Abstract

This study examines cloud responses to global warming using a global nonhydrostatic model with two different cloud microphysics schemes. The cloud microphysics schemes tested here are the single- and double-moment schemes with six water categories: these schemes are referred to as NSW6 and NDW6, respectively. Simulations of one year for NSW6 and one boreal summer for NDW6 are performed using the nonhydrostatic icosahedral atmospheric model with a mesh size of approximately 14 km. NSW6 and NDW6 exhibit similar changes in the visible cloud fraction under conditions of global warming. The longwave (LW) cloud radiative feedbacks in NSW6 and NDW6 are within the upper half of the phase 5 of the Coupled Model Intercomparison Project (CMIP5)–Cloud Feedback Model Intercomparison Project 2 (CFMIP2) range. The LW cloud radiative feedbacks are mainly attributed to cirrus clouds, which prevail more in the tropics under global warming conditions. For NDW6, the LW cloud radiative feedbacks from cirrus clouds also extend to midlatitudes. The changes in cirrus clouds and their effects on LW cloud radiative forcing (LWCRF) are assessed based on changes in the effective radii of ice hydrometeors () and the cloud fraction. It was determined that an increase in has a nonnegligible impact on LWCRF compared with an increase in cloud fraction.

1. Introduction

Clouds play an important role in the energy balance of the atmosphere (e.g., Hartmann et al. 1992), and their response to global warming is the most uncertain factor in estimations of climate sensitivity. It has been reported that the uncertainty of climate predictions is mainly derived from the uncertainty of cloud feedbacks in general circulation models (GCMs; Soden and Held 2006; Dufresne and Bony 2008; Vial et al. 2013). By analyzing the phase 5 of the Coupled Model Intercomparison Project (CMIP5) dataset, Vial et al. (2013) indicated that the large spread of cloud feedbacks in models, especially cloud feedbacks in the tropics, is the main contributor to uncertainties in climate sensitivity. The large cloud radiative feedback spread in models originates from GCM-simulated cloud uncertainties, which includes uncertainties in the properties of liquid and ice hydrometeors. In particular, by analyzing CMIP5 models, a large spread in simulated cloud ice has been reported (Waliser et al. 2009; Li et al. 2012). This is thought to be a result of the choice and combination of cloud and convective parameterization schemes in CMIP5.

The explicit use of cloud microphysics schemes in high-resolution global models may lead to more consistent reproductions of the distributions of liquid and ice hydrometeors (Tomita 2008; Seiki et al. 2015a). In addition to cloud water and cloud ice, precipitating hydrometeors such as rain, snow, and graupel should be explicitly calculated over resolvable grid scales because of the importance of the recently recognized radiative effects of precipitating hydrometeors (e.g., Li et al. 2012, 2013). Using the Nonhydrostatic Icosahedral Atmospheric Model (NICAM; Tomita and Satoh 2004; Soden et al. 2008), the cloud processes of which are calculated solely using a cloud microphysics scheme without convective parameterization, the simulated cloud ice distributions are consistent with those obtained from satellite observations (Kodama et al. 2012; Hashino et al. 2013; Roh and Satoh 2014; Seiki et al. 2015b). Multiple sensitivity studies estimating cloud responses to global warming have been conducted using NICAM (Collins and Satoh 2009; Satoh et al. 2012; Tsushima et al. 2014; Noda et al. 2014). The responses of high clouds to the global warming simulated by NICAM are consistently different from those simulated by conventional GCMs (Bretherton 2015). NICAM simulations indicate that high clouds in the tropics increase, whereas the ice water path (IWP) decreases (Collins and Satoh 2009; Satoh et al. 2012). Noda et al. (2014) revealed that the change in the high cloud fraction is mainly due to relatively smaller clouds, which are not resolved by conventional GCMs. Tsushima et al. (2014) found that the large longwave (LW) cloud feedbacks in NICAM are mainly attributed to an increase in high clouds and that the cloud response to global warming in NICAM is strongly correlated with the positive biases in atmospheric temperature and the amount of cirrus clouds in the present climate.

