First-order Closure Scheme Research Articles

Improvement of the vertical mixing in chemistry transport modeling based on a 1.5-order turbulence kinetic energy-based eddy diffusivity closure scheme

In this study, firstly, three first-order local closure vertical diffusion schemes used in multiple mesoscale chemical transport models (CTM) were embedded in the CHIMERE CTM model and tested over a one-year simulation covering the whole France. Three model configurations present a fair reproduction of pollutant concentrations both in urban and rural areas, indicating it is an effective way to reproduce the dispersion of pollutants in chemistry transport modeling. However, it cannot be expected to significantly improve the vertical mixing under the first-order closure scheme. Secondly, a 1.5-order turbulence kinetic energy-based eddy diffusivity closure scheme called as the new eddy diffusion (NED) is implemented in CHIMERE to describe more realistic diffusion processes near the surface. A fifteen-day simulation encompassing a winter pollution episode was performed for three major cities with horizontal resolution of 1.67 km and the first layer height at 12 m, respectively. The NED scheme improved NO2 simulations at most urban sites compared to the initial Kz diffusion scheme (IKD). Taking the root mean square error as evaluation criteria, the average improvements are 18.8%, 24.5% and 9.5% for NO2 simualtion in Paris, Lyon and Bordeaux respectively. For the model performance of PM2.5 and PM10 simulations in the urban areas of Paris, the improvements are 13.5% and 19.1%, respectively. Overall, preliminary outcomes of this study are encouraging. The simulation with more sophisticated and realistic eddy viscosities are better than for IKD that is widely used in CTMs, but we need to realize that this is only a fifteen-day simulation for three cities, for further research, longer periods are needed with a greater variety of meteorological situations to prove the universality of the NED scheme.

Uncertainty quantification of turbulent systems via physically consistent and data-informed reduced-order models

This work presents a data-driven, energy-conserving closure method for the coarse-scale evolution of the mean and covariance of turbulent systems. Spatiotemporally non-local neural networks are employed for calculating the impact of non-Gaussian effects to the low-order statistics of dynamical systems with an energy-preserving quadratic nonlinearity. This property, which characterizes the advection term of turbulent flows, is encoded via an appropriate physical constraint in the training process of the data-informed closure. This condition is essential for the stability and accuracy of the simulations as it appropriately captures the energy transfers between unstable and stable modes of the system. The numerical scheme is implemented for a variety of turbulent systems, with prominent forward and inverse energy cascades. These problems include prototypical models such as an unstable triad system and the Lorentz-96 system, as well as more complex models: The two-layer quasi-geostrophic flows and incompressible, anisotropic jets where passive inertial tracers are being advected on. Training data are obtained through high-fidelity direct numerical simulations. In all cases, the hybrid scheme displays its ability to accurately capture the energy spectrum and high-order statistics of the systems under discussion. The generalizability properties of the trained closure models in all the test cases are explored, using out-of-sample realizations of the systems. The presented method is compared with existing first-order closure schemes, where only the mean equation is evolved. This comparison showcases that correctly evolving the covariance of the system outperforms first-order schemes in accuracy, at the expense of increased computational cost.

Role of Planetary Boundary Layer Processes in the Simulation of Tropical Cyclones Over the Bay of Bengal

The behaviour of planetary boundary layer (PBL) schemes initialized at different life stages of a tropical cyclone (TC) is studied by considering seven Bay of Bengal TC cases. In each TC case, the Advanced Research Weather Research and Forecasting (WRF-ARW) model is initialized at four life stages (depression to very severe cyclone storm) with National Center for Environmental Prediction (NCEP) Global analysis and integrated up to 96 h. A set of six PBL sensitivity experiments are conducted at four stages for all seven TC cases to analyse the impact of the model boundary layer in simulating the TC track and intensity parameters. The model-produced track, intensity and rainfall patterns are evaluated with the best track, intensity and gridded rainfall estimates obtained from the India Meteorological Department (IMD). The spatial and radius/height section simulated fields are evaluated with satellite retrievals. Results depict that the six PBL schemes during model initialization at different stages of a TC have produced sizable differences in the simulation of track and intensity parameters. The local and nonlocal schemes produced different results based on the TC stage at which the model is initialized. The results also suggest that if the model is initialized with a non-organized cyclonic vortex such as depression stage of the storm, PBL schemes exhibit high sensitivity and spread in terms of both track and intensity. While the spread between PBL schemes was significantly reduced and found close to the observed estimates when the model was initialized at the advanced stages of the TC. In addition, the local 1.5-order closure scheme simulated the storm parameters relatively better when the cyclone vortex was not well organized in the model's initial conditions, while the non-local and first-order closure schemes perform better with initial model conditions of a well-defined cyclonic vortex.

