1. Introduction
Mesoscale convective systems (MCSs), characterized by active convective towers and expansive stratiform regions, represent the largest form of cumulonimbus clouds in our climate system (Hartmann et al. 1984; Mapes and Houze 1993; Houze 2004). They contribute significantly to tropical and midlatitude rainfall, playing a crucial role in redistributing heat, moisture, and momentum (Laing and Fritsch 1997; Maddox 1980; Yuan and Houze 2010; Roca et al. 2014). Under favorable environmental conditions, MCSs can persist for extended periods, leading to extreme precipitation events and hazardous weather like flooding, gusty winds, hail, and tornadoes (Laing and Fritsch 2000; Schumacher and Johnson 2006; Feng et al. 2018; Hu et al. 2021). Therefore, accurately simulating MCSs is key to understanding both mean climate and extreme weather events.
MCSs exhibit distinct features compared to individual convection, including the vertically tilted circulations known as slantwise layer overturning, that occur across scales (Moncrieff 2004; Houze et al. 2015); interaction between embedded fine-scale cumulus convection, low-level environmental shear, and the cold outflows (Yano et al. 2014); and the upscale growth and interactions with large-scale dynamics (Houze 2018). Coarser resolution (>100 km) general circulation models (GCMs) struggle to simulate these processes due to scale-separation assumptions in convection parameterizations, resulting in challenges modeling extreme weather linked to MCS activities (Donner 1993; Donner et al. 2001; Moncrieff 2004; Lin et al. 2017; Moncrieff et al. 2017; Lin et al. 2019). Various approaches have been explored to tackle these challenges, including additional parameterization schemes, high-resolution models, and superparameterization. The multiscale coherent structure parameterization (MCSP) scheme (Moncrieff et al. 2017), implemented in models like NCAR CAM and the Energy Exascale Earth System Model (E3SM) (Chen et al. 2021), has shown some improvements, but significant biases remain. High-resolution (∼50 km) GCMs have demonstrated enhanced capabilities compared to coarser-resolution models, but results vary largely across different models (Dong et al. 2021; Feng et al. 2021a; Zhao 2022; Dong et al. 2023a). Superparameterization, embedding a cloud-resolving model (CRM) within the GCM column (Grabowski and Smolarkiewicz 1999), also improves MCS simulations, though biases in MCS-related precipitation remain (Kooperman et al. 2013; Chen and Kirtman 2018; Lin et al. 2022; Hsu et al. 2023).
These studies highlights that high resolution and explicit convection representation improve MCS simulations. There is a growing interest in using 25-km GCMs to investigate MCS features. For instance, Feng et al. (2021a) compared 50- and 25-km MCS simulations over North America with a variable-resolution CAM, and Hsu et al. (2023) evaluated MCS simulations using ∼150- and 25-km superparameterized E3SM. Advancements in computational resources have enabled the utilization of convection-permitting models (CPMs) or global CPMs, also known as global storm-resolving models (GSRMs), with resolutions of a few kilometers. These models provide finely detailed and physically consistent simulations that are seamlessly across various scales. GSRMs, in particular, achieve all the goals of parameterizations and CPMs, capturing fine-scale convective features, their impact on larger-scale circulations, and the direct effects of external forcings. Regional CPMs and GSRMs have demonstrated the ability to capture important MCS characteristics (Feng et al. 2018; Prein et al. 2020; Feng et al. 2023; Li et al. 2023; Dong et al. 2024b). While these simulations offer valuable insights, they also have limitations. For instance, the analyses of MCS simulated in GSRMs based on the Dynamics of the Atmospheric General Circulation Modeled on Nonhydrostatic Domains (DYAMOND) summer and winter projects (Stevens et al. 2019) are constrained to 40 days due to high computational cost (Feng et al. 2023; Song et al. 2024). This could introduce biases due to the small sample size and single-season focus. Regional CPMs cover longer durations but are confined to smaller domains, limiting their ability to capture larger-scale impacts of MCSs (Prein et al. 2017). Additionally, CPMs often rely on pseudoglobal warming experiment to explore future changes (Schär et al. 1996). However, these experiments are limited in capturing global-scale interactions, boundary effects, and feedback—limitations that are not present in GSRMs. Simulations based on 25-km GCMs offer longer durations on a global scale, but existing studies primarily focus on the central United States, leaving gaps in understanding MCS characteristics elsewhere (Feng et al. 2021a; Hsu et al. 2023).
To address these gaps, in this study, we analyze global MCS features simulated by a 2-yr-long GSRM (∼3.25 km) and a 10-yr-long atmospheric GCM (∼25 km) developed at the Geophysical Fluid Dynamics Laboratory (GFDL). MCS features, including occurrence/genesis frequency, seasonality, and diurnal cycle, are examined. Moreover, we provide a statistical overview of MCS characteristics such as duration, size, intensity, propagation speed, and direction. The relationships among MCS size, duration, and intensity are explored at the event scale. MCS-associated precipitation is also analyzed in this study. The paper is organized as follows. The model simulation, observational datasets, and MCS tracking algorithm are described in section 2. The comparison of global MCS simulations in these two models with two satellite products is presented in section 3. Discussion and conclusions are given in section 4.
2. Data and methods
a. X-SHiELD and C384AM4 model simulations
The GSRM simulation analyzed in this study is based on the Experimental System for High-Resolution Prediction on Earth-to-Local Domains (X-SHiELD) (Cheng et al. 2022; Harris et al. 2023), a configuration of SHiELD (Harris et al. 2020). It uses a C3072 (∼3.25 km) cubed-sphere grid with 79 vertical levels from 10 m above the surface to 3 hPa. X-SHiELD couples the nonhydrostatic Finite-Volume Cubed-Sphere (FV3) dynamical core with GFDL microphysics (Zhou et al. 2019, 2022), the prognostic turbulence kinetic energy (TKE) form of the eddy-diffusivity mass-flux (EDMF) turbulence scheme (Han and Bretherton 2019), and the Noah land surface model (LSM). Shallow convection is parameterized using scale-aware simplified Arakawa–Schubert (SAS; Han et al. 2017), while deep convection is explicitly simulated. The model incorporates a mixed layer ocean model (Pollard et al. 1973), calculating the mixed layer depth and heat content as prognostic variables with tendencies derived from the net surface heat flux. SST is nudged toward real-time ECMWF SST analyses on a 15-day time scale. The simulation spans from late October 2019 to early January 2022, but our analysis focuses on a 2-yr period (2020–21). The outputs from this simulation have been used to evaluate mean precipitation and important weather phenomena, such as intense convection, rotating convective updrafts, and tropical MCSs (Cheng et al. 2022; Harris et al. 2023; Guendelman et al. 2024; Dong et al. 2024b).
