1. Introduction
In the tropics, we observe a wide variety of organization of deep cumulus convection, which emerges out of interactions among moisture, clouds, radiation, and circulations. The Madden–Julian oscillation (MJO; Madden and Julian 1972; Zhang 2005), a special type of the organized tropical convection, is characterized by its planetary-scale (i.e., wavenumber < 4), 30–60-day period with phase speed of ~5 m s−1 over the Indo-Pacific warm pool area, and eastward propagation. Fluctuations of moisture, clouds, and circulation in the tropics associated with the MJO are about or more than half of the total intraseasonal (i.e., 20–90-day time scale) variability of each parameter.
Our understanding of this dominant mode of the tropical intraseasonal variability is unsatisfactory in a sense that none of the existing theories of the phenomenon is able to explain all distinct features of it successfully. The limited understanding of the MJO, which is surprising given that it has been more than four decades since its discovery, reflects the lack of our understanding of the tropical atmosphere, especially the interactions among moisture, clouds, radiation, and circulations.
Partly originating from the deficient understanding, an accurate representation of the MJO is still a stringent test for many contemporary climate models (Lin et al. 2006; Kim et al. 2009; Hung et al. 2013). Regarding the profound impacts of the MJO on a wide variety of weather and climate phenomena (Lau and Waliser 2011; Zhang 2013), the lack of capability of climate models to adequately simulate the MJO challenges the reliability of simulation results made with them. The poor representation of the dominant tropical intraseasonal mode is also disturbing because it might indicate that some physical processes that are essential to the MJO are misrepresented or missing in the models. Given that all phenomena simulated by climate models emerge from complex interaction among all processes implemented in the models, any misrepresented or missing processes would affect not only the MJO but also all other phenomena that are influenced by the processes.
In this study, among the processes that have been suggested to be important in the dynamics of the MJO, we focus on the interplay of moisture, clouds, and radiation. An active phase of the MJO at a particular location can be depicted as a period of increased (i.e., positive deviation from climatology) tropospheric water vapor, clouds, and precipitation. The enhanced amount of water vapor and clouds reduces radiative cooling of the atmosphere by the “greenhouse effect.” In the heat budget, this reduced cooling means anomalous heating, which would be added to the condensational diabatic heating in the convection. Under the weak temperature gradient approximation (Sobel et al. 2001), the additional heating needs to be balanced with a stronger upward or weaker downward motion, which would act as additional moistening in the moisture budget. With the additional heating and moistening that come out of the cloud–moisture–radiation interaction, there would be a greater chance for the enhanced convection associated with the MJO to be maintained or grow if all other conditions are equal. In other words, the interaction between cloud–moisture and longwave (LW) radiation could provide a positive feedback to anomalous convection.
The role of the greenhouse feedback mechanism (GFM) in the dynamics of the MJO and tropical intraseasonal variability has been emphasized from different points of view. Theoretical works have suggested that the GFM is an essential ingredient in the dynamics of the MJO and tropical intraseasonal variability (Raymond 2001; Bretherton and Sobel 2002; Sobel and Gildor 2003; Fuchs and Raymond 2002, 2005, 2007; Bony and Emanuel 2005; Sobel and Maloney 2012, 2013). Raymond (2001) indicated based on results of an intermediate-complexity model of the tropical atmosphere that the GFM is a prerequisite for the existence of the MJO. Bony and Emanuel (2005) showed that the strength of the LW cloud–moisture–radiation feedback affects the growth rate and phase speed of tropical intraseasonal oscillation simulated in their idealized model of the tropical atmosphere.
GCM modeling studies have also stressed the role of the GFM in the simulation of the MJO (e.g., Zurovac-Jevtić et al. 2006; Kim et al. 2011; Landu and Maloney 2011a), supporting the arguments of the theoretical works. For example, Kim et al. (2011) turned off the cloud–radiation interaction in a GCM simulation by prescribing in all grid boxes the daily climatology of radiative heating, which is obtained from a control simulation where clouds and radiation are fully interactive. They found that the GCM loses its capability to simulate the MJO when the cloud–radiation interaction is disabled. Landu and Maloney (2011a) performed a similar experiment using a different GCM, and also concluded that the cloud–radiation interaction is a crucial process for the GCM they used to simulate a reasonable MJO. Unlike these studies, Lee et al. (2001) found that the cloud–radiation interaction weakens the tropical intraseasonal variability. Note that the phase speed of the eastward propagating wave in their aquaplanet simulations is closer to that of the convectively coupled Kelvin wave, rather than that of the MJO.
Recent studies of the moist static energy or moisture budget of the MJO represented either in global reanalysis products (Landu and Maloney 2011b; Kim et al. 2014a; Yokoi 2015) or in GCM simulations (Andersen and Kuang 2012; Arnold et al. 2013; Chikira 2014; Hannah and Maloney 2014) have further elucidated that the GFM is responsible for the maintenance and growth of the MJO. For example, Yokoi (2015) suggested by comparing three global reanalysis products that a reanalysis product that exhibits a relatively strong MJO simulates a greater LW cloud–moisture–radiation feedback process in the forecast model used in the data assimilation system. Chikira (2014) analyzed the moisture budget of an MJO simulated by a GCM, which showed that the additional diabatic heating from the GFM is responsible for moistening in the middle to lower troposphere that maintains and amplifies the moisture anomaly associated with the MJO.
Although the importance of the GFM in the MJO dynamics has been suggested and demonstrated in the above-mentioned studies, it has not been investigated how well the contemporary climate models simulate the process. In this study, we evaluate 29 climate models participating in phase 5 of the Coupled Model Intercomparison Project (CMIP5) with respect to the GFM. We then examine the statistical relationship between MJO simulation fidelity and effectiveness of the GFM to indicate how important the mechanism is in the MJO simulation. If a significant portion of the intermodel spread of MJO simulation capability can be explained by the effectiveness of the GFM, it will demonstrate that the GFM is an important process for the simulation of the MJO.
This paper is organized as follows. In section 2, data and methodology used will be described. The observed dependence of the GFM on precipitation anomaly and on the MJO cycle will be examined in section 3. A statistical relationship between the effectiveness of the GFM and MJO simulation capability of CMIP5 models will be investigated in section 4. Two IPSL models will be further analyzed in section 5. A summary and conclusion will be given in section 6.
2. Data and method
a. Observations and reanalysis products
We use 14-yr (1997–2010) observations of outgoing LW radiation (OLR) from the NOAA Advanced Very High Resolution Radiometer (AVHRR; Liebmann and Smith 1996), and precipitation from the Global Precipitation Climatology Project (GPCP; Huffman et al. 2001) and the Tropical Rainfall Measuring Mission (TRMM 3B42 version 6; Huffman et al. 2007) for validation of model simulations. The use of two precipitation datasets is to account for the uncertainty associated with the choice of the observation dataset. Note that another OLR product (Clouds and Earth’s Radiant Energy Systems; Loeb et al. 2009) was used to find that our results are not sensitive to the choice of the OLR dataset. Zonal winds in the lower (850 hPa) and upper (250 hPa) troposphere obtained from the NCEP–NCAR reanalysis (Kalnay et al. 1996) are employed together with the OLR dataset when assessing the MJO life cycle composite of various fields (Wheeler and Hendon 2004).
b. CMIP5 simulation dataset
Time series of daily OLR and precipitation are obtained from the CMIP5 archive (Taylor et al. 2012) for 29 CMIP5 models. We select a 20-yr (1985–2004) period from the twentieth-century historical run. For two versions of the IPSL models—IPSL-CM5A-LR and IPSL-CM5B-LR—zonal wind at 850- and 250-hPa levels, precipitable water, total cloud fraction, column liquid, and ice water content are also obtained for further analysis. Both observations and model simulations are daily-averaged and interpolated into a 2.5° × 2.5° grid before being used in any calculations. We limit our analysis to boreal winter months (November–April), when the MJO is relatively strong during its seasonal cycle. We excluded land area from our analysis because MJO variability is weak over the land because of a negligible heat capacity (Sobel et al. 2010).