Performance Analysis of Planetary Boundary Layer Parameterization Schemes in WRF Modeling Set Up over Southern Italy

Predictions of boundary layer meteorological parameters with accuracy are essential for achieving good weather and air quality regional forecast. In the present work, we have analyzed seven planetary boundary layer (PBL) parameterization schemes in a Weather Research and Forecasting (WRF) model over the Naples-Caserta region of Southern Italy. WRF model simulations were performed with 1-km horizontal resolution, and the results were compared against data collected by the small aircraft Sky Arrow Environmental Research Aircraft (ERA) during 7–9 October 2014. The selected PBL schemes include three first-order closure PBL schemes (ACM2, MRF, YSU) and four turbulent kinetic energy (TKE) closure schemes (MYJ, UW, MYNN2, and BouLac). A performance analysis of these PBL schemes has been investigated by validating them with aircraft measurements of meteorological parameters profiles (air temperature, specific humidity, wind speed, wind direction) and PBL height to assess their efficiency in terms of the reproduction of observed weather conditions. Results suggested that the TKE closure schemes perform better than first-order closure schemes, and the MYNN2 closure scheme is close to observed values most of the time. It is observed that the inland locations are better simulated than sea locations, and the morning periods are better simulated than those in the afternoon. The results are emphasizing that meteorology-induced variability is larger than the variability in PBL schemes.

Two approaches to describing the turbulent exchange within the atmospheric surface layer

Two two-dimensional models that describe the interaction of an air flow with a heterogeneous vegetation cover within the atmospheric surface layer are presented. These models differ in the closure techniques used to obtain approximate solutions of the averaged system of the Navier−Stokes and continuity equations and also in the approaches applied to describe the interaction of the air flow with elements of vegetation. The first model is based on the one-and-a-half-order closure scheme and the interaction of an air flow with elements of vegetation is described using the viscous drag force. The second model uses the first-order closure scheme and the distribution of the wind velocity field near the boundaries between different vegetation communities is simulated by the theory of contrast structures (CSs). We also present the results of the numerical experiments to compare the developed models with respect to their ability to describe the wind-velocity field within the atmospheric surface layer above the different forms of vegetation cover heterogeneity, such as a forest belt and clear-cut areas.

Evaluations on Profiles of the Eddy Diffusion Coefficients through Simulations of Super Typhoons in the Northwestern Pacific

The modeling of the eddy diffusion coefficients (also known as eddy diffusivity) in the first-order turbulence closure schemes is important for the typhoon simulations, since the coefficients control the magnitude of the sensible heat flux and the latent heat flux, which are energy sources for the typhoon intensification. Profiles of the eddy diffusion coefficients in the YSU planetary boundary layer (PBL) scheme are evaluated in the advanced research WRF (ARW) system. Three versions of the YSU scheme (original, K025, and K200) are included in this study. The simulation results are compared with the observational data from track, center sea-level pressure (CSLP), and maximum surface wind speed (MWSP). Comparing with the original version, the K200 improves the averaged mean absolute errors (MAE) of track, CSLP, and MWSP by 6.0%, 3.7%, and 23.1%, respectively, while the K025 deteriorates the averaged MAEs of track, CSLP, and MWSP by 25.1%, 19.0%, and 95.0%, respectively. Our results suggest that the enlarged eddy diffusion coefficients may be more suitable for super typhoon simulations.

Intercomparison of planetary boundary layer parameterization and its impacts on surface ozone concentration in the WRF/Chem model for a case study in Houston/Texas