We also use a high-resolution (∼25 km) version of the GFDL atmospheric model 4 (AM4) (referred to as C384AM4) with the two-moment Morrison–Gettelman (MG2) microphysics scheme. C384 denotes that there are 384 × 384 grid boxes (approximately 25-km horizontal grid spacing) in each of the six cubed-sphere faces. Developed from the original AM4 model (Zhao et al. 2018a,b), C384AM4 increases the horizontal resolution and replaces the one-moment Rotstayn–Klein (RK) microphysics scheme with the MG2 microphysics scheme (Guo et al. 2021, 2022). C384AM4 utilizes the double plume convection scheme introduced by Zhao et al. (2018a), which represents both shallow and deep convection. The assumptions regarding the determination of entrainment and detrainment rates, as well as the representation of vertical velocity in each plume, are consistent with those in Bretherton et al. (2004). However, the fractional lateral mixing rates for the two plumes are parameterized differently. The tunable values used for this parameterization are detailed in Zhao et al. (2018b). The simulation is an 11-yr climatological run, forced by SST and sea ice concentrations from Taylor et al. (2000), averaged over 1980–2014, with fixed solar, radiative gases and 2010 aerosol emissions. The first year is discarded as spinup. Since these simulations do not include interannual variability in forcing, the 10-yr period can be considered as 10 ensemble members. Previous versions of AM4 have been extensively employed in assessing atmospheric rivers, MCSs, and precipitation diurnal cycles (Zhao 2020, 2022; Dong et al. 2023a,b, 2024a). The findings consistently affirm the model’s reasonable capability in simulating these systems, bolstering our confidence in using this high-resolution version for in-depth examination of MCSs.
b. Observational datasets
Two satellite-based datasets are used to track MCSs: the Cloud Archive User Service (CLAUS) infrared brightness temperature Tb dataset (Hodges et al. 2000) and the NASA merged infrared Tb dataset (Janowiak et al. 2001). Both datasets have been validated across different climate regimes (Huang et al. 2018; Dong et al. 2021; Feng et al. 2023; Zhao 2022; Dong et al. 2023a). CLAUS offers 3-hourly Tb at a 1/3° resolution during 1985–2008, offering extensive temporal coverage that aligns well with the C384AM4 experiment. The NASA dataset, starting in 2000, provides a near-global (60°S–60°N) 4-km pixel-resolution infrared Tb dataset, which serves as a suitable reference for evaluating the X-SHiELD simulation, given its matching temporal range and resolution. In addition, both the 3-hourly Multi-Source Weighted-Ensemble Precipitation, version 2 (MSWEP V2; Beck et al. 2019), gridded precipitation dataset (0.1° × 0.1°; 1979–2020) and the half-hourly NASA Integrated Multi-satellitE Retrievals for Global Precipitation Measurement, version 7 (GPM/IMERG V07; Huffman et al. 2015), precipitation dataset (0.1° × 0.1°; 2020–21) are used to explore MCS-related precipitation.
For fair comparison between models and observations, all variables analyzed in this study are converted to 3-h intervals and regridded onto a 25-km grid resolution. Spatial maps presented in this study are limited to the range of 60°S–60°N to align with the spatial coverage of NASA observations. It is important to note that while MCSs and associated precipitation in X-SHiELD are primarily compared with NASA Tb and GPM precipitation during 2020–21 and C384AM4 is compared with CLAUS Tb and MSWEP precipitation during 1999–2008 due to their similar time coverage and original resolution, they are cross compared to consider both observational uncertainties and model differences.
c. MCS tracking algorithm
We adopt the two-step automated MCS detection and tracking algorithm developed by Huang et al. (2018), which identifies MCS candidates based on a 221-K Tb threshold and a minimum area of 30 000 km2 threshold. These threshold values have been used extensively in previous studies (Pope et al. 2009; Goyens et al. 2012; Fiolleau and Roca 2013). Sensitivity analyses (Huang et al. 2018; Dong et al. 2021, 2022; Hsu et al. 2023) suggest that while variations of 10 K for the Tb threshold and 5000–40 000 km2 for the minimum area coverage may affect the absolute number of tracked MCSs, their spatial distribution and statistical characteristics remain largely consistent. Following the initial identification, MCS candidates are tracked over consecutive time frames using area-overlapping criteria. They will be considered the same system if they spatially overlap by at least 15%. For small- or fast-moving MCSs with minimal or no overlap, a Kalman filter approach is employed to predict the potential movement of MCSs. The details on how the Kalman filter predicts MCS movement and the methods for selecting optimal objects are provided in appendix A of Huang et al. (2018). Tacked MCSs lasting less than 6 h are excluded from the analysis. Spatial collocation of each MCS and precipitation is performed within the region encompassed by the convective system. MCS occurrence frequency refers to the number of MCSs within a specific time and spatial domain, while MCS genesis frequency tracks only the initial occurrence of each MCS. Following Dong et al. (2023a), we utilize the harmonic analysis to determine the peak time and amplitude of the MCS occurrence frequency at different frequencies. The first harmonic component, namely, the diurnal periods, is analyzed in this study. MCS duration is defined as the total time an identified MCS persists, while MCS size is the grid area covered by the MCS. MCS intensity is measured by the difference between the minimum Tb of each MCS and 221 K (i.e., 221 − Tb), following a similar approach used in previous studies (Feng et al. 2023; You et al. 2024). The propagation direction of the MCSs is measured clockwise from due east within a range from −180° to 179.75°.
Spatial distribution of time-average brightness temperature (K) from (a) NASA observations during 2020–21, (b) X-SHiELD simulation during 2020–21, (d) CLAUS observations during 1999–2008, and (e) C384AM4 simulation during 0002–0011. (c) The difference between (b) and (a). (f) The difference between (e) and (d). The area-averaged (60°S–60°N) value (mean), PCorr, mean absolute error (MAE), and RMSE between NASA observation and X-SHiELD simulation; CLAUS observation and C384AM4 are listed in the top-right corner.
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0303.1
3. Results
a. MCS occurrence frequency
Figure 2 shows the observed and simulated mean MCS occurrence frequency. Over oceans, MCSs are frequently observed along the intertropical convergence zone (ITCZ), the South Pacific convergence zone (SPCZ), the Indian Ocean (IO), and the Maritime Continent. On land, the Maritime Continent, tropical Africa (TA), and Amazon region (AR) have intense MCS activities. Both observational datasets align with past studies in showing similar spatial patterns, though CLAUS indicates a higher overall MCS frequency (Yuan and Houze 2010; Houze et al. 2015; Feng et al. 2021b). Simulated MCS frequencies are similar across models, with prominent MCS regions well represented. Pattern correlations between NASA and X-SHiELD and CLAUS and C384AM4 exceed 0.85 (p < 0.01). When comparing X-SHiELD with CLAUS or C384AM4 with NASA, the pattern correlations remain consistently high. However, there are substantial absolute differences. As shown in Figs. 2c and 2f, both models tend to produce more MCSs in the oceans, notably in the Indian Ocean and Maritime Continent, with pronounced positive biases observed in C384AM4. Over land, C384AM4 generally exhibits an overall overestimation, and the observed relationship between orography and MCS occurrence is exaggerated in the model. These biases can be linked to Tb distribution. For example, over the Amazon, while C384AM4 has a higher mean Tb, it includes more samples with Tb below 221 K—the MCS tracking threshold (Fig. S1 in the online supplemental material). This is offset by a higher frequency of very high Tb values in C384AM4, resulting in a higher overall mean Tb. The discrepancy likely arises from the convective parameterization scheme, which struggles to accurately capture real-world intermittency. By contrast, the differences in X-SHiELD are more varied and of smaller magnitude. There is even a tendency to underestimate the MCS occurrence over sporadic regions over the tropical Africa, Amazon region, and Maritime Continent. The normalized root-mean-square error (RMSE) (RMSE divided by the mean) of MCS occurrence frequency between X-SHiELD and NASA is 1.3, compared to 1.8 when comparing C384AM4 with CLAUS.
