NCAR Community Atmosphere Model Research Articles

Sharp decline of dust events induces regional wetting over arid and semi-arid Northwest China in the NCAR Community atmosphere model

Multiple lines of observational evidence have indicated a significant wetting over the arid and semi-arid Northwest China (NWC) during recent decades, coinciding with a simultaneous sharp decline of dust events. Although recent studies have attributed NWC wetting to different anthropogenic and natural forcings, the mechanisms are not definitive and the regional wetting has been greatly underestimated in the Coupled Model Intercomparison Project historical simulations. Based on sensitivity experiments with different dust emission amounts using the NCAR Community Atmospheric Model version 5 (CAM5), here we find that decreasing dusts exert significant impacts on mixed-phase clouds through reducing the concentration of ice nucleating particles, increase the NWC precipitation and thus induce regional wetting through enhancing convection precipitation. A possible convection invigoration mechanism whereby the atmospheric vertical temperature gradient and convective instability are strengthened by reduced dusts, leading to convection invigoration and increased precipitation. These results are reinforced by simulations over the dust region in North Africa where mixed-phase and ice clouds are rare and reduced dusts do not increase precipitation. This study highlights the possible mechanism of dust-ice cloud interactions in recent NWC wetting and future regional climate change.

Insights of warm-cloud biases in Community Atmospheric Model 5 and 6 from the single-column modeling framework and Aerosol and Cloud Experiments in the Eastern North Atlantic (ACE-ENA) observations

Abstract. There has been a growing concern that most climate models predict precipitation that is too frequent, likely due to lack of reliable subgrid variability and vertical variations in microphysical processes in low-level warm clouds. In this study, the warm-cloud physics parameterizations in the singe-column configurations of NCAR Community Atmospheric Model version 6 and 5 (SCAM6 and SCAM5, respectively) are evaluated using ground-based and airborne observations from the Department of Energy (DOE) Atmospheric Radiation Measurement (ARM) Aerosol and Cloud Experiments in the Eastern North Atlantic (ACE-ENA) field campaign near the Azores islands during 2017–2018. The 8-month single-column model (SCM) simulations show that both SCAM6 and SCAM5 can generally reproduce marine boundary layer cloud structure, major macrophysical properties, and their transition. The improvement in warm-cloud properties from the Community Atmospheric Model 5 and 6 (CAM5 to CAM6) physics can be found through comparison with the observations. Meanwhile, both physical schemes underestimate cloud liquid water content, cloud droplet size, and rain liquid water content but overestimate surface rainfall. Modeled cloud condensation nuclei (CCN) concentrations are comparable with aircraft-observed ones in the summer but are overestimated by a factor of 2 in winter, largely due to the biases in the long-range transport of anthropogenic aerosols like sulfate. We also test the newly recalibrated autoconversion and accretion parameterizations that account for vertical variations in droplet size. Compared to the observations, more significant improvement is found in SCAM5 than in SCAM6. This result is likely explained by the introduction of subgrid variations in cloud properties in CAM6 cloud microphysics, which further suppresses the scheme's sensitivity to individual warm-rain microphysical parameters. The predicted cloud susceptibilities to CCN perturbations in CAM6 are within a reasonable range, indicating significant progress since CAM5 which produces an aerosol indirect effect that is too strong. The present study emphasizes the importance of understanding biases in cloud physics parameterizations by combining SCM with in situ observations.

Indian summer monsoon simulations in successive generations of the NCAR Community Atmosphere Model

Indian summer monsoon simulations in successive generations of the NCAR Community Atmosphere Model

Probing the Skill of Random Forest Emulators for Physical Parameterizations Via a Hierarchy of Simple CAM6 Configurations

AbstractMachine learning approaches, such as random forests (RF), have been used to effectively emulate various aspects of climate and weather models in recent years. The limitations to these approaches are not yet known, particularly with regards to varying complexity of the underlying physical parameterization scheme within the climate model. Utilizing a hierarchy of model configurations, we explore the limits of random forest emulator skill using simplified model frameworks within NCAR's Community Atmosphere Model, version 6 (CAM6). These include a dry CAM6 configuration, a moist extension of the dry model, and an extension of the moist case that includes an additional convection scheme. Each model configuration is run with identical resolution and over the same time period. With unique RF being optimized for each tendency or precipitation rate across the hierarchy, we create a variety of “best case” emulators. The random forest emulators are then evaluated against the CAM6 output as well as a baseline neural network emulator for completeness. All emulators show significant skill when compared to the “truth” (CAM6), often in line with or exceeding similar approaches within the literature. In addition, as the CAM6 complexity is increased, the random forest skill noticeably decreases, regardless of the extensive tuning and training process each random forest goes through. This indicates a limit on the feasibility of RF to act as physics emulators in climate models and encourages further exploration in order to identify ideal uses in the context of state‐of‐the‐art climate model configurations.