With over 6 million inhabitants the Houston metropolitan area is the fourth-largest in the United States. Ozone concentration in this southeast Texas region frequently exceeds the National Ambient Air Quality Standard (NAAQS). For this reason our study employed the Weather Research and Forecasting model with Chemistry (WRF/Chem) to quantify meteorological prediction differences produced by four widely used PBL schemes and analyzed its impact on ozone predictions. The model results were compared to observational data in order to identify one superior PBL scheme better suited for the area. The four PBL schemes include two first-order closure schemes, the Yonsei University (YSU) and the Asymmetric Convective Model version 2 (ACM2); as well as two turbulent kinetic energy closure schemes, the Mellor–Yamada–Janjic (MYJ) and Quasi-Normal Scale Elimination (QNSE). Four 24 h forecasts were performed, one for each PBL scheme. Simulated vertical profiles for temperature, potential temperature, relative humidity, water vapor mixing ratio, and the u–v components of the wind were compared to measurements collected during the Second Texas Air Quality Study (TexAQS-II) Radical and Aerosol Measurements Project (TRAMP) experiment in summer 2006. Simulated ozone was compared against TRAMP data, and air quality stations from Continuous Monitoring Station (CAMS). Also, the evolutions of the PBL height and vertical mixing properties within the PBL for the four simulations were explored. Although the results yielded high correlation coefficients and small biases in almost all meteorological variables, the overall results did not indicate any preferred PBL scheme for the Houston case. However, for ozone prediction the YSU scheme showed greatest agreements with observed values.

An Analytical Model for Mean Wind Profiles in Sparse Canopies

Existing analytical models for mean wind profiles within canopies are applicable only in dense canopy scenarios, where all momentum is absorbed by canopy elements and, hence, the effect of the ground on turbulent mixing is not important. Here, we propose a new analytical model that can simulate mean wind profiles within sparse canopies under neutral conditions. The model adopts a linearized canopy-drag parametrization and a first-order turbulence closure scheme taking into account the effects of both the ground and canopy elements on turbulent mixing. The resulting wind profile within a sparser canopy appears to be more like a logarithmic form, with the no-slip condition at the ground being satisfied. The analytical solution converges exactly to the standard surface-layer logarithmic wind profile in the case of zero canopy density (i.e., no-canopy scenario) and tends to be an exponential wind profile for a dense canopy; this feature is unique compared with existing analytical models for canopy wind profiles. Results from the new model are in good agreement with those from laboratory experiments and numerical simulations.

Improved low-cloud simulation from a multiscale modeling framework with a third-order turbulence closure in its cloud-resolving model component

[1] In the original multiscale modeling framework (MMF), the Community Atmosphere Model (CAM3.5) is used as the host general circulation model (GCM), and the System for Atmospheric Modeling model with a first-order turbulence closure is used as the cloud resolving model (CRM) for representing cloud physical processes in each grid column of the GCM. This study introduces an upgrade of the MMF in which the first-order turbulence closure scheme is replaced by an advanced third-order turbulence closure in its CRM component. The results are compared between the upgraded and original MMFs, CAM3.5, and observations. The global distributions of low-level cloud amounts in the subtropics in the upgraded MMF show substantial improvement relative to the original MMF when both are compared with observations. The improved simulation of low-level clouds is attributed not only to the representation of subgrid-scale condensation in the embedded CRM but also is closely related to the increased surface sensible and latent heat fluxes, the increased lower tropospheric stability (LTS), and stronger longwave radiative cooling. Both MMF simulations show close agreement in the vertical structures of cloud amount and liquid water content of midlatitude storm-track clouds and subtropical low-level clouds, compared with observations, with the upgraded MMF being better at simulating the low-level cumulus regime. Since the upgraded MMF produces more subtropical low-level clouds and does not produce an excessive amount of optically thick high-level clouds in either the tropics or midlatitudes as the original MMF does, the global mean albedo decreases. The positive bias in albedo and longwave cloud radiative forcing (CRF) and negative bias in shortwave CRF are reduced in the tropical convective regions.

The wind-field structure in a stably stratified atmospheric boundary layer over a rough surface

Both wind turning with height and ageostrophic flow in a stably stratified atmospheric boundary layer are analyzed using a three-parameter turbulence model. For a quasi-steady state of the boundary layer, the cross-isobaric flow is determined only by turbulent stress at the surface in the direction of geostrophic wind. The “operative” prediction models, in which the first-order turbulence closure schemes are used, tend to overestimate the boundary-layer depth and underestimate the angle between the surface and geostrophic winds when compared to “research” models (schemes of high-level turbulence closure). The true value of the angle between the surface and geostrophic winds is significant for the presentation of a large-scale flow. A nocturnal low-level jet is a mesoscale phenomenon reflected in data obtained from measurements in a stably stratified atmospheric boundary layer. It is found that such jets are of great importance in transporting humidity, momentum, and air pollution. In this study, the difference between jet flows over a homogeneous underlying surface and over a spatially localized large-scale aerodynamic roughness is shown.