Table 1 lists the CMIP5 models used in this study with references to the convection and cloud schemes of each model. All models except INM-CM4 use the mass-flux-type convection scheme. A wide variety of methods ranging from a diagnostic to a prognostic cloud fraction scheme are used as the cloud scheme. These two schemes are key players in the representation of the cloud–radiation interaction in GCMs. They determine for a given large-scale circulation regime the vertical distribution of clouds and their optical properties (e.g., Su et al. 2013) that influences vertical distribution of radiative heating and thereby affects the large-scale circulation. In the current multimodel study, we will not try to identify a scheme that is good for a decent simulation of the MJO. Rather, we will focus on a particular aspect of the cloud–radiation interaction (i.e., the GFM) simulated by a group of GCMs as a result of the complex interactions between the two schemes and others. Understanding the cloud–radiation interaction simulated by a GCM with respect to its convection and cloud scheme is beyond the scope of the current work.
Description of CMIP5 models used in this study. (Expansions of acronyms are available online at http://www.ametsoc.org/PubsAcronymList.)
c. Greenhouse enhancement factor
3. Observed GEF and its variation with the MJO cycle
Figure 1 shows the GEF and OLR anomalies as a function of rain rate anomaly together with frequency of rain rate anomaly events. Grid points over the Indo-Pacific warm pool area (15°S–15°N, 60°E–180°) are used in the calculations. We focus on observed features (thick black lines) in this section, and simulation results (thin gray lines) will be discussed later.
(a) GEF and (b) OLR anomaly (W m−2) as a function of precipitation anomaly (mm day−1). (c) Probability density (%) of precipitation anomaly (mm day−1). The thick black and thin gray lines represent observations and CMIP5 model simulations, respectively. In (a) and (b), the observed quantities are estimated using combinations of a single OLR dataset (AVHRR) and two precipitation datasets (GPCP: solid, TRMM: dashed). In (c), results from the GPCP and TRMM datasets are represented with solid and dashed line, respectively. Grid points over the central Indian Ocean (15°S–15°N, 60°E–180°) are used. Note that the precipitation anomaly is on a log scale.
Citation: Journal of Climate 28, 17; 10.1175/JCLI-D-14-00767.1
In observations, the GEF is positive and OLR anomaly is negative for all precipitation anomaly bins, indicating that anomalous water vapor and clouds associated with the positive precipitation anomaly trap LW radiation in the atmosphere, preventing it from going out to the space, and thereby provide additional heating to the atmosphere. In general, the GEF is greater when the precipitation anomaly is weaker and it decreases with increasing precipitation anomaly. When the precipitation anomaly is around 3 mm day−1, the observed GEF is about 0.3, indicating that the enhancement of column-integrated diabatic heating due to cloud–moisture–radiation interaction is about 30% of the anomalous condensational heating. The GEF is greater than 0.2 when the rain rate anomaly is weaker than 5 mm day−1 and it decreases and approaches about 0.04 for rain anomalies greater than 80 mm day−1. Overall features are insensitive to the choice of precipitation dataset, although the GEF is lower when TRMM is used for the strong precipitation anomalies greater than 10 mm day−1. This is consistent with the fact that the TRMM dataset exhibits a greater frequency of strong precipitation anomaly events (Fig. 1c). This nonuniform behavior of the observed GEF is in contrast to the constant GEF assumption made in many theoretical studies emphasizing the importance of the GEF (e.g., Sobel and Gildor 2003; Sobel and Maloney 2012). The fixed values used in the theoretical studies range from 0.1 to 0.25. Lin and Mapes (2004a) analyzed observational data and suggested 0.1–0.15, without separating rain anomaly regimes as we do in this study. Our results suggest that when a fixed value of 0.1–0.25 is used as a GEF parameter, the role of the GFM could be underestimated (overestimated) when rain rate anomaly is smaller than about 5 mm day−1 (greater than about 30 mm day−1).
Next we investigate whether the GEF varies with the life cycle of the MJO. If the GEF at a certain stage of the MJO cycle is preferentially greater than that in other stages, it might suggest that the GFM plays an important role in that stage. To distinguish different stages in the MJO cycle, the “phase” of the MJO is introduced following Wheeler and Hendon (2004). Wheeler and Hendon (2004) first create a combined field of 15°S–15°N averaged OLR, and zonal wind in 850 and 200 hPa, and perform an empirical orthogonal function (EOF) analysis with the combined field. They then use the principal components of the leading pair of EOFs to define the phase and amplitude of the MJO. Unlike Wheeler and Hendon (2004), whose primary objective was to develop an index for real-time monitoring and prediction and therefore avoided any prefiltering of time series, we used 20–100-day bandpass filtered data. Also, we use zonal wind at 250 hPa instead of 200 hPa because 200-hPa zonal wind is not available from the two IPSL models that will be used later. Only days with MJO amplitude greater than 1 are used in our calculation as in Wheeler and Hendon (2004). As a result, phases 2 and 3 indicate stages of enhanced convection over the area of interest (Indian Ocean), while phases 6 and 7 represent the suppressed period (see Fig. 3). Other phases (1, 4, 5, and 8) are transitioning stages between the enhanced and suppressed phases.
Figure 2 displays the percentage anomaly of the GEF from its mean value (Fig. 1a) at different MJO phases. The result shows that the GEF increases in the period of enhanced convection (phase 2 + 3, solid line) and decreases during suppressed phases (phase 6 + 7, dashed line). The magnitude of this change, in relative terms, is greater in the weaker precipitation anomaly regime, where the mean GEF is greater (Fig. 1a). The fact that the GEF of the weak precipitation anomaly events systematically varies with the MJO phase and that its magnitude is preferentially greater during active phases suggests that those events might play an important role in the MJO evolution through the GFM, especially during the phase of enhanced convection. Hereafter we focus on precipitation anomalies with magnitude of 1–5 mm day−1, which we will refer to as weak precipitation anomalies. The two boundary values are determined rather subjectively with the following considerations: 1) when rain rate anomaly is too small, it would be difficult to ensure the causality relationship of condensational heating with radiative heating assumed in the GFM; 2) there is uncertainty in satellite estimation of precipitation that might not allow us to discuss O(0.1) mm day−1 changes of precipitation anomaly seriously (e.g., Berg et al. 2010); and 3) the mean GEF and changes of the GEF with respect to MJO phase become weaker as rain rate anomaly increases. Note that our results presented below are not critically sensitive to the choice of the two boundary values.
Percentage change of the GEF at different MJO phases: phases 1–8 (Figs. 3a–h). Solid line represents phases 2 and 3 and dotted line shows phases 6 and 7. Results of phases 4 + 5 and 8 + 1 are in between these lines and are not shown.