Climate Impacts of Convective Cloud Microphysics in NCAR CAM5

Abstract We improved the treatments of convective cloud microphysics in the NCAR Community Atmosphere Model version 5.3 (CAM5.3) by 1) implementing new terminal velocity parameterizations for convective ice and snow particles, 2) adding graupel microphysics, 3) considering convective snow detrainment, and 4) enhancing rain initiation and generation rate in warm clouds. We evaluated the impacts of improved microphysics on simulated global climate, focusing on simulated cloud radiative forcing, graupel microphysics, convective cloud ice amount, and tropical precipitation. Compared to CAM5.3 with the default convective microphysics, the too-strong cloud shortwave radiative forcing due primarily to excessive convective cloud liquid is largely alleviated over the tropics and midlatitudes after rain initiation and generation rate is enhanced, in better agreement with the CERES-EBAF estimates. Geographic distributions of graupel occurrence are reasonably simulated over continents; whereas the graupel occurrence remains highly uncertain over the oceanic storm-track regions. When evaluated against the CloudSat–CALIPSO estimates, the overestimation of convective ice mass is alleviated with the improved convective ice microphysics, among which adding graupel microphysics and the accompanying increase in hydrometeor fall speed play the most important role. The probability distribution function (PDF) of rainfall intensity is sensitive to warm rain processes in convective clouds, and enhancement in warm rain production shifts the PDF toward heavier precipitation, which agrees better with the TRMM observations. Common biases of overestimating the light rain frequency and underestimating the heavy rain frequency in GCMs are mitigated.

Description and Evaluation of a New Deep Convective Cloud Model Considering In‐Cloud Inhomogeneity

AbstractConvections still need to be parameterized in general circulation models (GCMs) in the coming decades. Performances of GCMs are significantly influenced by the convection schemes used. In contrast to most conventional cloud models that ignore in‐cloud inhomogeneities, a new convective cloud model explicitly considering in‐cloud inhomogeneities is developed. The new model adopts a single bulk plume approach, but divides the plume into a series of interacting sub‐plumes in order to mimic the transition from the plume core to its edges. We implemented the new cloud model in NCAR Community Atmosphere Model version 5.0 (CAM5) and evaluated its performance. Single‐column tests show a stronger mass flux profile and low‐level detrainment with the new model. In global simulations, the long‐lasting wet bias in the tropical free troposphere of CAM5 is alleviated. Spatial pattern and intensity distribution of precipitation is improved, but the global hydrological cycle is over‐enhanced. The diurnal cycle of tropical precipitation is also better captured though the precipitation peak still occurs earlier than the observations and the amplitude of the cycle tends to weaken. MJO simulation is slightly improved with the new scheme. In addition, the new cloud model generated more tropical cyclones, in better agreement with observations.

Remote effect of model systematic bias in tropical SST on the cold bias over the Tibetan Plateau

Most state-of-the-art climate models substantially underestimate the near-surface air temperature (SAT) over the Tibetan Plateau (TP), especially for the cold season. While previous studies have attributed this cold bias to local factors such as the elevation difference, coarse resolution, and excessive snow cover, this investigation addresses the potential contributions of the systematic bias of tropical sea surface temperature (SST) to the TP cold bias. Experiments with the NCAR Community Atmosphere Model demonstrate that tropical SST bias results in an apparent cold bias over the TP, especially in boreal winter, and explains about 40% of the deviation in multi-model mean SAT over the TP relative to station observations. Forced by the tropical SST bias, heat flux exhibits an anomalous divergence over the plateau, causing a cooling center in the mid- and lower-troposphere over the TP. This atmospheric cooling in turn leads to a reduction of the downward longwave radiative fluxes reaching the surface, less energy supply, and thus a cold bias over the TP.