Intercomparison of Planetary Boundary-Layer Parametrizations in the WRF Model for a Single Day from CASES-99

This study compares five planetary boundary-layer (PBL) parametrizations in the Weather Research and Forecasting (WRF) numerical model for a single day from the Cooperative Atmosphere-Surface Exchange Study (CASES-99) field program. The five schemes include two first-order closure schemes—the Yonsei University (YSU) PBL and Asymmetric Convective Model version 2 (ACM2), and three turbulent kinetic energy (TKE) closure schemes—the Mellor–Yamada–Janjic (MYJ), quasi-normal scale elimination (QNSE), and Bougeault–Lacarrere (BouLac) PBL. The comparison results reveal that discrepancies among thermodynamic surface variables from different schemes are large at daytime, while the variables converge at nighttime with large deviations from those observed. On the other hand, wind components are more divergent at nighttime with significant biases. Regarding PBL structures, a non-local scheme with the entrainment flux proportional to the surface flux is favourable in unstable conditions. In stable conditions, the local TKE closure schemes show better performance. The sensitivity of simulated variables to surface-layer parametrizations is also investigated to assess relative contributions of the surface-layer parametrizations to typical features of each PBL scheme. In the surface layer, temperature and moisture are more strongly influenced by surface-layer formulations than by PBL mixing algorithms in both convective and stable regimes, while wind speed depends on vertical diffusion formulations in the convective regime. Regarding PBL structures, surface-layer formulations only contribute to near-surface variability and then PBL mean properties, whereas shapes of the profiles are determined by PBL mixing algorithms.

Numerical simulation of fog using non-hydrostatic NWP model with third-order turbulence closure

Numerical simulation of a fog episode is conducted using the Hong Kong Observatory’s Non-Hydrostatic Model (NHM) to study various physical processes in the formation of a shallow saturated layer. It is found that a third-order turbulence closure model and partial condensation scheme implemented in NHM can improve the parametrization of turbulent mixing and condensation processes. Physically, the formation of fog or low-level clouds over the coastal waters is primarily due to advection of warm and humid maritime air mass over a cooler sea surface. The numerical experiments show that the radiative cooling above the low-level cloud top further enhances the mixing process during nighttime and results in condensation of cloud liquid water near the surface. Differences in the forecast low-level clouds, wind and moisture fields within the boundary layer using the original first-order turbulence closure scheme will also be illustrated.

Ekman layers in the Southern Ocean: spectral models and observations, vertical viscosity and boundary layer depth

Abstract. Spectral characteristics of the oceanic boundary-layer response to wind stress forcing are assessed by comparing surface drifter observations from the Southern Ocean to a suite of idealized models that parameterize the vertical flux of horizontal momentum using a first-order turbulence closure scheme. The models vary in their representation of vertical viscosity and boundary conditions. Each is used to derive a theoretical transfer function for the spectral linear response of the ocean to wind stress. The transfer functions are evaluated using observational data. The ageostrophic component of near-surface velocity is computed by subtracting altimeter-derived geostrophic velocities from observed drifter velocities (nominally drogued to represent motions at 15-m depth). Then the transfer function is computed to link these ageostrophic velocities to observed wind stresses. The traditional Ekman model, with infinite depth and constant vertical viscosity is among the worst of the models considered in this study. The model that most successfully describes the variability in the drifter data has a shallow layer of depth O(30–50 m), in which the viscosity is constant and O(100–1000 m2 s−1), with a no-slip bottom boundary condition. The second best model has a vertical viscosity with a surface value O(200 m2 s−1), which increases linearly with depth at a rate O(0.1–1 cm s−1) and a no-slip boundary condition at the base of the boundary layer of depth O(103 m). The best model shows little latitudinal or seasonal variability, and there is no obvious link to wind stress or climatological mixed-layer depth. In contrast, in the second best model, the linear coefficient and the boundary layer depth seem to covary with wind stress. The depth of the boundary layer for this model is found to be unphysically large at some latitudes and seasons, possibly a consequence of the inability of Ekman models to remove energy from the system by other means than shear-induced dissipation. However, the Ekman depth scale appears to scale like the climatological mixed-layer depth.