Citation: Journal of Climate 28, 17; 10.1175/JCLI-D-14-00767.1
Figure 3 illustrates the evolution of the GEF averaged over 1–5 mm day−1 bins during a full life cycle of the MJO. The composite OLR anomaly is also presented to show the relationship between the status of local convection and the magnitude of GEF of weak precipitation anomaly events. A negative OLR anomaly, which indicates anomalously enhanced convection, starts to develop over the western Indian Ocean in phase 1. The Indian Ocean experiences development of the negative OLR anomaly in phases 2 and 3. The developed anomaly propagates toward the east in subsequent phases until it weakens near the date line in phases 8 and 1. Over the Indo-Pacific warm pool area, the GEF for the weak precipitation anomaly events systematically increases (decreases) when MJO-associated convection is active (suppressed). Figure 4 summarizes this relationship between OLR and GEF anomaly for the Indian Ocean and the western Pacific. This suggests that in the period of active convection, the same amount of precipitation anomaly accompanies greater amounts of water vapor and/or cloud anomalies, leading to a greater feedback to the original precipitation anomaly.
MJO life cycle composite of anomalous OLR (shaded, W m−2) and the GEF for the weak precipitation anomaly events (solid contour, unitless). Both OLR and GEF anomalies are averaged over 15°S–15°N.
Citation: Journal of Climate 28, 17; 10.1175/JCLI-D-14-00767.1
MJO life cycle composite of anomalous OLR (blue, W m−2) and GEF for the weak precipitation anomaly events (black, unitless) averaged over the (a) Indian Ocean (15°S–15°N, 60°–90°E) and (b) western Pacific (15°S–15°N, 150°E–180°) region.
Citation: Journal of Climate 28, 17; 10.1175/JCLI-D-14-00767.1
Figures 3 and 4 show that the GEF of the weak precipitation anomaly events increases (decreases) with an increase (decrease) of the MJO-associated convective anomaly. This systematic relationship suggests that the weak precipitation anomaly events are probably related to the stratiform stage in the life cycle of tropical convective systems. That is, within the envelope of the large-scale convective anomaly, convective systems produces stronger, or optically thicker, anvil clouds, thus enhancing the GEF. The strengthening of the stratiform stage could be a result of enhanced environmental moisture (Schumacher and Houze 2006).
4. GEF simulated in CMIP5 models and its relationship with MJO simulation capability
In this section, we turn our attention to CMIP5 simulations and examine how well the CMIP5 models simulate the observed GEF, and investigate the relationship the GEF has with the MJO simulation capability of the models. The CMIP5 models overall underestimate the GEF, especially in relatively weak (<5 mm day−1) rain rate anomaly bins (Fig. 1a). When averaged across all models, the multimodel mean GEF is about 60%–90% of the observed value in 1–5 mm day−1 bins (Fig. 5). Also, there is a large scatter among simulated values of GEF; the highest GEF is about or more than 2 times the lowest value in almost all rain anomaly bins (Fig. 1a). This suggests that moisture and cloud anomalies simulated in CMIP5 models might be weaker than observed, in particular in weak precipitation anomaly regimes. It is interesting to see that where the models perform better is where the frequency of rain rate anomaly events is greater (Fig. 1c). In other words, the variation of moisture/cloud associated with convection in models is similar to that of observations for the events that are most frequent, but the models systematically underestimate the fluctuations when rain anomaly becomes smaller.
Percentage ratio of multimodel mean GEF to observed values (solid: GPCP, dashed: TRMM). The GEF is calculated over the Indo-Pacific warm pool area (15°S–15°N, 60°E–180°).
Citation: Journal of Climate 28, 17; 10.1175/JCLI-D-14-00767.1
If the GFM is important in the simulation of the MJO, the scatter among models in the GEF (Fig. 1a) should have a tight relationship with that of some MJO characteristics. Here we test this possibility by employing scalars that represents the MJO simulation capability of the models. To derive a scalar that indicates the strength of the simulated MJO (SMJO), we first calculate the wavenumber–frequency power spectrum of equatorial (i.e., 10°S–10°N averaged) precipitation anomalies. The term SMJO is derived from the spectrum to represent how dominant eastward propagation is over the MJO band, as the sum of eastward propagating power within the MJO band (i.e., zonal wavenumbers 1–3 and periods 30–60 days) divided by its westward propagating counterpart. The SMJO has often been used in previous studies to measure the fidelity of climate models to represent the MJO (Lin et al. 2006; Kim et al. 2009; Sperber and Kim 2012; Hung et al. 2013; Kim et al. 2014b; Benedict et al. 2014). Another scalar metric, which represents a dominant period of the eastward propagating signal (PMJO), is derived from the eastward propagating components of the power spectrum by performing a power-weighted average of the period over wavenumbers 1–3 and for periods of 20–100 days. Figure 6 shows SMJO and PMJO obtained from each model and observations. Observed SMJO is about 2.56, while the CMIP5 climate models exhibit SMJO ranging from 0.70 to 3.74. There are only 5 among 29 models that exhibit SMJO greater than the observed value (Fig. 6a). The CMIP5 climate models exhibit PMJO that spans from 30 to 45 days, and the observed value (~39 days) is in the middle of model scatter.
Metrics that represent characteristic of the observed (red bar) and simulated (blue bars) MJO: (a) SMJO, which is the sum of power within the MJO band (eastward propagating, zonal wavenumbers 1–3, and periods 30–60 days) divided by its westward propagating counterpart; and (b) PMJO, which is the power-weighted average of period. The averaging is performed over the extended MJO band (eastward propagating, zonal wavenumbers 1–3 and periods 20–100 days).
Citation: Journal of Climate 28, 17; 10.1175/JCLI-D-14-00767.1
Using the scalar metrics (SMJO and PMJO) obtained from each model, we investigate the statistical relationship between the GEF and the MJO simulation capability. If the intermodel spread of the GEF in all rain rate anomaly bins is tightly coupled to the intermodel spread of the metrics, it would mean that the GEF is in general a crucial process for the dynamics of the simulated MJO. If the relationship is stronger in some particular precipitation anomaly regimes than in other regimes, it might suggest that those regimes are particularly important.
Figure 7 shows linear correlation coefficients of SMJO (red) and PMJO (blue) with the GEF in each precipitation anomaly bin. The GEF exhibits a positive linear relationship with SMJO in all precipitation anomaly bins, indicating that the MJO is stronger in models that simulate a stronger GEF. Interestingly, this linear relationship is strong particularly in the rain rate anomaly bins that are weaker than about 5 mm day−1. The correlation coefficient in this rain rate anomaly regime is between 0.4 and 0.6, which is statistically significant at the 5% significance level. When one model that is an outlier in the scatter of the GEF and SMJO (CNRM-CM5; Fig. 8) is excluded in the calculation, correlation coefficients become even higher (dashed line in Fig. 7). For rain rate anomalies greater than 5 mm day−1, correlation decreases with increase of rain rate anomaly until rain rate anomaly becomes 10–20 mm day−1, and it increases and exhibits a secondary peak near 70 mm day−1. Note that this secondary peak is not statistically significant at the 5% significance level. In the case of PMJO, the correlation coefficient monotonically increases with the rain rate anomaly and shows a statistical relationship that is significant at the 95% confidence level in the high rain rate anomaly regimes (>30 mm day−1).