As in Fig. 1, but for the annual mean occurrence frequency of MCSs (# per 2° × 2° grid). Six subregions, including TA, IO, EA, MC, U.S., and AR, indicated by green rectangles in (c) are used for subsequent regional analysis.
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0303.1
MCS occurrence frequency is influenced by both genesis and duration. We start by focusing on the genesis, while duration will be discussed in section 3c. Figure S2 shows that MCS genesis frequency closely mirrors occurrence frequency. Considering these spatial differences and aiming to facilitate comparison with previous studies, we have identified six hotspots (outlined by the green rectangles in Fig. 2c) for a detailed examination. When averaged over different subregions, the observed annual mean counts (see numbers listed in Fig. 3) from NASA indicate fewer MCS genesis compared to CLAUS, hovering near the lower boundary of ±1 standard deviation of CLAUS, except for the Indian Ocean and Maritime Continent, where NASA reports comparable MCS genesis counts. X-SHiELD aligns better with both NASA and CLAUS with biases ranging from −35% to 66%. The largest bias occurs over eastern Asia (EA) when compared with NASA. By contrast, C384AM4 demonstrates substantial overestimation, ranging from 30% to 284%, with the most significant bias observed over the central United States when compared with NASA.
Seasonal cycle of MCS genesis frequency (# per month) for each subregion based on observations (thick black lines) and simulations (thick blue lines). Lines with (without) circles are from the NASA and X-SHiELD (CLAUS and C384AM4). The light blue (gray) shading denotes the interannual spread for CLAUS and C384AM4, while the blue (black) dashed line denotes the individual year for NASA and X-SHiELD. The numbers in each panel represent the mean (±1 s.d.) of the MCS genesis number. Correlation coefficients between NASA and X-SHiELD and CLAUS and C384AM4 are listed in the top right of each panel.
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0303.1
b. Seasonality and diurnal cycle
In this section, we first explore the seasonal cycles of MCSs. Figure 3 compares the simulated seasonal cycles of MCS genesis frequency across these six subregions with observations. Overall, the simulated seasonal cycles align well with observations, with statistically significant correlation coefficients (p < 0.05) in most subregions. However, a notable bias appears in the Indian Ocean, where observations are inconsistent and models diverge. This inconsistency in observations may be attributed to gaps in satellite coverage over this region, while the discrepancies between the two models may suggest a misrepresentation of the strong air–sea interaction in this area in GCMs (Lau and Nath 2003; Held et al. 2019; Dong et al. 2020). Elsewhere, MCS genesis exhibits a single peak during the boreal/austral warm season in eastern Asia, central United States, and Amazon, while tropical Africa shows double peaks in boreal spring and fall. The Indian Ocean and Maritime Continent display a relatively minor seasonal cycle. These results remain consistent when we repeat the analysis for MCS occurrence frequency (Fig. S3).
We next examine the diurnal cycle of MCS frequency. Figure S3 and Fig. 4 show the first diurnal harmonic amplitude and phase for both the boreal summer (June–August) and winter (December–February) seasons. Regions like the ITCZ, monsoon areas, and Maritime Continent display larger diurnal amplitudes, aligning with the mean MCS occurrence patterns (Fig. S4 cf. Fig. 2). Regarding the diurnal phase, observed MCS peaks typically occur in the afternoon to evening over land and early morning over oceans (Wei and Pu 2022; Song et al. 2024). These characteristics are more pronounced in the summer hemisphere. For example, during boreal summer, nocturnal peaks are evident over tropical Africa, eastern Asia, and central United States, while austral summer shows peaks over southern Africa, Australia, and Amazon. These patterns remain largely consistent between the two observational datasets, although CLAUS exhibits more features compared to NASA, likely due to the much larger sample sizes used, especially noticeable in winter when MCS frequencies are lower.
Spatial distribution of the first diurnal phase (LST) of MCS occurrence frequency for (a)–(d) composite JJA and (e)–(h) composite DJF based on (a),(e) NASA during 2020–21, (b),(f) X-SHiELD during 2020–21, (c),(g) CLAUS during 1999–2008, and (d),(h) C384AM4 during 0002–0011. Regions where the diurnal amplitude to monthly mean precipitation ratio is smaller than 0.25 are masked in the diurnal phase panels.
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0303.1
Simulated diurnal amplitudes mirror their mean distribution and match observations in magnitude (Fig. S4). However, diurnal phases exhibit significant differences between models (Fig. 4). Over land, nocturnal peaks are inaccurately represented, with C384AM4 showing morning peaks instead of nocturnal peaks while X-SHiELD simulating peaks around midnight. Diurnal patterns over the ocean show smaller differences between models and observations, though systematic biases remain evident. These differences are clearly evident in the harmonic dial diagram presented in Fig. 5. In this diagram, each point represents either an observation or a model simulation for a specific season. The distance from the center represents the diurnal amplitude, while the angle from due north indicates the diurnal phase. Observational datasets generally show close agreement with each other on the dial diagram, although diurnal phase differences are notable over eastern Asia and central United States. Again, these discrepancies may be attributed in part to the relatively smaller sample sizes in the NASA dataset. Simulated diurnal amplitudes and phases vary from the observations across all subregions, but notably, X-SHiELD aligns more closely with the observed diurnal phase over land. The diurnal cycle analyses for the boreal spring and fall seasons are provided in Figs. S5–S7.
Harmonic dial plots of the amplitude (mm day−1) and phase (LST) of Fourier components, after vector averaging over each subregion based on 1) NASA, 2) X-SHiELD, 3) CLAUS, and 4) C384AM4 during JJA season (black dots) and DJF season (blue dots). Note the different radial scales among panels.
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0303.1
c. Duration, size, intensity, and their mutual dependence
This section examines several key MCS features, including duration, size, and intensity, all of which significantly influence the potential impacts of MCSs. For instance, long-lived MCSs often generate large accumulations of precipitation, increasing the risk of flooding events. Similarly, larger and more intense MCSs can cause more extensive damage to infrastructure, agriculture, and property, resulting in higher economic losses and posing greater challenges for preparedness, response, and recovery efforts. Figure 6 demonstrates the probability distribution function (PDF) of MCS duration for all subregions. The PDF is used instead of a density curve due to the discrete nature of MCS duration. Consistent with previous studies, the observed and simulated MCS duration PDFs show a similar monotonic decrease of probability with duration (Pope et al. 2008; Roca et al. 2017; Dong et al. 2021). Oceanic MCSs generally last longer than their continental counterpart. However, differences between NASA and CLAUS are noted over some subregions, particularly in short-duration bins. Both models capture the overall distribution, with the observed land–sea contrast evident in the simulations. Notably, X-SHiELD aligns more closely with observations across all subregions, except for the central United States. C384AM4 simulates more long-lived (>36 h) MCSs over the Indian Ocean and Maritime Continent.