Improved Climate Simulation by Using a Double‐Plume Convection Scheme in a Global Model

AbstractConvective parameterization can drastically regulate the mean climate and tropical transient activity of a General circulation model (GCM). In this study, the physics suite of the NCAR Community Atmosphere Model, version 5 (CAM5) was first ported to the Global‐to‐Regional Integrated Forecast System model. Then, the original convective parameterization of CAM5—with a separate representation of deep convection Zhang–Mcfarlane (ZM) and shallow convection University of Washington (UW)—was replaced by a double‐plume (DP) scheme. This DP scheme adopts a quasi‐unified representation of shallow and deep convection within a single framework. Results demonstrate that the new scheme brings about several improvements in the modeled climate. The differences in the trigger and closure assumptions, lateral mixing rate, and cloud model for the deep convection result in systematic regional differences in the simulated precipitation pattern, cloud vertical structure, and the associated radiative forcing. Compared with ZM‐UW, DP reduces the biases in precipitation over the Indian Ocean, ameliorates the “high‐frequency and low‐intensity” problem of tropical precipitation, and leads to an improved representation of tropical variability, including the Madden–Julian Oscillation. Double‐plume reduces low clouds and increases high clouds in the tropics, due to its internal parallel‐split convective processes and smaller cumulus cloud fraction. Discussions related to parametric tuning of convective parameterization are also presented.

Implementation and Evaluation of a Double-Plume Convective Parameterization in NCAR CAM5

AbstractPerformance of global climate models (GCMs) is strongly affected by the cumulus parameterization (CP) used. Similar to the approach in GFDL AM4, a double-plume CP, which unifies the deep and shallow convection in one framework, is implemented and tested in the NCAR Community Atmospheric Model version 5 (CAM5). Based on the University of Washington (UW) shallow convection scheme, an additional plume was added to represent the deep convection. The shallow and deep plumes share the same cloud model, but use different triggers, fractional mixing rates, and closures. The scheme was tested in single-column, short-term hindcast, and AMIP simulations. Compared with the default combination of the Zhang–McFarlane scheme and UW scheme in CAM5, the new scheme tends to produce a top-heavy mass flux profile during the active monsoon period in the single-column simulations. The scheme increases the intensity of tropical precipitation, closer to TRMM observations. The new scheme increased subtropical marine boundary layer clouds and high clouds over the deep tropics, both in better agreement with observations. Sensitivity tests indicate that regime-dependent fractional entrainment rates of the deep plume are desired to improve tropical precipitation distribution and upper troposphere temperature. This study suggests that a double-plume approach is a promising way to combine shallow and deep convections in a unified framework.

Interdecadal Variations of Different Types of Summer Heat Waves in Northeast China Associated with AMO and PDO

AbstractThe summer heat waves (HWs) in Northeast China (NEC) during 1961–2016 can be classified into two types, namely, wave train HWs and blocking HWs based on the hierarchical clustering algorithm by using ERA-Interim daily datasets. Wave train HWs occurred accompanied by eastward-moving wave trains with a “− + − +” structure formed over Eurasia, while the blocking HWs occurred with blocking circulation anomalies over Eurasia. In general, the blocking HWs could cause the positive temperature anomalies in NEC to last longer than wave train HWs. During the period from 1961 to 2016, the wave train HWs experienced an interdecadal variation from less to more, while the blocking HWs experienced interdecadal variations of less–more–less. Regression analysis and information flow indicate that the interdecadal variation of the wave train HWs is associated with Atlantic multidecadal oscillation (AMO) and Pacific decadal oscillation (PDO), while the interdecadal variation of the blocking HWs is more likely associated with PDO. The positive phase of AMO (negative phase of PDO) could increase the wave train (blocking) HWs by strengthening the zonal wave train similar to the Silk Road pattern (the arched wave train like the polar–Eurasian pattern). The observed results are in agreement with the numerical experiments with the NCAR Community Atmosphere Model, version 5.3.