Correlation coefficients of SMJO (red) and PMJO (blue) with the GEF as a function of rain rate anomaly. The correlation coefficients are obtained using CMIP5 model simulations. Solid lines represent results obtained using all 29 models, while dashed lines show results obtained without one outlier model (CNRM-CM5 for SMJO and MIROC4h for PMJO). The GEF is calculated over the Indo-Pacific warm pool area (15°S–15°N, 60°E–180°). The gray dotted line indicates the 95% confidence level.
Citation: Journal of Climate 28, 17; 10.1175/JCLI-D-14-00767.1
(a) Scatterplot of GEF, averaged over 1–5 mm day−1 precipitation anomaly bin, with SMJO. (b) As in (a), except that the averaged GEF is obtained from relatively strong rain rate anomaly regimes (30–60 mm day−1) and PMJO. The closed circles indicate CMIP5 models and the open circle and open square represents GPCP and TRMM, respectively. Correlation coefficients between two metrics derived from model simulations are shown in each panel. The number in parentheses is the correlation coefficient obtained without an outlier model [CNRM-CM5 in (a) and MIROC4h in (b)]. The GEF is calculated over the Indo-Pacific warm pool area (15°S–15°N, 60°E–180°).
Citation: Journal of Climate 28, 17; 10.1175/JCLI-D-14-00767.1
Figure 8 displays scatterplots of SMJO and PMJO with the GEF averaged over the weak (1–5 mm day−1) and strong (30–50 mm day−1) rain rate anomalies, respectively. When averaging the GEF over rain anomaly bins, we weighted the GEF by frequencies of precipitation anomaly bins. The two observation datasets exhibits similar GEF for the weak precipitation anomaly regime (~0.32), whereas when averaged over strong precipitation anomaly regime the GPCP value (~0.085) is noticeably greater than that of TRMM (~0.05). This again suggests that uncertainty of the GEF in the strong precipitation anomaly regimes is greater than that of the weak precipitation anomaly regimes. In the weak precipitation anomaly regime, only seven models simulate the GEF comparable to or higher than 0.3, while most models underestimate it. And the scatter in the models is well represented by the linear fit between SMJO and the GEF among the models, demonstrating the robustness of the statistical relationship. CNRM-CM5 exhibits much stronger SMJO than what can be inferred from its GEF and the linear relationship between GEF and SMJO from other models. When this model is removed from the model set, the correlation increases from about 0.58 (R2 = 0.34) to 0.67 (R2 = 0.45), which are both statistically significant at the 1% significance level. This means that the intermodel spread of GEF in the weak rain rate anomaly bins explains about 45% of the intermodel spread of SMJO. In the strong precipitation anomaly regime, the intermodel spread of GEF explains about 29% of intermodel variability in PMJO, when an outlier model in the scatter (MIROC4h) is not considered. This suggests that, although the relationship of the GEF in the strong precipitation anomaly regime with PMJO is statistically significant at 95% confidence level, there are other factors that influence the dominant period of eastward propagating signal in precipitation field.
5. IPSL-CM5A-LR versus IPSL-CM5B-LR
Interestingly, two models from L’Institut Pierre-Simon Laplace (IPSL) des sciences de l’environnement—IPSL-CM5A-LR and IPSL-CM5B-LR—exhibit a considerable difference in their simulation of the MJO and GEF; IPSL-CM5B-LR simulates a stronger MJO and a greater GEF for 1–5 mm day−1 rain rate anomaly events than those of IPSL-CM5A-LR (Fig. 10). Therefore, it is worthwhile trying to understand the difference of the GEF represented in the two IPSL models. The two models are different essentially in their representation of boundary layer turbulence, how it triggers deep convection and regulates its mass flux, and how the resulting clouds are determined (Hourdin et al. 2013). This is accomplished in IPSL-CM5B-LR by predicting thermal plumes and cold pools in addition to smaller-scale turbulence. Hourdin et al. (2013) documented that IPSL-CM5B-LR simulates a much stronger intraseasonal variability of precipitation and a more realistic MJO than IPSL-CM5A-LR, which is consistent with our results. This provides us an opportunity to further investigate the physical mechanism behind the difference between the two IPSL models in simulating GEF. Figure 9 shows that the better MJO model, IPSL-CM5B-LR, simulates well the relationship between the MJO-associated convection and the GEF of the weak precipitation anomaly regime over the Indo-Pacific region. This suggests that the role the GFM plays in the evolution of the MJO cycle in this model is similar to that in observations.
As in Fig. 3, but for IPSL-CM5B-LR.
Citation: Journal of Climate 28, 17; 10.1175/JCLI-D-14-00767.1
Figure 10 provides clues for why IPSL-CM5B-LR simulates a greater GEF in the weak precipitation anomaly regime. IPSL-CM5B-LR exhibits a precipitable water anomaly in this regime that is greater than 2 times what IPSL-CM5A-LR shows. Anomalous cloud cover is also much larger in IPSL-CM5B-LR than in IPSL-CM5A-LR. The greater amount of water vapor anomaly in the column and the larger coverage of cloud anomaly should have a greater greenhouse effect. The two models show a comparable amount of ice water path anomaly, and liquid water path anomaly is greater in IPSL-CM5A-LR, implying that the ice and liquid water path anomalies are not responsible for the difference in GEF of the two models. Figure 11 shows that the vertical velocity anomaly simulated by IPSL-CM5B-LR in the 1–5 mm day−1 rain rate anomaly regime is greater than that in IPSL-CM5A-LR. This might suggest that, with a greater GEF, IPSL-CM5B-LR simulates a greater diabatic heating, which would induce a greater upward motion in the weak rain rate anomaly regime. The greater upward motion and accompanying vertical moisture advection would increase the amount of moisture in the lower troposphere, thereby providing a favorable condition for the initial anomalies to further develop. The results suggest that parameterization of moist physics in a climate model changes the way convection interacts with water vapor, clouds, and radiation, and it could affect the simulated GEF significantly, and the organization of large-scale convection.
(a) SMJO of the two IPSL models. (b) The GEF; and anomalous (c) precipitable water (mm), (d) cloud fraction (%), (e) ice cloud water path (mm), and (f) liquid cloud water path (mm) for the weak precipitation anomaly events of the two IPSL models. Blue and red bars represent IPSL-CM5A-LR and IPSL-CM5B-LR, respectively. The GEF is calculated over the Indo-Pacific warm pool area (15°S–15°N, 60°E–180°).
Citation: Journal of Climate 28, 17; 10.1175/JCLI-D-14-00767.1
Vertical profiles of pressure velocity (hPa s−1) anomaly averaged over 1–5 mm day−1 rain rate anomaly bins. Two IPSL models, IPSL-CM5A-LR (blue) and IPSL-CM5B-LR (red), are used.
Citation: Journal of Climate 28, 17; 10.1175/JCLI-D-14-00767.1
6. Summary and conclusions
The positive feedback process between tropical cumulus convection and large-scale clouds and moisture, and its role in the dynamics of the MJO are investigated using observations and CMIP5 simulations. Through the greenhouse effect, large-scale clouds and moisture that increase with enhancement of tropical cumulus convection give favorable conditions for convection to grow by providing radiative heating and upward motion that balances the additional radiative heating. The effectiveness of this positive process, the greenhouse feedback mechanism (GFM), is measured for different positive rain rate anomaly regimes as a ratio of negative OLR anomaly, which we regard as an estimation of anomalous longwave radiative heating in the atmosphere, to precipitation anomaly.