Normalized histograms (%) of the duration of MCS for each subregion based on observations (black lines) and simulations (blue lines). Solid (dashed) lines are from CLAUS and C384AM4 (NASA and X-SHiELD). The numbers in each panel represent the mean (±1 s.d.) of MCS duration.
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0303.1
Figure 7 shows the MCS size density curve, which follows a gamma-shaped pattern across all subregions, peaking between 40 000 and 50 000 km2 with a long tail. The NASA dataset has a higher portion of smaller MCSs than CLAUS, resulting in smaller mean MCS sizes. This difference might be attributed, in part, to the original resolution of the source datasets (4 km for NASA vs 33 km for CLAUS), which have finer structures in NASA while smoother distributions for CLAUS. Despite these variations, larger MCS sizes are observed over oceans compared to land areas. Both models reasonably capture the gamma-shaped pattern and land–sea differences, with X-SHiELD aligning more closely with the NASA and C384AM4 with CLAUS, suggesting that original horizontal resolution impacts MCS size distribution.
Probability density distribution of MCS size (104 km2) for each subregion based on observations (black lines) and simulations (blue lines). Solid (dashed) lines are from NASA and X-SHiELD (CLAUS and C384AM4). The numbers in each panel represent the mean (±1 s.d.) of MCS size.
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0303.1
Figure 8 illustrates the density curve of MCS intensity. The observed intensity curves approach Gaussian distributions, with peak densities between 8 and 12 K except in the central United States, where it peaks around 5–7 K. Similar to the MCS size distribution, the NASA dataset has a higher peak with less intense MCS compared to CLAUS, resulting in larger mean intensities for CLAUS. In terms of model simulations, X-SHiELD reasonably captures the observed distribution, while C384AM4 tends to exhibit biases toward weaker intensities with a positively skewed distribution. Consequently, C384AM4 has a larger proportion of intense MCSs in most subregions and even shows a relatively flat distribution over the tropical Africa and Amazon region. Longer tails in C384AM4 help offset the skewness, resulting in comparable mean MCS intensities between the models.
As in Fig. 7, but for MCS intensity (K), measured by the difference between the minimum Tb of each MCS and 221 K (i.e., 221 − Tb).
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0303.1
Previous studies have shown that longer-lived and stronger MCSs tend to be larger in size. We investigate these relationships using a binning method. Figure 9 illustrates the dependence of MCS size on duration, with MCS sizes grouped into 10 duration bins (6–33 with a 3-h interval). Mean MCS size and relative change rates are calculated for each bin, revealing significant positive correlations between MCS size and duration in observational datasets, with size increasing by 1.6%–4.2% per 1-h increase in duration. Model simulations capture this relationship, particularly in the C384AM4 model, though correlations are weaker in X-SHiELD. The slightly weaker relationships noted in NASA and X-SHiELD over eastern Asia and the central United States may be attributed to smaller sample sizes in these areas. A similar method applied to MCS intensity and size (Fig. 10) also shows significant positive correlations, except for the central United States, with size increasing by 2.6%–5.3% per 1-K increase in intensity based on observational datasets. The range is 1.9%–5.3% for model simulations, with larger values observed in X-SHiELD. It is worth noting that both observations and model simulations indicate a nonlinear relationship, with size increasing more rapidly for more intense MCSs.
Relationship between MCS duration (h) and size (104 km2) for each subregion based on each MCS event. Individual events are binned based on duration with an interval of 3 h. The lines are the respective best linear fits. The normalized rates of change (% per hour) with the correlation coefficient and p value are listed.
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0303.1
As in Fig. 9, but for the relationship between MCS intensity (K) and size (104 km2).
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0303.1
d. Propagation speed and direction
This section discusses MCS propagation speed and direction, which are closely linked to the large-scale circulation. Figure 11 shows the observed and simulated PDF of MCS propagation speed. Similarly to MCS duration, the PDF is selected due to the dependence of propagation speed on the discrete time interval. Observations show that MCSs in eastern Asia, the Indian Ocean, and the Maritime Continent peak within 20–40 km h−1, while those in tropical Africa, central United States, and Amazon region peak around 40–60 km h−1. Overall, model simulations generally capture the observed distribution, albeit with a slight underestimation over tropical Africa and the Amazon region. X-SHiELD matches observations more closely than C384AM4.
As in Fig. 6, but for the MCS propagation speed (km h−1).
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0303.1
Figure 12 shows the density curve of MCS propagation direction. Midlatitude regions, such as eastern Asia and central United States, predominantly show eastward propagation (0°) due to prevailing westerlies. In contrast, MCSs in tropical and subtropical regions (e.g., tropical Africa and Amazon region) mainly propagate westward with a slight tilt toward north (−160°) or south (160°) and minor peaks in eastward propagation. Oceanic MCSs, such as those over the Indian Ocean and the Maritime Continent, also display a significant portion propagating westward with northward or southward tilts. But they also show another distinct peak in eastward propagation. The models generally capture the propagation directions well but inaccurately simulate a false secondary peak in westward propagation (160°) in eastern Asia and a higher proportion of eastward-propagating MCSs in tropical Africa, central United States, and Amazon region.
As in Fig. 7, but for the MCS propagation direction (°), measured clockwise from the east.
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0303.1
e. MCS-associated precipitation
MCSs are efficient rain-producing systems, significantly influencing the hydrological cycle and precipitation pattern. This section explores the statistical characteristics of MCS-related precipitation. Figure 13 shows its spatial distribution, which generally aligns with the annual mean pattern (Fig. S8) but differs in magnitude between observations and simulations. The area-weighted mean MCS-related precipitation is about 0.41 mm day−1 in GPM and 0.29 mm day−1 in MSWEP. The X-SHiELD model underestimates MCS precipitation at 0.19 mm day−1, while C384AM4 overestimates it at 0.57 mm day−1. Observations show hotspots with MCS-related precipitation ratios of 30%–50% in tropical Africa, parts of the Maritime Continent, and southern part of South America (Fig. 14), with global averages of 9.4% based on GPM and 7.1% based on MSWEP. These values are relatively smaller than those reported in previous studies, potentially due to the use of a 3-h dataset instead of a daily dataset and the stringent MCS tracking thresholds (Zhao 2022; Feng et al. 2023). The models exhibit significant discrepancies: X-SHiELD underestimates the global mean value by 5.6%, especially over the ocean, likely due to excessive total precipitation (Fig. S8), while C384AM4 overestimates it by about 6%, particularly in the identified hotspots. This overestimation aligns with the model’s positive biases in simulating MCSs.
As in Fig. 1, but for the mean MCS-related precipitation (mm day−1).
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0303.1
As in Fig. 1, but for the ratio of MCS-related precipitation to total precipitation (%).
