermore, the amount of horizontal diffusion plays a crucial role for deep convection with more diffusion generally leading to larger convective cells and higher precipitation intensities. The study also shows that for both models the parameterization of deep convection leads to lower updraft and precipitation intensities and biases in the diurnal cycle with a precipitation peak which is too early.

Impact of Stochastic Entrainment in the NCAR CAM Deep Convection Parameterization on the Simulation of South Asian Summer Monsoon

Model simulations are highly sensitive to the formulation of the atmospheric mixing process or entrainment in the deep convective parameterizations used in their atmospheric component. In this paper, we have implemented stochastic entrainment in the deep convection scheme of NCAR CAM5 and analyzed the improvements in model simulation, focusing on the South Asian Summer Monsoon (SASM), as compared to the deterministic entrainment formulation in the default version of the model. Simulations using stochastic entrainment (StochCAM5) outperformed default model simulations (DefCAM5), as inferred from multiple metrics associated with the SASM. StochCAM5 significantly alleviated some of the longstanding SASM biases seen in DefCAM5, such as precipitation pattern and magnitude over the Arabian Sea and western Equatorial Indian ocean, early monsoon withdrawal, and the overestimation in the frequency of light precipitation and the underestimation in the frequency of large-to-extreme precipitation. Related SASM dynamical and thermodynamical features, such as Somali Jet, low-level westerly winds, and meridional tropospheric temperature gradient (MTTG), are improved in StochCAM5. Further, the simulation of monsoon intra-seasonal oscillation (MISO), Madden Julian Oscillation (MJO), and equatorial Kelvin waves are improved in StochCAM5. Many essential climate variables, such as shortwave and longwave cloud forcing, cloud cover, relative and specific humidity, and precipitable water, show significant improvement in StochCAM5.

Improving Convection Trigger Functions in Deep Convective Parameterization Schemes Using Machine Learning

AbstractDeficiencies in convection trigger functions, used in deep convection parameterizations in General Circulation Models (GCMs), have critical impacts on climate simulations. A novel convection trigger function is developed using the machine learning (ML) classification model XGBoost. The large‐scale environmental information associated with convective events is obtained from the long‐term constrained variational analysis forcing data from the Atmospheric Radiation Measurement (ARM) program at its Southern Great Plains (SGP) and Manaus (MAO) sites representing, respectively, continental mid‐latitude and tropical convection. The ML trigger is separately trained and evaluated per site, and jointly trained and evaluated at both sites as a unified trigger. The performance of the ML trigger is compared with four convective trigger functions commonly used in GCMs: dilute convective available potential energy (CAPE), undilute CAPE, dilute dynamic CAPE (dCAPE), and undilute dCAPE. The ML trigger substantially outperforms the four CAPE‐based triggers in terms of the F1 score metric, widely used to estimate the performance of ML methods. The site‐specific ML trigger functions can achieve, respectively, 91% and 93% F1 scores at SGP and MAO. The unified trigger also has a 91% F1 score, with virtually no degradation from the site‐specific training, suggesting the potential of a global ML trigger function. The ML trigger alleviates a GCM deficiency regarding the overprediction of convection occurrence, offering a promising improvement to the simulation of the diurnal cycle of precipitation. Furthermore, to overcome the black box issue of the ML methods, insights derived from the ML model are discussed, which may be leveraged to improve traditional CAPE‐based triggers.

Effects of coupling a stochastic convective parameterization with the Zhang–McFarlane scheme on precipitation simulation in the DOE E3SMv1.0 atmosphere model

Abstract. A stochastic deep convection parameterization is implemented into the US Department of Energy (DOE) Energy Exascale Earth System Model (E3SM) Atmosphere Model version 1.0 (EAMv1). This study evaluates its performance in simulating precipitation. Compared to the default model, the probability distribution function (PDF) of rainfall intensity in the new simulation is greatly improved. The well-known problem of “too much light rain and too little heavy rain” is alleviated, especially over the tropics. As a result, the contribution from different rain rates to the total precipitation amount is shifted toward heavier rain. The less frequent occurrence of convection contributes to suppressed light rain, while more intense large-scale and convective precipitation contributes to enhanced heavy total rain. The synoptic and intraseasonal variabilities of precipitation are enhanced as well to be closer to observations. The sensitivity of the rainfall intensity PDF to the model vertical resolution is examined. The relationship between precipitation and dilute convective available potential energy in the stochastic simulation agrees better with that in the Atmospheric Radiation Measurement (ARM) observations compared with the standard model simulation. The annual mean precipitation is largely unchanged with the use of the stochastic scheme except over the tropical western Pacific, where a moderate increase in precipitation represents a slight improvement. The responses of precipitation and its extremes to climate warming are similar with or without the stochastic deep convection scheme.

Improving the Physical Basis for Updraft Dynamics in Deep Convection Parameterizations

AbstractThis article presents a new deep convective parameterization that determines cloud characteristics based on a specified cloud size distribution. The vertical profiles of cloud properties are determined by analytic equations, which formulate entrainment with an inverse relationship to cloud width. In line with recent studies of large eddy simulations (LES), cloud widths are assumed to be constant with height and vertical mass flux (M) characteristics of the clouds are therefore regulated by the vertical velocity profile. The parameterization is configured to work with existing cloud baseMclosure formulations, with the closure predicting the total cloud area rather than the cloud baseMdirectly. Analytic formula are also used to connect the vertical wind shear magnitude to the cloud size distribution, wherein larger shear magnitudes result in more numerous large updrafts than weaker shear magnitudes, which is in line with recent research results. The parameterization is compared against 10 deep convective LES with varying thermodynamic and vertical wind shear profiles. Results show dramatic improvements in the prediction of normalizedM, detrainment, and the properties of detrained air over the existing Zhang and McFarlane (1995) scheme. In particular, the new model is able to correctly portray the transition from a bottom‐heavyMprofile in weakly sheared environments, to a top‐heavyMprofile in strongly sheared environments.