Results of observational analysis show that the ratio, which is referred to as the greenhouse enhancement factor (GEF), varies vastly depending on precipitation anomaly. Overall, the GEF is greater in the regime of weaker precipitation anomaly and monotonically decreases as rain rate anomaly increases. The GEF of 2 mm day−1 precipitation anomaly events (about 1) is an order of magnitude greater than that of 20 mm day−1 events (about 0.1). In many theoretical studies of the MJO, however, it is assumed that this parameter is a constant, ranging from 0.1 to 0.25. According to our results, the role of the GFM may have been underestimated in the theoretical works that used a fixed number for the GEF, especially when rain rate anomaly is weak. Relaxation of this constant GEF assumption might yield solutions that are substantially different from those with the assumption. Further works are desired along this line. The GEF also varies with the cycle of the MJO, having greater values during the phases of the MJO with enhanced convection. Weak precipitation anomaly regimes exhibit particularly strong sensitivity of the GEF to the MJO phase, suggesting that the rain anomaly events in this regime play important role during the enhanced phase of the MJO.
A statistical relationship between the effectiveness of GFM and the MJO simulation fidelity of the CMIP5 models is investigated using 29 climate model simulations participating in CMIP5. Two scalars that characterize specific aspects of the MJO (SMJO: strength, PMJO: period) of each model are obtained using the space–time power spectra of equatorial precipitation, and the statistical relationship between them and the GEF is examined. Results show that the GEF has a robust statistical relationship with both scalars. In the case of SMJO, the relationship peaks at weak (<5 mm day−1) precipitation anomaly regimes, while PMJO is better correlated with the GEF in the strong (>30 mm day−1) precipitation anomaly regime. When the mean state is weakly perturbed in the relatively strong-GEF models, the GFM would push the perturbation further from the mean state by providing a greater LW radiative heating. This means that the GFM makes the mean state less stable in the strong-GEF models compared to that of weak-GEF models. A large-scale organization of tropical convection, such as the MJO, could possibly be preferred in those models.
Two IPSL models that differ from each other essentially in their representation of turbulence and its interaction with convection and clouds are used to gain further insights about possible cause of the different GEF simulations. IPSL-CM5B-LR simulates a GEF in the weak rain rate anomalies (between 1 and 5 mm day−1) that is much stronger than in IPSL-CM5A-LR, and also has a stronger MJO. Compared to IPSL-CM5A-LR, IPSL-CM5B-LR produces a greater amount of column water vapor and cloud fraction anomaly, and a stronger anomalous upward motion in the weak precipitation anomaly regime. With the stronger GEF associated with the greater amount of column water vapor and cloud fraction anomaly, the atmosphere of IPSL-CM5B-LR would feel a greater total diabatic heating, which needs to be balanced by a stronger upward motion in the tropics where any horizontal buoyancy gradient is quickly removed by the gravity waves. This stronger upward motion might provide a more favorable condition for the rain rate anomaly to grow or be maintained through vertical advection of moisture. In this way, the GEF process, especially that in weak precipitation anomaly regimes, could be an important process in the simulation of the MJO.
Although the results presented in Figs. 10 and 11 suggests that the GFM is an important mechanism in the simulation of the MJO at least in IPSL models, one should be cautious in making conclusions about the role of the GFM in the MJO. The GFM is certainly one possible mechanism that could be responsible for the difference between the two IPSL models, but we cannot rule out possibilities of the existence of other mechanisms that also could cause the difference. To address this issue, we investigated other mechanisms that have been suggested to be important in the simulation of the MJO. Our investigation of the possible mechanisms that might have affected the simulation of the MJO reveals that the two IPSL models exhibit a similar surface latent heat flux feedback process and a similar normalized gross moist stability (not shown). This suggests that the surface heat flux feedback mechanism (Sobel et al. 2010) and the gross moist instability mechanism (Raymond and Fuchs 2009) are not responsible for the difference between the two IPSL models. On the other hand, we found that the two models exhibit a quite different moisture sensitivity of convection. Kim et al. (2014b) demonstrated that the moisture sensitivity of convection, measured as the lower-tropospheric relative humidity difference between strong (upper 10% events) and weak (lower 20% events) precipitation regimes, has a strong relationship with model’s MJO simulation fidelity. IPSL-CM5A-LR and IPSL-CM5B-LR exhibit a moisture sensitivity metric of 33.2% and 42.9%, respectively. Further study is warranted to understand the relationship between and relative role of the GFM and the moisture sensitivity of convection.
Del Genio et al. (2015) recently conducted a set of reforecast experiments of an observed MJO event with a GCM using different configurations of its convection scheme. They found that versions that simulate relatively strong greenhouse feedback process in the weak rain rate anomaly regime better forecast the onset and evolution of the MJO event. The results of Del Genio et al. (2015) support our argument that the GFM, especially when accompanied with weak precipitation anomaly events, plays an important role in the simulation of the MJO.
Most climate models participating in CMIP5 systematically underestimate the strength of the GEF process, especially in relatively weak (<5 mm day−1) rain rate anomaly regimes. There are a few possible model deficiencies that possibly influence the GEF either directly or indirectly. A bias in the vertical structure of diabatic heating or the shallow-to-deep transition in the cloud life cycle could affect the level of detrainment, and thereby degrade the cloud–radiation interaction. Both processes have been suggested to be important in the simulation of the MJO (Li et al. 2009; Zhang and Song 2009).
The lack of representation of the mesoscale convective organization (Tobin et al. 2013) and the role of wind shear in the horizontal cloud distribution could also affect the simulation of the GEF. Lin and Mapes (2004b) showed that in observations vertical wind shear strongly affects the GEF: for a given amount of precipitation, reduction of OLR is greater when vertical wind shear is stronger. This is because a greater anvil cloud fraction, and therefore a greater cloud–radiation feedback, is preferred when vertical wind shear is stronger through its impacts on horizontal extent of the anvil clouds, and perhaps also because of the impact of shear on the mesoscale organization of convective systems that sustains these clouds. Lin and Mapes (2004b) also showed that a climate model (Community Atmospheric Model, version 2) was not able to simulate the association of the GEF with vertical wind shear. As pointed out in Lin and Mapes (2004b), vertical wind shear is not accounted for in most of the cumulus and stratiform cloud schemes. Our results suggest that a further investigation of the relationship between vertical wind shear and cloud–radiation interaction using observations as well as state-of-the-art comprehensive models (e.g., CMIP5), and development of cloud parameterizations for GCMs that incorporate the effects of vertical wind shear on cloud fraction, needs to be done.
Acknowledgments
We thank Adam H. Sobel for his thoughtful comments on an earlier version of the manuscript. D. Kim was supported by the NASA Grant NNX13AM18G and the Korea Meteorological Administration Research and Development Program under Grant CATER 2013-3142. M.-S. Ahn and I.-S. Kang were supported by a National Research Foundation of Korea (NRF) grant funded by the Korean government (MEST; NRF-2012M1A2A2671775) and by the Brain Korea 21 Plus. A. D. Del Genio was supported by the NASA Modeling and Analysis and Precipitation Measurement Mission Programs.
REFERENCES
Andersen, J. A., and Z. Kuang, 2012: Moist static energy budget of MJO-like disturbances in the atmosphere of a zonally symmetric aquaplanet. J. Climate, 25, 2782–2804, doi:10.1175/JCLI-D-11-00168.1.