Understanding the Roles of Convective Trigger Functions in the Diurnal Cycle of Precipitation in the NCAR CAM5

AbstractThe wrong diurnal cycle of precipitation is a common weakness of current global climate models (GCMs). To improve the simulation of the diurnal cycle of precipitation and understand what physical processes control it, we test a convective trigger function described in Xie et al. with additional optimizations in the NCAR Community Atmosphere Model version 5 (CAM5). The revised trigger function consists of three modifications: 1) replacing the convective available potential energy (CAPE) trigger with a dynamic CAPE (dCAPE) trigger, 2) allowing convection to originate above the top of planetary boundary layer [i.e., the unrestricted air parcel launch level (ULL)], and 3) optimizing the entrainment rate and threshold value of the dynamic CAPE generation rate for convection onset based on observations. Results from 1° resolution simulations show that the revised trigger can alleviate the long-standing GCM problem of too early maximum precipitation during the day and missing the nocturnal precipitation peak that is observed in many regions, including the U.S. southern Great Plains (SGP). The revised trigger also improves the simulation of the propagation of precipitation systems downstream of the Rockies and the Amazon region. A further composite analysis over the SGP unravels the mechanisms through which the revised trigger affects convection. Additional sensitivity tests show that both the peak time and the amplitude of the diurnal cycle of precipitation are sensitive to the entrainment rate and dCAPE threshold values.

Quantitative study of the interannual variability of South Asian summer monsoon rainfall regulated by SST

AbstractMost of the rainfall over South Asia occurs from June to September, and its variation is strongly related to the fluctuation of sea surface temperature (SST) on a global scale. In this study, we investigate how and to what degree the interannual variance (IAV) of South Asian summer monsoon (SASM) rainfall is regulated by SST anomalies through comparing the multi‐model simulations forced by observed interannual SST variation and climatological SST just with annual cycle. Results indicate that the interannual change in global SST can explain nearly 50% of the IAV of SASM rainfall, in which dynamic process (via atmospheric circulation), instead of thermodynamic process (via water‐vapour effect), plays a dominated role. SST influences the IAV of SASM rainfall mainly through changing the lower‐ and upper‐level atmospheric circulations, namely, the large‐scale vertical wind shears. Additional sensitivity experiments with the NCAR Community Atmosphere Model show that the effect of tropical Pacific SST (~24.0%) is more important than that of tropical Indian Ocean SST (~15.3%) on regulating the IAV of SASM rainfall.

Uncertainty quantification based cloud parameterization sensitivity analysis in the NCAR community atmosphere model

Using uncertainty quantification techniques, we carry out a sensitivity analysis of a large number (17) of parameters used in the NCAR CAM5 cloud parameterization schemes. The LLNL PSUADE software is used to identify the most sensitive parameters by performing sensitivity analysis. Using Morris One-At-a-Time (MOAT) method, we find that the simulations of global annual mean total precipitation, convective, large-scale precipitation, cloud fractions (total, low, mid, and high), shortwave cloud forcing, longwave cloud forcing, sensible heat flux, and latent heat flux are very sensitive to the threshold-relative-humidity-for-stratiform-low-clouds (rhminl) and the auto-conversion-size-threshold-for-ice-to-snow left( {dcs} right). The seasonal and regime specific dependence of some parameters in the simulation of precipitation is also found for the global monsoons and storm track regions. Through sensitivity analysis, we find that the Somali jet strength and the tropical easterly jet associated with the south Asian summer monsoon (SASM) show a systematic dependence on dcs and rhminl. The timing of the withdrawal of SASM over India shows a monotonic increase (delayed withdrawal) with an increase in dcs. Overall, we find that rhminl, dcs, ai, and as are the most sensitive cloud parameters and thus are of high priority in the model tuning process, in order to reduce uncertainty in the simulation of past, present, and future climate.

Interdecadal Variation of Winter Cold Surge Path in East Asia and Its Relationship with Arctic Sea Ice

AbstractThe paths of winter cold surge (CS) events in East Asia (EA) from 1979 to 2017 are tracked by the Flexible Particle (FLEXPART) model using ERA-Interim daily datasets, and the probability density distribution of the paths is calculated by the kernel density estimation (KDE) method. The results showed that the paths of CSs are significantly correlated with the intensity of the CSs, which shows an interdecadal transition from weak to strong around 1995. CS paths can be classified into two types, namely, the western path type and the northern path type, which were more likely to occur before and after 1995, respectively. Before 1995, the cold air mainly originated from Europe and moved from west to east, and the synoptic features were associated with the zonal wave train. After 1995, cold air accumulated over western Siberia and then invaded EA along the northern path, and the synoptic features were mainly associated with the blocking structure. The geopotential height (GPH) anomalies over the Arctic were abnormally strong. This paper further analyzes the relationship between CSs and winter sea ice concentration (SIC) in the Arctic. The results show that the intensity of CSs is negatively correlated with the Barents SIC (BSIC). When the BSIC declines, the upward wave flux over the Barents Sea is enhanced and expanded to the midlatitude region. GPH anomalies over the Arctic are positive and form a negative AO-like pattern, which is conducive to the formation of the northern path CS. Furthermore, the observed results are supported by numerical experiments with the NCAR Community Atmosphere Model, version 5.3 (CAM5.3).