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0303.1
On an event basis, the density curves of MCS-related precipitation across six subregions are shown in Fig. 15. Based on GPM data, a single peak typically occurs around 3–5 mm h−1. The MSWEP dataset shows similar structures but with smaller peaks, approximately 1–2 mm h−1. GPM also exhibits a broader distribution, with more precipitation exceeding 5 mm h−1. In model simulations, both models display gamma-shaped distributions, but their peaks are significantly lower than observations, averaging around 0.5 mm h−1, followed by a sharp decrease. While both models produce less intense precipitation than GPM, they generate more intense precipitation than MSWEP. As a result, the simulated mean MCS-related precipitation falls short compared to GPM but similar to MSWEP. Notably, the X-SHiELD simulation features a plateau structure (even an evident second peak in tropical Africa) after the peak, near 1–2 mm h−1. These event-based results, along with the overall MCS-related precipitation ratio, suggest that X-SHiELD underestimates MCS precipitation due to smaller precipitation associated with each event. Conversely, the overestimation in C384AM4 is primarily due to an excessive number of simulated MCS events, with individual event precipitation playing a smaller compensating role.
As in Fig. 7, but for the MCS-related precipitation (mm h−1).
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0303.1
4. Discussion and conclusions
In this study, we evaluated the global simulation of MCSs using two GFDL models: the ∼3.25-km resolution GSRM (X-SHiELD) and the ∼25-km high-resolution GCM (C384AM4). Simulations were compared with two satellite datasets to evaluate MCS features, including spatial distributions, seasonal/diurnal cycles, and event-based statistical characteristics like duration, size, intensity, and propagation speed/direction. We also explored the relationship among MCS size, duration, and intensity, as well as MCS-related precipitation. These analyses were conducted across six selected MCS hotspot regions to account for different geographical locations and distinct MCS features.
Our analysis suggests that both types of models are skillful at simulating many MCS features on a global scale, making them valuable for further exploration of MCS interaction with large-scale circulation, changes in MCS-related weather extremes, and responses to global warming. Yet each has its strengths and limitations. X-SHiELD performs better in replicating the observed nocturnal peak over land, while C384AM4 inaccurately simulating morning peaks. MCS-related precipitation patterns generally align between observations and model simulations, but their absolute values and contributions to total precipitation vary considerably. X-SHiELD notably underestimates MCS-related precipitation due to fewer simulated MCSs and lower precipitation rates associated with individual MCS, while C384AM4 overestimates it primarily due to a higher number of MCSs. Future works should address the weak and misrepresented diurnal cycle of MCSs over land, biases in MCS occurrence over oceans, and issues with MCS-associated precipitation. A detailed, process-level investigation of the physical mechanisms driving MCS genesis and upscale growth is essential. Additionally, more sensitivity tests on the model’s microphysics and convective parameterizations are needed to better understand and correct these biases.
Besides, these biases may partly stem from our MCS detection algorithm, which relies solely on Tb data and may misidentify cirrus or stratiform clouds as MCSs. Figure S9 shows the results by slightly perturbing the three thresholds in the tracking algorithm. As expected, relaxing the thresholds yields more MCS being tracked, while stricter thresholds reduce the tracked MCSs. Despite these changes, the spatial patterns remain highly consistent, suggesting that the Tb-related criteria in our tracking algorithm do not significantly affect our main findings. Besides, the effectiveness of incorporating surface precipitation into tracking methods is still debated (Leung et al. 2022; Zhao 2022; Feng et al. 2023; Hsu et al. 2023), as combing Tb and precipitation may introduce arbitrary threshold, reduce sample size, and weaken statistical robustness. Figure S10 presents the results of incorporating surface precipitation into the tracking. Excluding small precipitation events filters out more MCSs in C384AM4, as it simulates more MCSs with little or no precipitation, making the results appear more comparable to observations. However, we caution that applying such restrictions may obscure underlying model biases, leading to seemingly “correct” outcomes for the “wrong” reasons.
Furthermore, variations in Tb used for MCS tracking, along with differences in precipitation observations, may also contribute to these discrepancies. For example, CLAUS shows lower Tb values than NASA over most land regions (except parts of Africa and the Maritime Continent; Fig. S11) and over the central Pacific and the Indian Ocean, resulting in more MCSs tracked with CLAUS. Precipitation discrepancies are also evident when comparing MCS-related precipitation from NASA Tb with GPM and MSWEP (2020–21), with MSWEP showing higher precipitation despite similar totals (Fig. S12). This contrast is reflected in the density distribution of precipitation within each MCS (Fig. S13). Evaluating observational uncertainties is thus essential for future analyses.
Acknowledgments.
We thank Editor Kevin Reed and the three anonymous reviewers for their constructive feedback, which greatly improved the paper. We also thank Ilai Guendelman and Alex Kaltenbaugh for commenting on earlier versions of this paper. This research from the Geophysical Fluid Dynamics Laboratory is supported by NOAA’s Science Collaboration Program and administered by UCAR’s Cooperative Programs for the Advancement of Earth System Science (CPAESS) under Awards NA16NWS4620043 (to W. D.) and NA18NWS4620043B (to W. D.). Cheng is supported under Awards NA18OAR4320123, NA19OAR0220146, and NA19OAR0220147 from NOAA. Cheng and Zhou are additionally supported under the NOAA Research Global-Nest initiative.
Data availability statement.
The half-hourly NASA globally merged 4-km pixel-resolution infrared brightness temperature dataset can be downloaded from https://disc2.gesdisc.eosdis.nasa.gov/data/MERGED_IR/GPM_MERGIR.1/. The half-hourly NASA Integrated Multi-satellitE Retrievals for Global Precipitation Measurement (GPM/IMERG V07) final run product can be downloaded from https://gpm1.gesdisc.eosdis.nasa.gov/data/GPM_L3/GPM_3IMERGHH.07/. The brightness temperature datasets from the Cloud Archive User Service (CLAUS) product are available at https://data.ceda.ac.uk/badc/claus/data/. The Multi-Source Weighted-Ensemble Precipitation, version 2 (MSWEP V2), precipitation data can be found at https://gloh2o.org/mswep/. Public releases of SHiELD are accessible at https://github.com/NOAA-GFDL/SHiELD_build. The code used for the version of X-SHiELD in this paper is available in Harris et al. (2023). The AM4-MG2 source codes can be found at https://github.com/NOAA-GFDL/AM4/tree/MG2_xanadu_2020.02.01. All custom codes are direct implementation of standard methods and techniques, described in detail in materials and methods.
REFERENCES
Abel, P., and A. Gruber, 1979: An improved model for the calculation of longwave flux at 11 μm. NOAA Tech. Rep. NESS 106, NTIS PB80-119431, 24 pp.
Beck, H. E., and Coauthors, 2019: MSWEP V2 global 3-hourly 0.1° precipitation: Methodology and quantitative assessment. Bull. Amer. Meteor. Soc., 100, 473–500, https://doi.org/10.1175/BAMS-D-17-0138.1.
Bretherton, C. S., J. R. McCaa, and H. Grenier, 2004: A new parameterization for shallow cumulus convection and its application to marine subtropical cloud-topped boundary layers. Part I: Description and 1D results. Mon. Wea. Rev., 132, 864–882, https://doi.org/10.1175/1520-0493(2004)132<0864:ANPFSC>2.0.CO;2.