Anderson, J. L., and Coauthors, 2004: The new GFDL global atmosphere and land model AM2-LM2: Evaluation with prescribed SST simulations. J. Climate, 17, 4641–4673, doi:10.1175/JCLI-3223.1.
Arakawa, A., and W. H. Schubert, 1974: Interaction of a cumulus cloud ensemble with the large-scale environment, Part I. J. Atmos. Sci., 31, 674–701, doi:10.1175/1520-0469(1974)031<0674:IOACCE>2.0.CO;2.
Arnold, N. P., Z. Kuang, and E. Tziperman, 2013: Enhanced MJO-like variability at high SST. J. Climate, 26, 988–1001, doi:10.1175/JCLI-D-12-00272.1.
Benedict, J. J., E. D. Maloney, A. H. Sobel, and D. M. W. Frierson, 2014: Gross moist stability and MJO simulation skill in three full-physics GCMs. J. Atmos. Sci., 71, 3327–3349, doi:10.1175/JAS-D-13-0240.1.
Berg, W., T. L’Ecuyer, and J. M. Haynes, 2010: The distribution of rainfall over oceans from spaceborne radars. J. Appl. Meteor. Climatol., 49, 535–543, doi:10.1175/2009JAMC2330.1.
Betts, A. K., 1986: A new convective adjustment scheme. Part I. Observational and theoretical basis. Quart. J. Roy. Meteor. Soc., 112, 677–691, doi:10.1002/qj.49711247307.
Bony, S., and K. A. Emanuel, 2001: A parameterization of the cloudiness associated with cumulus convection; evaluation using TOGA COARE data. J. Atmos. Sci., 58, 3158–3183, doi:10.1175/1520-0469(2001)058<3158:APOTCA>2.0.CO;2.
Bony, S., and K. A. Emanuel, 2005: On the role of moist processes in tropical intraseasonal variability: Cloud–radiation and moisture–convection feedbacks. J. Atmos. Sci., 62, 2770–2789, doi:10.1175/JAS3506.1.
Bougeault, P., 1985: A simple parameterization of the large-scale effects of cumulus convection. Mon. Wea. Rev., 113, 2108–2121, doi:10.1175/1520-0493(1985)113<2108:ASPOTL>2.0.CO;2.
Bretherton, C. S., and A. H. Sobel, 2002: A simple model of a convectively coupled Walker circulation using the weak temperature gradient approximation. J. Climate, 15, 2907–2920, doi:10.1175/1520-0442(2002)015<2907:ASMOAC>2.0.CO;2.
Chikira, M., 2014: Eastward-propagating intraseasonal oscillation represented by Chikira–Sugiyama cumulus parameterization. Part II: Understanding moisture variation under weak temperature gradient balance. J. Atmos. Sci., 71, 615–639, doi:10.1175/JAS-D-13-038.1.
Chikira, M., and M. Sugiyama, 2010: A cumulus parameterization with state-dependent entrainment rate. Part I: Description and sensitivity to temperature and humidity profiles. J. Atmos. Sci., 67, 2171–2193, doi:10.1175/2010JAS3316.1.
Del Genio, A. D., J. Wu, A. B. Wolf, Y. Chen, M.-S. Yao, and D. Kim, 2015: Constraints on cumulus parameterization from simulation of observed MJO events. J. Climate,28, 6419–6442, doi:10.1175/JCLI-D-14-00832.1.
Derbyshire, S. H., A. V. Maidens, S. F. Milton, R. A. Stratton, and M. R. Willett, 2011: Adaptive detrainment in a convective parametrization. Quart. J. Roy. Meteor. Soc., 137, 1856–1871, doi:10.1002/qj.875.
Donner, L. J., 1993: A cumulus parameterization including mass fluxes, vertical momentum dynamics, and mesoscale effects. J. Atmos. Sci., 50, 889–906, doi:10.1175/1520-0469(1993)050<0889:ACPIMF>2.0.CO;2.
Donner, L. J., C. J. Seman, R. S. Hemler, and S. 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, doi:10.1175/1520-0442(2001)014<3444:ACPIMF>2.0.CO;2.
ECMWF, 2004: IFS Documentation CY28r1, Part IV, chapter 6 (Clouds and large-scale precipitation). European Centre for Medium-Range Weather Forecasts. [Available online at https://software.ecmwf.int/wiki/display/IFS/CY28R1+Official+IFS+Documentation.]
Emanuel, K. A., 1991: A scheme for representing cumulus convection in large-scale models. J. Atmos. Sci., 48, 2313–2329, doi:10.1175/1520-0469(1991)048<2313:ASFRCC>2.0.CO;2.
Emori, S., T. Nozawa, A. Numaguti, and I. Uno, 2001: Importance of cumulus parameterization for precipitation simulation over East Asia in June. J. Meteor. Soc. Japan, 79, 939–947, doi:10.2151/jmsj.79.939.
Fuchs, Ž., and D. J. Raymond, 2002: Large-scale modes of a nonrotating atmosphere with water vapor and cloud–radiation feedbacks. J. Atmos. Sci., 59, 1669–1679, doi:10.1175/1520-0469(2002)059<1669:LSMOAN>2.0.CO;2.
Fuchs, Ž., and D. J. Raymond, 2005: Large-scale modes in a rotating atmosphere with radiative–convective instability and WISHE. J. Atmos. Sci., 62, 4084–4094, doi:10.1175/JAS3582.1.
Fuchs, Ž., and D. J. Raymond, 2007: A simple, vertically resolved model of tropical disturbances with a humidity closure. Tellus, 59A, 344–354, doi:10.1111/j.1600-0870.2007.00230.x.
Grandpeix, J.-Y., and J.-P. Lafore, 2010: A density current parameterization coupled with Emanuel’s convection scheme. Part I: The models. J. Atmos. Sci., 67, 881–897, doi:10.1175/2009JAS3044.1.
Grandpeix, J.-Y., J.-P. Lafore, and F. Cheruy, 2010: A density current parameterization coupled with Emanuel’s convection scheme. Part II: 1D simulations. J. Atmos. Sci., 67, 898–922, doi:10.1175/2009JAS3045.1.
Gregory, D., 1995: The representation of moist convection in atmospheric models. Hadley Centre Climate Research Tech. Rep. 54, 36 pp.
Gregory, D., and P. R. Rowntree, 1990: A mass-flux convection scheme with representation of cloud ensemble characteristics and stability-dependent closure. Mon. Wea. Rev., 118, 1483–1506, doi:10.1175/1520-0493(1990)118<1483:AMFCSW>2.0.CO;2.
Hannah, W. M., and E. D. Maloney, 2014: The moist static energy budget in NCAR CAM5 hindcasts during DYNAMO. J. Adv. Model. Earth Syst., 6, 420–440, doi:10.1002/2013MS000272.
Hourdin, F., and Coauthors, 2013: LMDZ5B: The atmospheric component of the IPSL climate model with revisited parameterizations for clouds and convection. Climate Dyn., 40, 2193–2222, doi:10.1007/s00382-012-1343-y.
Huffman, G. J., and Coauthors, 2001: Global precipitation at one-degree daily resolution from multisatellite observations. J. Hydrometeor., 2, 36–50, doi:10.1175/1525-7541(2001)002<0036:GPAODD>2.0.CO;2.
Huffman, G. J., and Coauthors, 2007: The TRMM Multisatellite Precipitation Analysis (TMPA): Quasi-global, multiyear, combined-sensor precipitation estimates at fine scales. J. Hydrometeor., 8, 38–55, doi:10.1175/JHM560.1.