Investigating a Hierarchy of Eulerian Closure Models for Scalar Transfer Inside Forested Canopies

Modelling the transfer of heat, water vapour, and CO2 between the biosphere and the atmosphere is made difficult by the complex two-way interaction between leaves and their immediate microclimate. When simulating scalar sources and sinks inside canopies on seasonal, inter-annual, or forest development time scales, the so-called well-mixed assumption (WMA) of mean concentration (i.e. vertically constant inside the canopy but dynamically evolving in time) is often employed. The WMA eliminates the need to model how vegetation alters its immediate microclimate, which necessitates formulations that utilize turbulent transport theories. Here, two inter-related questions pertinent to the WMA for modelling scalar sources, sinks, and fluxes at seasonal to inter-annual time scales are explored: (1) if the WMA is to be replaced so as to resolve this two-way interaction, how detailed must the turbulent transport model be? And (2) what are the added predictive skills gained by resolving the two-way interaction vis-a-vis other uncertainties such as seasonal variations in physiological parameters. These two questions are addressed by simulating multi-year mean scalar concentration and eddy-covariance scalar flux measurements collected in a Loblolly pine (P. taeda L.) plantation near Durham, North Carolina, U.S.A. using turbulent transport models ranging from K-theory (or first-order closure) to third-order closure schemes. The multi-layer model calculations with these closure schemes were contrasted with model calculations employing the WMA. These comparisons suggested that (i) among the three scalars, sensible heat flux predictions are most biased with respect to eddy-covariance measurements when using the WMA, (ii) first-order closure schemes are sufficient to reproduce the seasonal to inter-annual variations in scalar fluxes provided the canonical length scale of turbulence is properly specified, (iii) second-order closure models best agree with measured mean scalar concentration (and temperature) profiles inside the canopy as well as scalar fluxes above the canopy, (iv) there are no clear gains in predictive skills when using third-order closure schemes over their second-order closure counterparts. At inter-annual time scales, biases in modelled scalar fluxes incurred by using the WMA exceed those incurred when correcting for the seasonal amplitude in the maximum carboxylation capacity (V cmax, 25) provided its mean value is unbiased. The role of local thermal stratification inside the canopy and possible computational simplifications in decoupling scalar transfer from the generation of the flow statistics are also discussed. “The tree, tilting its leaves to capture bullets of light; inhaling, exhaling; its many thousand stomata breathing, creating the air”. Ruth Stone, 2002, In the Next Galaxy

Sensitivity of simulated wintertime Arctic atmosphere to vertical resolution in the ARPEGE/IFS model

The current state-of-the-art general circulation models, including several of those used by the IPCC, show considerable biases in the simulated present day high-latitude climate compared to observations and reanalysis data. These biases are most pronounced during the winter season. We here employ ideal vertical profiles of temperature and wind from turbulence-resolving simulations to perform a priori studies of the first-order eddy-viscosity closure scheme employed in the ARPEGE/IFS model. This reveals that the coarse vertical resolution (31 layers) of the model cannot be expected to realistically resolve the Arctic stable boundary layer. The curvature of the Arctic inversion and thus also the vertical turbulent-exchange processes cannot be reproduced by the coarse vertical mesh employed. To investigate how turbulent vertical exchange processes in the Arctic boundary layer are represented by the model parameterization, a simulation with high vertical resolution (90 layers in total) in the lower troposphere is performed. Results from the model simulations are validated against data from the ERA-40 reanalysis. The dependence of the surface air temperature on surface winds, surface energy fluxes, free atmosphere stability and boundary layer height is investigated. The coarse-resolution run reveals considerable biases in these parameters, and in their physical relations to surface air temperature. In the simulation with fine vertical resolution, these biases are clearly reduced. The physical relation between governing parameters for the vertical turbulent-exchange processes improves in comparison with ERA-40 data.

Effect of stability on mixing in open canopies

In open canopies, the within-canopy flux from the ground surface and understory can account for a significant fraction of the total flux above the canopy. Here, “open” canopies refer to sufficient spacing between vegetation elements to allow significant clear-sky radiative heating and cooling of the ground or understory surface, which in turn significantly influences the subcanopy turbulence. This study incorporates the important influence of within-canopy stability on turbulent mixing and subcanopy fluxes into a first-order closure scheme. Toward this goal, we analyze within-canopy eddy-correlation data from the old aspen site in the Boreal ecosystem–atmosphere study (BOREAS) and a mature ponderosa pine site in Central Oregon, USA. A formulation of within-canopy transport is framed in terms of a stability-dependent eddy diffusivity, which approaches Monin–Obukhov similarity theory above the canopy roughness sublayer. The new simple formulation for the subcanopy is an improvement upon the usual neglect of the influence of stability on turbulent mixing. However, frequent well-defined cold air drainage within the pine subcanopy inversion reduces the utility of simple models for nocturnal transport. Other shortcomings of the formulation are discussed.