Chen, C.-C., J. H. Richter, C. Liu, M. W. Moncrieff, Q. Tang, W. Lin, S. Xie, and P. J. Rasch, 2021: Effects of organized convection parameterization on the MJO and precipitation in E3SMv1. Part I: Mesoscale heating. J. Adv. Model. Earth Syst., 13, e2020MS002401, https://doi.org/10.1029/2020MS002401.
Chen, G., and B. P. Kirtman, 2018: Long‐lived mesoscale convective systems of superparameterized CAM and the response of CAM. J. Adv. Model. Earth Syst., 10, 2269–2286, https://doi.org/10.1029/2018MS001339.
Cheng, K.-Y., and Coauthors, 2022: Impact of warmer sea surface temperature on the global pattern of intense convection: Insights from a global storm resolving model. Geophys. Res. Lett., 49, e2022GL099796, https://doi.org/10.1029/2022GL099796.
Dong, W., M. Zhao, Y. Ming, and V. Ramaswamy, 2021: Representation of tropical mesoscale convective systems in a general circulation model: Climatology and response to global warming. J. Climate, 34, 5657–5671, https://doi.org/10.1175/JCLI-D-20-0535.1.
Dong, W., M. Zhao, Y. Ming, and V. Ramaswamy, 2022: Significant increase in sea surface temperature at the genesis of tropical mesoscale convective systems. Geophys. Res. Lett., 49, e2022GL101950, https://doi.org/10.1029/2022GL101950.
Dong, W., M. Zhao, Y. Ming, J. P. Krasting, and V. Ramaswamy, 2023a: Simulation of United States mesoscale convective systems using GFDL’s new high-resolution general circulation model. J. Climate, 36, 6967–6990, https://doi.org/10.1175/JCLI-D-22-0529.1.
Dong, W., J. P. Krasting, and H. Guo, 2023b: Analysis of precipitation diurnal cycle and variance in multiple observations, CMIP6 models, and a series of GFDL-AM4.0 simulations. J. Climate, 36, 8637–8655, https://doi.org/10.1175/JCLI-D-23-0268.1.
Dong, W., M. Zhao, Z. Tan, and V. Ramaswamy, 2024a: Atmospheric rivers over eastern US affected by Pacific/North America pattern. Sci. Adv., 10, eadj3325, https://doi.org/10.1126/sciadv.adj3325.
Dong, W., M. Zhao, L. Harris, K.-Y. Cheng, L. Zhou, and V. Ramaswamy, 2024b: Contrasting response of mesoscale convective systems occurrence over tropical land and ocean to increased sea surface temperature. Geophys. Res. Lett., 51, e2024GL109251, https://doi.org/10.1029/2024GL109251.
Dong, W. H., Y. L. Lin, M. H. Zhang, and X. M. Huang, 2020: Footprint of tropical mesoscale convective system variability on stratospheric water vapor. Geophys. Res. Lett., 47, e2019GL086320, https://doi.org/10.1029/2019GL086320.
Donner, L. J., 1993: A cumulus parameterization including mass fluxes, vertical momentum dynamics, and mesoscale effects. J. Atmos. Sci., 50, 889–906, https://doi.org/10.1175/1520-0469(1993)050<0889:ACPIMF>2.0.CO;2.
Donner, L. J., C. J. Seman, R. S. Hemler, and S. M. Fan, 2001: A cumulus parameterization including mass fluxes, convective vertical velocities, and mesoscale effects: Thermodynamic and hydrological aspects in a general circulation model. J. Climate, 14, 3444–3463, https://doi.org/10.1175/1520-0442(2001)014<3444:ACPIMF>2.0.CO;2.
Ellingson, R. G., and R. R. Ferraro, 1983: An examination of a technique for estimating the longwave radiation budget from satellite radiance observations. J. Climate Appl. Meteor., 22, 1416–1423, https://doi.org/10.1175/1520-0450(1983)022<1416:AEOATF>2.0.CO;2.
Feng, Z., L. R. Leung, R. A. Houze Jr., S. Hagos, J. Hardin, Q. Yang, B. Han, and J. Fan, 2018: Structure and evolution of mesoscale convective systems: Sensitivity to cloud microphysics in convection-permitting simulations over the United States. J. Adv. Model. Earth Syst., 10, 1470–1494, https://doi.org/10.1029/2018MS001305.
Feng, Z., F. Song, K. Sakaguchi, and L. R. Leung, 2021a: Evaluation of mesoscale convective systems in climate simulations: Methodological development and results from MPAS-CAM over the United States. J. Climate, 34, 2611–2633, https://doi.org/10.1175/JCLI-D-20-0136.1.
Feng, Z., and Coauthors, 2021b: A global high-resolution mesoscale convective system database using satellite-derived cloud tops, surface precipitation, and tracking. J. Geophys. Res. Atmos., 126, e2020JD034202, https://doi.org/10.1029/2020JD034202.
Feng, Z., L. R. Leung, J. Hardin, C. R. Terai, F. Song, and P. Caldwell, 2023: Mesoscale convective systems in DYAMOND global convection‐permitting simulations. Geophys. Res. Lett., 50, e2022GL102603, https://doi.org/10.1029/2022GL102603.
Fiolleau, T., and R. Roca, 2013: Composite life cycle of tropical mesoscale convective systems from geostationary and low Earth orbit satellite observations: Method and sampling considerations. Quart. J. Roy. Meteor. Soc., 139, 941–953, https://doi.org/10.1002/qj.2174.
Goyens, C., D. Lauwaet, M. Schröder, M. Demuzere, and N. P. M. Van Lipzig, 2012: Tracking mesoscale convective systems in the Sahel: Relation between cloud parameters and precipitation. Int. J. Climatol., 32, 1921–1934, https://doi.org/10.1002/joc.2407.
Grabowski, W. W., and P. K. Smolarkiewicz, 1999: CRCP: A cloud resolving convection parameterization for modeling the tropical convecting atmosphere. Physica D, 133, 171–178, https://doi.org/10.1016/S0167-2789(99)00104-9.
Guendelman, I., and Coauthors, 2024: The precipitation response to warming and CO2 increase: A comparison of a global storm resolving model and CMIP6 models. Geophys. Res. Lett., 51, e2023GL107008, https://doi.org/10.1029/2023GL107008.
Guo, H., Y. Ming, S. Fan, L. Zhou, L. Harris, and M. Zhao, 2021: Two‐moment bulk cloud microphysics with prognostic precipitation in GFDL’s Atmosphere Model AM4.0: Configuration and performance. J. Adv. Model. Earth Syst., 13, e2020MS002453, https://doi.org/10.1029/2020MS002453.
Guo, H., Y. Ming, S. Fan, A. T. Wittenberg, R. Zhang, M. Zhao, and L. Zhou, 2022: Performance of two‐moment stratiform microphysics with prognostic precipitation in GFDL’s CM4.0. J. Adv. Model. Earth Syst., 14, e2022MS003111, https://doi.org/10.1029/2022MS003111.