Hung, M.-P., J.-L. Lin, W. Wang, D. Kim, T. Shinoda, and S. J. Weaver, 2013: MJO and convectively coupled equatorial waves simulated by CMIP5 climate models. J. Climate, 26, 6185–6214, doi:10.1175/JCLI-D-12-00541.1.
Jakob, C., 2000: The representation of cloud cover in atmospheric general circulation models. Ph.D. thesis, Ludwig-Maximilians-Universitaet Muenchen, 193 pp.
Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437–471, doi:10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.
Kim, D., and Coauthors, 2009: Application of MJO simulation diagnostics to climate models. J. Climate, 22, 6413–6436, doi:10.1175/2009JCLI3063.1.
Kim, D., A. H. Sobel, E. D. Maloney, D. M. W. Frierson, and I.-S. Kang, 2011: A systematic relationship between intraseasonal variability and mean state bias in AGCM simulations. J. Climate, 24, 5506–5520, doi:10.1175/2011JCLI4177.1.
Kim, D., J.-S. Kug, and A. H. Sobel, 2014a: Propagating versus nonpropagating Madden–Julian oscillation events. J. Climate, 27, 111–125, doi:10.1175/JCLI-D-13-00084.1.
Kim, D., and Coauthors, 2014b: Process-oriented MJO simulation diagnostic: Moisture sensitivity of simulated convection. J. Climate, 27, 5379–5395, doi:10.1175/JCLI-D-13-00497.1.
Landu, K., and E. D. Maloney, 2011a: Effect of SST distribution and radiative feedbacks on the simulation of intraseasonal variability in an aquaplanet GCM. J. Meteor. Soc. Japan,89, 195–210, doi:10.2151/jmsj.2011-302.
Landu, K., and E. D. Maloney, 2011b: Intraseasonal moist static energy budget in reanalysis data. J. Geophys. Res., 116, D21117, doi:10.1029/2011JD016031.
Lau, W. K. M., and D. E. Waliser, 2011: Intraseasonal Variability in the Atmosphere–Ocean Climate System. 2nd ed. Springer, 614 pp.
Lee, M.-I., I.-S. Kang, J.-K. Kim, and B. E. Mapes, 2001: Influence of cloud–radiation interaction on simulating tropical intraseasonal oscillation with an atmospheric general circulation model. J. Geophys. Res., 106, 14 219–14 233, doi:10.1029/2001JD900143.
Le Treut, H., and Z.-X. Li, 1991: Sensitivity of an atmospheric general circulation model to prescribed SST changes: Feedback effects associated with the simulation of cloud optical properties. Climate Dyn., 5, 175–187, doi:10.1007/BF00251808.
Li, C., X. Jia, J. Ling, W. Zhou, and C. Zhang, 2009: Sensitivity of MJO simulations to diabatic heating profiles. Climate Dyn., 32, 167–187, doi:10.1007/s00382-008-0455-x.
Liebmann, B., and C. A. Smith, 1996: Description of a complete (interpolated) outgoing longwave radiation dataset. Bull. Amer. Meteor. Soc., 77, 1275–1277.
Lin, J.-L., and B. E. Mapes, 2004a: Radiation budget of the tropical intraseasonal oscillation. J. Atmos. Sci., 61, 2050–2062, doi:10.1175/1520-0469(2004)061<2050:RBOTTI>2.0.CO;2.
Lin, J.-L., and B. E. Mapes, 2004b: Wind shear effects on cloud-radiation feedback in the western Pacific warm pool. Geophys. Res. Lett., 31, L16118, doi:10.1029/2004GL020199.
Lin, J.-L., and Coauthors, 2006: Tropical intraseasonal variability in 14 IPCC AR4 climate models. Part I: Convective signals. J. Climate, 19, 2665–2690, doi:10.1175/JCLI3735.1.
Liu, H., and G. Wu, 1997: Impacts of land surface on climate of July and onset of summer monsoon: A study with an AGCM plus SSiB. Adv. Atmos. Sci., 14, 289–308, doi:10.1007/s00376-997-0051-8.
Loeb, N. G., B. A. Wielicki, D. R. Doelling, G. L. Smith, D. F. Keyes, S. Kato, N. Manalo-Smith, and T. Wong, 2009: Toward optimal closure of the Earth’s top-of-atmosphere radiation budget. J. Climate, 22, 748–766, doi:10.1175/2008JCLI2637.1.
Lohmann, U., and E. Roeckner, 1996: Design and performance of a new cloud microphysics scheme developed for the ECHAM4 general circulation model. Climate Dyn., 12, 557–572, doi:10.1007/BF00207939.
Madden, R. A., and P. R. Julian, 1972: Description of global-scale circulation cells in the tropics with a 40–50 day period. J. Atmos. Sci., 29, 1109–1123, doi:10.1175/1520-0469(1972)029<1109:DOGSCC>2.0.CO;2.
McFarlane, N. A., J. F. Scinocca, M. Lazare, R. Harvey, D. Verseghy, and J. Li, 2005: The CCCma third generation atmospheric general circulation model. CCCma Internal Rep., 25 pp.
Moorthi, S., and M. J. Suarez, 1992: Relaxed Arakawa–Schubert: A parameterization of moist convection for general circulation models. Mon. Wea. Rev., 120, 978–1002, doi:10.1175/1520-0493(1992)120<0978:RASAPO>2.0.CO;2.
Morrison, H., and A. Gettelman, 2008: A new two-moment bulk stratiform cloud microphysics scheme in the Community Atmosphere Model, version 3 (CAM3). Part I: Description and numerical tests. J. Climate, 21, 3642–3659, doi:10.1175/2008JCLI2105.1.
Neale, R. B., J. H. Richter, and M. Jochum, 2008: The impact of convection on ENSO: From a delayed oscillator to a series of events. J. Climate, 21, 5904–5924, doi:10.1175/2008JCLI2244.1.
Nordeng, T. E., 1994: Extended versions of the convective parameterization scheme at ECMWF and their impact on the mean and transient activity of the model in the tropics. ECMWF Tech. Memo. 206, 41 pp.
Pan, D.-M., and D. A. Randall, 1998: A cumulus parameterization with a prognostic closure. Quart. J. Roy. Meteor. Soc., 124, 949–981, doi:10.1002/qj.49712454714.
Rasch, P. J., and J. E. Kristjansson, 1998: A comparison of the CCM3 model climate using diagnosed and predicted condensate parameterizations. J. Climate, 11, 1587–1614, doi:10.1175/1520-0442(1998)011<1587:ACOTCM>2.0.CO;2.
Raymond, D. J., 2001: A new model of the Madden–Julian oscillation. J. Atmos. Sci., 58, 2807–2819, doi:10.1175/1520-0469(2001)058<2807:ANMOTM>2.0.CO;2.
Raymond, D. J., and Ž. Fuchs, 2009: Moisture modes and the Madden–Julian oscillation. J. Climate, 22, 3031–3046, doi:10.1175/2008JCLI2739.1.
Ricard, J., and J. Royer, 1993: A statistical cloud scheme for use in an AGCM. Ann. Geophys., 11, 1095–1115.
Richter, J. H., and P. J. Rasch, 2008: Effects of convective momentum transport on the atmospheric circulation in the Community Atmosphere Model, version 3. J. Climate, 21, 1487–1499, doi:10.1175/2007JCLI1789.1.