Analysis of climate sensitivity via high-dimensional principal component regression

Uncertainty in the sensitivity of global mean air temperature to changes in atmospheric CO2 concentration arises from factors controlling the response of cloud amounts. Here, we present an analysis of these factors within an ensemble consisting of 164 variations of the NCAR Community Atmosphere Model version 3.1, each of which differs slightly in their treatment of convection and cloud radiative properties. The problem is to uniquely relate observable aspects of model output, which are inherently high dimensional such as the 64,000 spatially mapped values of 11 fields generated from each CAM3.1 model to a scalar prediction of each model’s sensitivity to CO2 forcing. We approach this problem using principal component regression and provide maps of observable predictors. Furthermore, based on this approach, we develop prediction intervals and show that they closely match their advertised coverage.

A Test of Emergent Constraints on Cloud Feedback and Climate Sensitivity Using a Calibrated Single-Model Ensemble

AbstractA calibrated single-model ensemble (SME) derived from the NCAR Community Atmosphere Model, version 3.1, is used to test two hypothesized emergent constraints on cloud feedback and equilibrium climate sensitivity (ECS). The Fasullo and Trenberth relative humidity (RH) metric and the Sherwood et al. lower-tropospheric mixing (LTMI) metric are computed for the present-day climate of the SME, and the relationships between the metrics, ECS, and cloud and other climate feedbacks are examined. The tropical convergence zone relative humidity (RHM) and the parameterized lower-tropospheric mixing (LTMIS) are positively correlated to ECS, and each is associated with a different spatial pattern of tropical shortwave cloud feedback in the SME. However, neither of those metrics is linked to the type of cloud response hypothesized by its authors. The resolved lower-tropospheric mixing (LTMID) is positively correlated to ECS for a subset of the SME having LTMID over a threshold value. LTMI and the RH for the dry, descending branch of the Hadley cell (RHD) narrow and shift upward the posterior estimates of ECS in the SME, but the SME bias in RHD and concerns over poorly understood physical mechanisms suggest the narrowing could be spurious for both constraints. While calibrated SME results may not generalize to multimodel ensembles, they can enhance the process understanding of emergent constraints and serve as out-of-sample tests of robustness.

Importance of the resolution of surface topography vis-à-vis atmospheric and surface processes in the simulation of the climate of Himalaya-Tibet highland

As Himalaya-Tibet Highland (HTH), a spatially extensive and complex terrain, plays influential roles in the regional and global climate, its representation in climate models is a decisive factor in climate simulations. It is established that higher spatial resolution improves the simulation of several seasonal mean features over this region, but how much improvement comes from the better representation of surface topography and how much comes from the better modeling of the atmospheric and surface processes are still not clear. To understand this, three sets of 6-member ensemble simulations are conducted using the NCAR Community Atmosphere Model version 5.1 (CAM5.1) at: (1) 1.9° × 2.5° resolution (Coarse), (2) 0.47° × 0.63° resolution (Fine), and (3) topography is prescribed at the Coarse resolution and rest of the model processes are computed at the Fine resolution (Hybrid). The Coarse resolution overlooks most of the intricate features of the topography with a severe bias of ~ 1–2 km but Fine resolution does a much better job in their representations. Surface air temperatures are found to be strongly dependent on the resolution of topography and therefore it is desirable to resolve topography in models for the realistic simulation and projection of temperatures. Rest of the variables, viz. seasonal mean, seasonal cycle and probability distribution of precipitation; tropospheric temperature and moisture; Tibetan anticyclone, show remarkable improvements with the increase in resolution of the atmospheric and surface processes. Influence of the resolution of topography is found to be limited to the model levels adjacent to the surface, and for higher model levels, the resolution of the atmospheric and surface processes is noted to play a more crucial role.