Han, J., and C. S. Bretherton, 2019: TKE-based moist Eddy-Diffusivity Mass-Flux (EDMF) parameterization for vertical turbulent mixing. Wea. Forecasting, 34, 869–886, https://doi.org/10.1175/WAF-D-18-0146.1.
Han, J., W. Wang, Y. C. Kwon, S.-Y. Hong, V. Tallapragada, and F. Yang, 2017: Updates in the NCEP GFS cumulus convection schemes with scale and aerosol awareness. Wea. Forecasting, 32, 2005–2017, https://doi.org/10.1175/WAF-D-17-0046.1.
Harris, L., and Coauthors, 2020: GFDL SHiELD: A unified system for weather‐to‐seasonal prediction. J. Adv. Model. Earth Syst., 12, e2020MS002223, https://doi.org/10.1029/2020MS002223.
Harris, L., L. Zhou, A. Kaltenbaugh, S. Clark, K.-Y. Cheng, and C. Bretherton, 2023: A global survey of rotating convective updrafts in the GFDL X‐SHiELD 2021 global storm resolving model. J. Geophys. Res. Atmos., 128, e2022JD037823, https://doi.org/10.1029/2022JD037823.
Hartmann, D. L., H. H. Hendon, and R. A. Houze Jr., 1984: Some implications of the mesoscale circulations in tropical cloud clusters for large-scale dynamics and climate. J. Atmos. Sci., 41, 113–121, https://doi.org/10.1175/1520-0469(1984)041<0113:SIOTMC>2.0.CO;2.
Held, I. M., and Coauthors, 2019: Structure and performance of GFDL’s CM4.0 climate model. J. Adv. Model. Earth Syst., 11, 3691–3727, https://doi.org/10.1029/2019MS001829.
Hodges, K. I., D. W. Chappell, G. J. Robinson, and G. Yang, 2000: An improved algorithm for generating global window brightness temperatures from multiple satellite infrared imagery. J. Atmos. Oceanic Technol., 17, 1296–1312, https://doi.org/10.1175/1520-0426(2000)017<1296:AIAFGG>2.0.CO;2.
Houze, R. A., Jr., 2004: Mesoscale convective systems. Rev. Geophys., 42, RG4003, https://doi.org/10.1029/2004RG000150.
Houze, R. A., Jr., 2018: 100 years of research on mesoscale convective systems. A Century of Progress in Atmospheric and Related Sciences: Celebrating the American Meteorological Society Centennial, Meteor. Monogr., No. 59, Amer. Meteor. Soc., https://doi.org/10.1175/AMSMONOGRAPHS-D-18-0001.1.
Houze, R. A., Jr., K. L. Rasmussen, M. D. Zuluaga, and S. R. Brodzik, 2015: The variable nature of convection in the tropics and subtropics: A legacy of 16 years of the Tropical Rainfall Measuring Mission satellite. Rev. Geophys., 53, 994–1021, https://doi.org/10.1002/2015RG000488.
Hsu, W.-C., G. J. Kooperman, W. M. Hannah, K. A. Reed, A. A. Akinsanola, and A. G. Pendergrass, 2023: Evaluating mesoscale convective systems over the US in conventional and multiscale modeling framework configurations of E3SMv1. J. Geophys. Res. Atmos., 128, e2023JD038740, https://doi.org/10.1029/2023JD038740.
Hu, H., Z. Feng, and L.-Y. R. Leung, 2021: Linking flood frequency with mesoscale convective systems in the US. Geophys. Res. Lett., 48, e2021GL092546, https://doi.org/10.1029/2021GL092546.
Huang, X., C. Hu, X. Huang, Y. Chu, Y.-h. Tseng, G. J. Zhang, and Y. Lin, 2018: A long-term tropical mesoscale convective systems dataset based on a novel objective automatic tracking algorithm. Climate Dyn., 51, 3145–3159, https://doi.org/10.1007/s00382-018-4071-0.
Huffman, G. J., D. T. Bolvin, D. Braithwaite, K. Hsu, R. Joyce, P. Xie, and S. H. Yoo, 2015: NASA Global Precipitation Measurement (GPM) integrated multi-satellite retrievals for GPM (IMERG). Algorithm Theoretical Basis Document (ATBD) version, 4, 30 pp.
Janowiak, J. E., R. J. Joyce, and Y. Yarosh, 2001: A real-time global half-hourly pixel-resolution infrared dataset and its applications. Bull. Amer. Meteor. Soc., 82, 205–218, https://doi.org/10.1175/1520-0477(2001)082<0205:ARTGHH>2.3.CO;2.
Kooperman, G. J., M. S. Pritchard, and R. C. J. Somerville, 2013: Robustness and sensitivities of central U.S. summer convection in the super‐parameterized CAM: Multi‐model intercomparison with a new regional EOF index. Geophys. Res. Lett., 40, 3287–3291, https://doi.org/10.1002/grl.50597.
Laing, A. G., and J. M. Fritsch, 1997: The global population of mesoscale convective complexes. Quart. J. Roy. Meteor. Soc., 123, 389–405, https://doi.org/10.1002/qj.49712353807.
Laing, A. G., and J. M. Fritsch, 2000: The large-scale environments of the global populations of mesoscale convective complexes. Mon. Wea. Rev., 128, 2756–2776, https://doi.org/10.1175/1520-0493(2000)128<2756:TLSEOT>2.0.CO;2.
Lau, N.-C., and M. J. Nath, 2003: Atmosphere–ocean variations in the Indo-Pacific sector during ENSO episodes. J. Climate, 16, 3–20, https://doi.org/10.1175/1520-0442(2003)016<0003:AOVITI>2.0.CO;2.
Leung, L. R., and Coauthors, 2022: Exploratory precipitation metrics: Spatiotemporal characteristics, process-oriented, and phenomena-based. J. Climate, 35, 3659–3686, https://doi.org/10.1175/JCLI-D-21-0590.1.
Li, P., and Coauthors, 2023: Intensification of mesoscale convective systems in the East Asian rainband over the past two decades. Geophys. Res. Lett., 50, e2023GL103595, https://doi.org/10.1029/2023GL103595.
Lin, G., J. Fan, Z. Feng, W. I. Gustafson Jr., P.-L. Ma, and K. Zhang, 2019: Can the Multiscale Modeling Framework (MMF) simulate the MCS-associated precipitation over the central United States? J. Adv. Model. Earth Syst., 11, 4669–4686, https://doi.org/10.1029/2019MS001849.
Lin, G., C. R. Jones, L. R. Leung, Z. Feng, and M. Ovchinnikov, 2022: Mesoscale convective systems in a superparameterized E3SM simulation at high resolution. J. Adv. Model. Earth Syst., 14, e2021MS002660, https://doi.org/10.1029/2021MS002660.
Lin, Y., W. Dong, M. Zhang, Y. Xie, W. Xue, J. Huang, and Y. Luo, 2017: Causes of model dry and warm bias over central U.S. and impact on climate projections. Nat. Commun., 8, 881, https://doi.org/10.1038/s41467-017-01040-2.