Rotstayn, L. D., 1997: A physically based scheme for the treatment of stratiform clouds and precipitation in large-scale models. I: Description and evaluation of the microphysical processes. Quart. J. Roy. Meteor. Soc., 123, 1227–1282, doi:10.1002/qj.49712354106.
Rotstayn, L. D., 1998: A physically based scheme for the treatment of stratiform clouds and precipitation in large-scale models. II: Comparison of modelled and observed climatological fields. Quart. J. Roy. Meteor. Soc., 124, 389–415, 10.1002/qj.49712454603.
Rotstayn, L. D., 2000: On the “tuning” of autoconversion parameterizations in climate models. J. Geophys. Res., 105, 15 495–15 507, doi:10.1029/2000JD900129.
Schumacher, C., and R. A. Houze Jr., 2006: Stratiform precipitation production over sub-Saharan Africa and the tropical East Atlantic as observed by TRMM. Quart. J. Roy. Meteor. Soc., 132, 2235–2255, doi:10.1256/qj.05.121.
Smith, R. N. B., 1990: A scheme for predicting layer clouds and their water content in a general circulation model. Quart. J. Roy. Meteor. Soc., 116, 435–460, doi:10.1002/qj.49711649210.
Sobel, A. H., and H. Gildor, 2003: A simple time-dependent model of SST hot spots. J. Climate, 16, 3978–3992, doi:10.1175/1520-0442(2003)016<3978:ASTMOS>2.0.CO;2.
Sobel, A. H., and E. Maloney, 2012: An idealized semi-empirical framework for modeling the Madden–Julian oscillation. J. Atmos. Sci., 69, 1691–1705, doi:10.1175/JAS-D-11-0118.1.
Sobel, A. H., and E. Maloney, 2013: Moisture modes and the eastward propagation of the MJO. J. Atmos. Sci., 70, 187–192, doi:10.1175/JAS-D-12-0189.1.
Sobel, A. H., J. Nilsson, and L. M. Polvani, 2001: The weak temperature gradient approximation and balanced tropical moisture waves. J. Atmos. Sci., 58, 3650–3665, doi:10.1175/1520-0469(2001)058<3650:TWTGAA>2.0.CO;2.
Sobel, A. H., E. D. Maloney, G. Bellon, and D. M. Frierson, 2010: Surface fluxes and tropical intraseasonal variability: A reassessment. J. Adv. Model. Earth Syst.,2 (2), doi:10.3894/JAMES.2010.2.2.
Sperber, K. R., and D. Kim, 2012: Simplified metrics for the identification of the Madden–Julian oscillation in models. Atmos. Sci. Lett., 13, 187–193, doi:10.1002/asl.378.
Su, H., and Coauthors, 2013: Diagnosis of regime-dependent cloud simulation errors in CMIP5 models using “A-Train” satellite observations and reanalysis data. J. Geophys. Res. Atmos., 118, 2762–2780, doi:10.1029/2012JD018575.
Sundqvist, H., E. Berge, and J. E. Kristjansson, 1989: Condensation and cloud parameterization studies with a mesoscale numerical weather prediction model. Mon. Wea. Rev., 117, 1641–1657, doi:10.1175/1520-0493(1989)117<1641:CACPSW>2.0.CO;2.
Taylor, K. E., R. J. Stouffer, and G. A. Meehl, 2012: An overview of CMIP5 and the experiment design. Bull. Amer. Meteor. Soc., 93, 485–498, doi:10.1175/BAMS-D-11-00094.1.
Tiedtke, M., 1989: A comprehensive mass flux scheme for cumulus parameterization in large-scale models. Mon. Wea. Rev., 117, 1779–1800, doi:10.1175/1520-0493(1989)117<1779:ACMFSF>2.0.CO;2.
Tiedtke, M., 1993: Representation of clouds in large-scale models. Mon. Wea. Rev., 121, 3040–3061, doi:10.1175/1520-0493(1993)121<3040:ROCILS>2.0.CO;2.
Tobin, I., S. Bony, C. E. Holloway, J. Y. Grandpeix, G. Sèze, D. Coppin, S. J. Woolnough, and R. Roca, 2013: Does convective aggregation need to be represented in cumulus parameterizations? J. Adv. Model. Earth Syst., 5, 692–703, doi:10.1002/jame.20047.
Tompkins, A., 2002: A prognostic parameterization for the subgrid-scale variability of water vapor and clouds in large-scale models and its use to diagnose cloud cover. J. Atmos. Sci., 59, 1917–1942, doi:10.1175/1520-0469(2002)059<1917:APPFTS>2.0.CO;2.
Watanabe, M., S. Emori, M. Satoh, and H. Miura, 2009: A PDF-based hybrid prognostic cloud scheme for general circulation models. Climate Dyn., 33, 795–816, doi:10.1007/s00382-008-0489-0.
Wheeler, M. C., and H. H. Hendon, 2004: An all-season real-time multivariate MJO index: Development of an index for monitoring and prediction. Mon. Wea. Rev., 132, 1917–1932, doi:10.1175/1520-0493(2004)132<1917:AARMMI>2.0.CO;2.
Wilcox, E. M., and L. J. Donner, 2007: The frequency of extreme rain events in satellite rain-rate estimates and an atmospheric general circulation model. J. Climate, 20, 53–69, doi:10.1175/JCLI3987.1.
Wilson, D. R., and S. P. Ballard, 1999: A microphysically based precipitation scheme for the UK Meteorological Office Unified Model. Quart. J. Roy. Meteor. Soc., 125, 1607–1636, doi:10.1002/qj.49712555707.
Yokoi, S., 2015: Multireanalysis comparison of variability in column water vapor and its analysis increment associated with the Madden–Julian oscillation. J. Climate, 28, 793–808, doi:10.1175/JCLI-D-14-00465.1.
Yukimoto, S., and Coauthors, 2011: Meteorological Research Institute Earth System Model version 1 (MRI-ESM1)—Model description. MRI Tech. Rep. 64, 83 pp. [Available online at http://www.mri-jma.go.jp/Publish/Technical/index_en.html.]
Zhang, C., 2005: Madden–Julian Oscillation. Rev. Geophys., 43, RG2003, doi:10.1029/2004RG000158.
Zhang, C., 2013: Madden–Julian oscillation: Bridging weather and climate. Bull. Amer. Meteor. Soc., 94, 1849–1870, doi:10.1175/BAMS-D-12-00026.1.
Zhang, G. J., and N. A. McFarlane, 1995: Sensitivity of climate simulations to the parameterization of cumulus convection in the Canadian Climate Centre general circulation model. Atmos.–Ocean, 33, 407–446, doi:10.1080/07055900.1995.9649539.
Zhang, G. J., and M. Mu, 2005: Effects of modifications to the Zhang–McFarlane convection parameterization on the simulation of the tropical precipitation in the National Center for Atmospheric Research Community Climate Model, version 3. J. Geophys. Res., 110, D09109, doi:10.1029/2004JD005617.
Zhang, G. J., and X. Song, 2009: Interaction of deep and shallow convection is key to Madden–Julian oscillation simulation. Geophys. Res. Lett., 36, L09708, doi:10.1029/2009GL037340.
Zurovac-Jevtić, D., S. Bony, and K. Emanuel, 2006: On the role of clouds and moisture in tropical waves: A two-dimensional model study. J. Atmos. Sci., 63, 2140–2155, doi:10.1175/JAS3738.1.