A Diagnostic PDF Cloud Scheme to Improve Subtropical Low Clouds in NCAR Community Atmosphere Model (CAM5)

AbstractLow clouds strongly impact the radiation budget of the climate system, but their simulation in most GCMs has remained a challenge, especially over the subtropical stratocumulus region. Assuming a Gaussian distribution for the subgrid‐scale total water and liquid water potential temperature, a new statistical cloud scheme is proposed and tested in NCAR Community Atmospheric Model version 5 (CAM5). The subgrid‐scale variance is diagnosed from the turbulent and shallow convective processes in CAM5. The approach is able to maintain the consistency between cloud fraction and cloud condensate and thus alleviates the adjustment needed in the default relative humidity‐based cloud fraction scheme. Short‐term forecast simulations indicate that low cloud fraction and liquid water content, including their diurnal cycle, are improved due to a proper consideration of subgrid‐scale variance over the southeastern Pacific Ocean region. Compared with the default cloud scheme, the new approach produced the mean climate reasonably well with improved shortwave cloud forcing (SWCF) due to more reasonable low cloud fraction and liquid water path over regions with predominant low clouds. Meanwhile, the SWCF bias over the tropical land regions is also alleviated. Furthermore, the simulated marine boundary layer clouds with the new approach extend further offshore and agree better with observations. The new approach is able to obtain the top of atmosphere (TOA) radiation balance with a slightly alleviated double ITCZ problem in preliminary coupled simulations. This study implies that a close coupling of cloud processes with other subgrid‐scale physical processes is a promising approach to improve cloud simulations.

Distinct Contributions of Ice Nucleation, Large‐Scale Environment, and Shallow Cumulus Detrainment to Cloud Phase Partitioning With NCAR CAM5

AbstractMixed‐phase clouds containing both liquid droplets and ice particles occur frequently at high latitudes and in the midlatitude storm track regions. Simulations of the cloud phase partitioning between liquid and ice hydrometeors in state‐of‐the‐art global climate models are still associated with large biases. In this study, the phase partitioning in terms of liquid mass phase ratio (MPRliq, defined as the ratio of liquid mass to total condensed water mass) simulated from the NCAR Community Atmosphere Model version 5 (CAM5) is evaluated against the observational data from A‐Train satellite remote sensors. Modeled MPRliq is significantly lower than observations on the global scale, especially in the Southern Hemisphere (e.g., Southern Ocean and the Antarctic). Sensitivity tests with CAM5 are conducted to investigate the distinct contributions of heterogeneous ice nucleation, shallow cumulus detrainment, and large‐scale environment (e.g., winds, temperature, and water vapor) to the low MPRliq biases. Our results show that an aerosol‐aware ice nucleation parameterization increases the MPRliq especially at temperatures colder than −20°C and significantly improves the model agreements with observations in the Polar regions in summer. The decrease of threshold temperature over which all detrained cloud water is liquid from 268 to 253 K enhances the MPRliq and improves the MPRliq mostly over the Southern Ocean. By constraining water vapor in CAM5 toward reanalysis, modeled low biases in many geographical regions are largely reduced through a significant decrease of cloud ice mass mixing ratio.

Role of Vertical Structure of Convective Heating in MJO Simulation in NCAR CAM5.3

Abstract Observational studies suggest that the vertical structure of diabatic heating is important to MJO development. In particular, the lack of a top-heavy heating profile was believed to be responsible for poor MJO simulations in global climate models. In this work, the role of the vertical heating profile in MJO simulation is investigated by modifying the convective heating profile to different shapes, from top-heavy heating to bottom cooling, to mimic mesoscale heating in the NCAR Community Atmosphere Model, version 5.3 (CAM5.3). Results suggest that incorporating a mesoscale stratiform heating structure can significantly improve the MJO simulation. By artificially adding stratiform-like heating and cooling in the experiments, many observed features of MJO are reproduced, including clear eastward propagation, a westward-tilted vertical structure of MJO-scale anomalies of dynamic and thermodynamic fields, and strong 20–80-day spectral power. Further analysis shows an abundance of shallow convection ahead of MJO deep convection, confirming the role of shallow convection in preconditioning the atmosphere by moistening the lower troposphere ahead of deep convection during the MJO life cycle. Additional experiments show that lower-level cooling contributes more to improving the MJO simulation. All these features are lacking in the control simulation, suggesting that the mesoscale stratiform heating, especially its lower-level cooling component, is important to MJO simulation.