Intermittent Turbulence and Oscillations in the Stable Boundary Layer over Land. Part I: A Bulk Model

In the stable boundary layer (SBL) it is observed often that turbulence is not continuous in space and time. This discontinuous, intermittent turbulence causes alternations from the mean evolution of the stratified atmospheric boundary layer, which may result in an oscillatory type of behavior of the near-surface wind speed and temperature. It is well known that intermittent turbulence in the SBL can be generated by various mechanisms. This paper focuses on an intermittency generating mechanism that arises from a direct interaction of the lower atmosphere (first tens of meters) with the vegetation surface, without interaction with the air aloft. It is shown that the essence of this mechanism can be captured by a 1D bulk model of three coupled nonlinear differential equations. In the present paper, numerical runs with the model show that intermittent turbulence is most likely to occur over land surfaces with low vegetation during clear-sky conditions in the presence of a moderate to low synoptical pressure gradient. The existence of a vegetation layer has a strong influence on intermittency dynamics. Due to its small heat capacity, the vegetation temperature is able to quickly respond to rapidly changing conditions. This in turn affects the stability of the lower atmosphere, causing an important feedback mechanism. In addition, it was found that intermittent behavior of SBL models occurs for various first-order closure schemes with different stability functions. However, stability functions that allow turbulent transport beyond the critical Richardson number effectively suppress intermittent‐oscillatory behavior. Currently, the latter type of formulations is often used in numerical weather prediction to prevent excessive SBL cooling in very stable conditions. The advantage of using a simplified SBL model, as proposed in the present paper, is that it allows an analytical study of the system, which, in turn, allows theoretical predictions about the occurrence of intermittent SBL behavior (see the companion paper).

High resolution flight observations and numerical simulations: horizontal variability in the wintertime boreal boundary layer

High resolution aircraft observations made along flight tracks over inhomogeneous surface in the late wintertime boreal zone are described and compared to 2D mesoscale model simulations with surface properties defined at 2 km resolution from maps. All observations displayed the expected small-scale turbulence. On top of that, the near-surface wind speeds (but not directions) showed mesoscale variations related to local topography and roughness. Upward (but not downward) SW and LW radiative fluxes and ground temperature also displayed mesoscale variability; in SW radiation this was clearly due to local albedo changes. In the sensible heat flux there was strong horizontal variation near the surface in correlation with surface types. The above observed mesoscale along-track variations were reasonably well represented by the mesoscale model simulation. The track-averaged observed sensible and latent heat flux profiles were in rough agreement with a mixing length approach, which used the track-averaged wind, temperature and moisture profiles as input (mimicking a first-order turbulence closure scheme of a GCM).

The Modeling of Variable Density Turbulent Flows. A review of first-order closure schemes

The paper is intended as a review of where eddy-viscosity turbulence models have reached in accounting for the several distinct effects of variable density and high-Mach number behaviour on the development of turbulent shear flows. Three generations of ``compressible models'' are depicted: a first one based on variable density and compressible adjustments of ``incompressible'' schemes, a second one in which it is presumed that explicit dilatational terms can account for variable density and compressibility effects, and a third one, where such effects are taken implicitly in association with structural changes of the turbulent field. The latter are hardly tractable when using first-order closure schemes, but more reliable than the former in accounting for most observed variable density and compressibility effects in turbulent shear flows.

Effect of three different boundary-layer parameterisations in a regional atmospheric model on the simulation of summer monsoon circulation

Bulk, first-order and turbulent kinetic energy (TKE) closure schemes are used to parameterise the boundary-layer physics in a high resolution, limited area model. The model was used to simulate the summer monsoon circulations over India. The domain selected included the monsoon trough over northern India, a region of mesoscale convection. A monsoon depression was present at the time of the simulation. The results indicate that the TKE closure scheme combined with the Monin–Obukhov surface-layer similarity relation provided the best 48-hour simulation of the circulation and the rainfall associated with the monsoon depression.