Maddox, R. A., 1980: Mesoscale convective complexes. Bull. Amer. Meteor. Soc., 61, 1374–1387, https://doi.org/10.1175/1520-0477(1980)061<1374:MCC>2.0.CO;2.
Mapes, B. E., and R. A. Houze Jr., 1993: Cloud clusters and superclusters over the oceanic warm pool. Mon. Wea. Rev., 121, 1398–1416, https://doi.org/10.1175/1520-0493(1993)121<1398:CCASOT>2.0.CO;2.
Moncrieff, M. W., 2004: Analytic representation of the large-scale organization of tropical convection. J. Atmos. Sci., 61, 1521–1538, https://doi.org/10.1175/1520-0469(2004)061<1521:AROTLO>2.0.CO;2.
Moncrieff, M. W., C. Liu, and P. Bogenschutz, 2017: Simulation, modeling, and dynamically based parameterization of organized tropical convection for global climate models. J. Atmos. Sci., 74, 1363–1380, https://doi.org/10.1175/JAS-D-16-0166.1.
Ohring, G., A. Gruber, and R. Ellingson, 1984: Satellite determinations of the relationship between total longwave radiation flux and infrared window radiance. J. Climate Appl. Meteor., 23, 416–425, https://doi.org/10.1175/1520-0450(1984)023<0416:SDOTRB>2.0.CO;2.
Pollard, R. T., P. B. Rhines, and R. O. Thompson, 1973: The deepening of the wind-mixed layer. Geophys. Fluid Dyn., 4, 381–404, https://doi.org/10.1080/03091927208236105.
Pope, M., C. Jakob, and M. J. Reeder, 2008: Convective systems of the north Australian monsoon. J. Climate, 21, 5091–5112, https://doi.org/10.1175/2008JCLI2304.1.
Pope, M., C. Jakob, and M. J. Reeder, 2009: Objective classification of tropical mesoscale convective systems. J. Climate, 22, 5797–5808, https://doi.org/10.1175/2009JCLI2777.1.
Prein, A. F., R. Rasmussen, and G. Stephens, 2017: Challenges and advances in convection-permitting climate modeling. Bull. Amer. Meteor. Soc., 98, 1027–1030, https://doi.org/10.1175/BAMS-D-16-0263.1.
Prein, A. F., C. Liu, K. Ikeda, R. Bullock, R. M. Rasmussen, G. J. Holland, and M. Clark, 2020: Simulating North American mesoscale convective systems with a convection-permitting climate model. Climate Dyn., 55, 95–110, https://doi.org/10.1007/s00382-017-3993-2.
Roca, R., J. Aublanc, P. Chambon, T. Fiolleau, and N. Viltard, 2014: Robust observational quantification of the contribution of mesoscale convective systems to rainfall in the tropics. J. Climate, 27, 4952–4958, https://doi.org/10.1175/JCLI-D-13-00628.1.
Roca, R., T. Fiolleau, and D. Bouniol, 2017: A simple model of the life cycle of mesoscale convective systems cloud shield in the tropics. J. Climate, 30, 4283–4298, https://doi.org/10.1175/JCLI-D-16-0556.1.
Schär, C., C. Frei, D. Lüthi, and H. C. Davies, 1996: Surrogate climate-change scenarios for regional climate models. Geophys. Res. Lett., 23, 669–672, https://doi.org/10.1029/96GL00265.
Schumacher, R. S., and R. H. Johnson, 2006: Characteristics of U.S. extreme rain events during 1999–2003. Wea. Forecasting, 21, 69–85, https://doi.org/10.1175/WAF900.1.
Song, J., F. Song, Z. Feng, L. R. Leung, C. Li, and L. Wu, 2024: Realistic precipitation diurnal cycle in global convection-permitting models by resolving mesoscale convective systems. Geophys. Res. Lett., 51, e2024GL109945, https://doi.org/10.1029/2024GL109945.
Stevens, B., and Coauthors, 2019: DYAMOND: The DYnamics of the Atmospheric general circulation Modeled On Non-hydrostatic Domains. Prog. Earth Planet. Sci., 6, 61, https://doi.org/10.1186/s40645-019-0304-z.
Taylor, K. E., D. Williamson, and F. Zwiers, 2000: The sea surface temperature and sea-ice concentration boundary conditions for AMIP II simulations. Program for Climate Model Diagnosis and Intercomparison, Lawrence Livermore National Laboratory, University of California Rep. 60, 28 pp., https://pcmdi.llnl.gov/report/ab60.html.
Wei, Y., and Z. Pu, 2022: Diurnal cycle of precipitation and near-surface atmospheric conditions over the maritime continent: Land–sea contrast and impacts of ambient winds in cloud-permitting simulations. Climate Dyn., 58, 2421–2449, https://doi.org/10.1007/s00382-021-06012-3.
Yano, J.-I., and Coauthors, 2014: Basic concepts for convection parameterization in weather forecast and climate models: COST action ES0905 final report. Atmosphere, 6, 88–147, https://doi.org/10.3390/atmos6010088.
You, Z., Y. Deng, Y. Ming, and W. Dong, 2024: A multiscale assessment of the springtime U.S. mesoscale convective systems in the NOAA GFDL AM4. Climate Dyn., 62, 4017–4030, https://doi.org/10.1007/s00382-024-07114-4.
Yuan, J., and R. A. Houze Jr., 2010: Global variability of mesoscale convective system anvil structure from a-train satellite data. J. Climate, 23, 5864–5888, https://doi.org/10.1175/2010JCLI3671.1.
Zhao, M., 2020: Simulations of atmospheric rivers, their variability, and response to global warming using GFDL’s new high-resolution general circulation model. J. Climate, 33, 10 287–10 303, https://doi.org/10.1175/JCLI-D-20-0241.1.
Zhao, M., 2022: A study of AR-, TS-, and MCS-associated precipitation and extreme precipitation in present and warmer climates. J. Climate, 35, 479–497, https://doi.org/10.1175/JCLI-D-21-0145.1.
Zhao, M., and Coauthors, 2018a: The GFDL global atmosphere and land model AM4.0/LM4.0: 1. Simulation characteristics with prescribed SSTs. J. Adv. Model. Earth Syst., 10, 691–734, https://doi.org/10.1002/2017MS001208.
Zhao, M., and Coauthors, 2018b: The GFDL global atmosphere and land model AM4.0/LM4.0: 2. Model description, sensitivity studies, and tuning strategies. J. Adv. Model. Earth Syst., 10, 735–769, https://doi.org/10.1002/2017MS001209.
Zhou, L., S.-J. Lin, J.-H. Chen, L. M. Harris, X. Chen, and S. L. Rees, 2019: Toward convective-scale prediction within the next generation global prediction system. Bull. Amer. Meteor. Soc., 100, 1225–1243, https://doi.org/10.1175/BAMS-D-17-0246.1.
Zhou, L., and Coauthors, 2022: Improving global weather prediction in GFDL SHiELD through an upgraded GFDL cloud microphysics scheme. J. Adv. Model. Earth Syst., 14, e2021MS002971, https://doi.org/10.1029/2021MS002971.