Adaptation of the Predicted Particles Properties (P3) Microphysics Scheme for Large-Scale Numerical Weather Prediction

AbstractA parameterization for the subgrid-scale cloud and precipitation fractions has been incorporated into the Predicted Particle Properties (P3) microphysics scheme for use in atmospheric models with relatively coarse horizontal resolution. The modified scheme was tested in a simple 1D kinematic model and in the Canadian Global Environmental Multiscale (GEM) model using an operational global NWP configuration with a 25-km grid spacing. A series of 5-day forecast simulations was run using P3 and the much simpler operational Sundqvist condensation scheme as a benchmark for comparison. The effects of using P3 in a global GEM configuration, with and without the modifications, were explored through statistical metrics of common forecast fields against upper-air and surface observations. Diagnostics of state variable tendencies from various physics parameterizations were examined to identify possible sources of errors resulting from the use of the modified scheme. Sensitivity tests were performed on the coupling between the deep convection parameterization scheme and the microphysics, specifically regarding assumptions in the physical properties of detrained ice. It was found that even without recalibration of the suite of moist physical parameterizations, substituting the Sundqvist condensation scheme with the modified P3 microphysics resulted in some significant improvements to the temperature and geopotential height bias throughout the troposphere and out to day 5, but with degradation to error standard deviation toward the end of the integrations, as well as an increase in the positive bias of precipitation quantities. The modified P3 scheme was thus shown to hold promise for potential use in coarse-resolution NWP systems.

Cascading Toward a Kilometer‐Scale GCM: Impacts of a Scale‐Aware Convection Parameterization in the Goddard Earth Observing System GCM

AbstractThe National Aeronautics and Space Administration (NASA) Goddard Earth Observing System global circulation model (GCM) is evaluated through a cascade of simulations with increasing horizontal resolution. This model employs a nonhydrostatic dynamical core and includes a scale‐aware, deep convection parameterization (DPCP). The 40‐day simulations at six resolutions (100 km to 3 km) with unvarying model formulation were produced. At the highest resolution, extreme experiments were carried out: one with no DPCP and one with its scale awareness eliminated. Simulated precipitation, radiative balance, and atmospheric thermodynamic and dynamical variables are well reproduced with respect to both observational and reanalysis data. As model resolution increases, the convective precipitation smoothly transitions from being mostly produced by the convection parameterization to the cloud microphysics parameterization. However, contrary to current thought, these extreme cases argue for maintaining, to some extent, the scale‐aware DPCP even at 3‐km scale, as the run relying solely on explicit grid‐scale production of rainfall performs more poorly at this resolution.

Interactions between Moisture and Tropical Convection. Part I: The Coevolution of Moisture and Convection

AbstractRealistically representing the multiscale interactions between moisture and tropical convection remains an ongoing challenge for weather prediction and climate models. In this study, we revisit the relationship between precipitation and column saturation fraction (CSF) by investigating their tendencies in CSF–precipitation space using satellite and radar observations, as well as reanalysis. A well-known, roughly exponential increase in precipitation occurs as CSF increases above a “critical point,” which acts as an attractor in CSF–precipitation space. Each movement away from and subsequent return toward the attractor results in a small net change of the coupled system, causing it to evolve in a cyclical fashion around the attractor. This cyclical evolution is characterized by shallow and convective precipitation progressively moistening the environment and strengthening convection, stratiform precipitation progressively weakening convection, and drying in the nonprecipitating and lightly precipitation regime. This behavior is evident across a range of spatiotemporal scales, suggesting that shortcomings in model representation of the joint evolution of convection and large-scale moisture will negatively impact a broad range of spatiotemporal scales. Novel process-level diagnostics indicate that several models, all implementing versions of the Zhang–McFarlane deep convective parameterization, exhibit unrealistic coupling between column moisture and convection.

Simulating model uncertainty of subgrid-scale processes by sampling model errors at convective scales

Abstract. Ideally, perturbation schemes in ensemble forecasts should be based on the statistical properties of the model errors. Often, however, the statistical properties of these model errors are unknown. In practice, the perturbations are pragmatically modelled and tuned to maximize the skill of the ensemble forecast. In this paper a general methodology is developed to diagnose the model error, linked to a specific physical process, based on a comparison between a target and a reference model. Here, the reference model is a configuration of the ALADIN (Aire Limitée Adaptation Dynamique Développement International) model with a parameterization of deep convection. This configuration is also run with the deep-convection parameterization scheme switched off, degrading the forecast skill. The model error is then defined as the difference of the energy and mass fluxes between the reference model with scale-aware deep-convection parameterization and the target model without deep-convection parameterization. In the second part of the paper, the diagnosed model-error characteristics are used to stochastically perturb the fluxes of the target model by sampling the model errors from a training period in such a way that the distribution and the vertical and multivariate correlation within a grid column are preserved. By perturbing the fluxes it is guaranteed that the total mass, heat and momentum are conserved. The tests, performed over the period 11–20 April 2009, show that the ensemble system with the stochastic flux perturbations combined with the initial condition perturbations not only outperforms the target ensemble, where deep convection is not parameterized, but for many variables it even performs better than the reference ensemble (with scale-aware deep-convection scheme). The introduction of the stochastic flux perturbations reduces the small-scale erroneous spread while increasing the overall spread, leading to a more skillful ensemble. The impact is largest in the upper troposphere with substantial improvements compared to other state-of-the-art stochastic perturbation schemes. At lower levels the improvements are smaller or neutral, except for temperature where the forecast skill is degraded.