Recently, two different bulk cloud microphysics schemes have been implemented in NICAM: the single moment of six water categories scheme (NSW6; Tomita 2008) and the double moment of six water categories scheme (NDW6; Seiki and Nakajima 2014; Seiki et al. 2015a). Cloud behaviors in the present climate in NSW6 and comparisons with various satellite observations have been studied by Kodama et al. (2012) and Hashino et al. (2013). The warm bias in the present atmospheric temperature, which is present in the tropical upper troposphere in NSW6, is reduced in NDW6 by improving the simulated cirrus fraction, cloud-top height, and consequent cloud-radiative effects (Seiki et al. 2015b). Although the cloud responses to global warming in NICAM are correlated with the present-climate biases in atmospheric temperature and simulated cirrus clouds (Tsushima et al. 2014), how the cloud responses to global warming act with different cloud microphysics schemes has not been tested. The aim of this study is to reveal whether the cloud responses, especially of cirrus clouds, to the global warming in NICAM are robust when different cloud microphysics schemes, NSW6 and NDW6, are implemented in NICAM. We also aim to investigate how clouds respond to global warming for different cloud microphysics schemes implemented in NICAM: NSW6 and NDW6. In particular, we focus on changes in high clouds. We examine cloud feedbacks and compare our results with CMIP5 and the Cloud Feedback Model Intercomparison Project 2 (CFMIP2) using a cloud radiative kernel proposed by Zelinka et al. (2012). Because the available integration time is limited, we mainly discuss the LW cloud feedbacks, which exhibit a smaller seasonal dependency compared with shortwave (SW) cloud feedbacks. We also discuss the difference in the two cloud microphysics schemes with regard to the change in LW cloud radiative forcing (LWCRF) caused by global warming.

This is the first attempt to examine cloud responses to global warming with a double-moment cloud microphysics scheme implemented in a high-resolution global model without convective parameterization. We investigate the sensitivity of cloud changes to the cloud microphysics scheme and the effect of the change in high clouds on the atmospheric energy balance. In section 2, we introduce the method used. The results are described in section 3. A discussion is provided in section 4, and concluding remarks are given in section 5.

2. Method

a. Model and cloud microphysics schemes

Simulations are conducted using NICAM (Tomita and Satoh 2004; Soden et al. 2008), implemented with the following two cloud microphysics schemes: the single moment (NSW6; Tomita 2008) and double moment of six water categories schemes (NDW6; Seiki and Nakajima 2014; Seiki et al. 2015a). NICAM simulates clouds over the global domain for all grid points with explicit microphysical processes but without a cumulus parameterization. The modeling grid domain has a horizontal mesh interval of approximately 14 km and 40 vertical layers up to 40 km in altitude. Although a 14-km horizontal resolution is not sufficient to resolve the structure of individual cumulus clouds, the global-scale statistical properties of clouds can be reproduced (Satoh et al. 2010; Iga et al. 2011; Kodama et al. 2012; Noda et al. 2012). Analysis of the 3.5-km mesh NICAM simulation by Hashino et al. (2013) showed that the cloud statistical properties are consistent between the 3.5-km and 14-km simulations. Since this study mainly examines cloud statistics, we expect that the cloud statistics can be represented by this kind of horizontal resolution. Although it is still unknown whether the cloud statistics converge if the model resolution is increased, this study focuses on the categorical differences and similarities between the two cloud microphysics schemes.

Both NSW6 and NDW6 solve the mass concentrations of six types of water categories: water vapor, cloud water, cloud ice, rain, snow, and graupel. In addition, NDW6 predicts the number concentrations of each hydrometeor, which are cloud water, cloud ice, rain, snow, and graupel. The advantage to using NICAM with explicit cloud microphysics processes and without a cumulus parameterization is that all of the hydrological cycles are consistently represented, including precipitating hydrometeors and cloud–radiative interactions, for both liquid and ice clouds, and its importance has very recently been recognized for GCMs (e.g., Li et al. 2012, 2013).

b. Experimental settings

The initial and boundary conditions for the present climate and future climate are described here. The initial conditions for the present climate in this study are similar to those used by Yamada et al. (2010), Kodama et al. (2012), and Noda et al. (2015). The initial dates for the present and future climate are 1 June and 1 May 2004, respectively. The start date in the analysis is 1 June considering a one-month spinup period for the future climate. A slab ocean model with 15-m depth is employed, and the sea surface temperature (SST) is nudged toward the National Centers for Environmental Prediction (NCEP) Reynolds optimally interpolated weekly SST (Reynolds et al. 2002) for the present climate run. The nudging relaxation time was 7 days. The slab ocean model is employed herein because better geographical distributions of the precipitation can be found in NICAM (Kodama et al. 2015). The sea ice concentration (SIC) is constrained to the monthly mean HadISST1 (Rayner et al. 2003) from the NCEP final reanalysis dataset (NCEP-FNL; http://rda.ucar.edu/datasets/ds083.2/). The SST and SIC for the warming experiment are created by the method proposed by Mizuta et al. (2008) by using the ensemble mean of 18 models, which are listed in Table 1, from phase 3 of the Coupled Model Intercomparison Project (CMIP3). The SST and SIC for the warming experiment add the difference between the future climate average (2075–99) and the present climate average (1979–2009), which shows an El Niño–like SST pattern. For details of the SST difference distribution, please see Fig. 1d of Yamada et al. (2010). The CO2 concentration is 348 ppmv for the present climate, and doubled for the future climate.

Table 1.

The 18 models in CMIP3 used for making the boundary condition in both NSW6 and NDW6.

The 18 models in CMIP3 used for making the boundary condition in both NSW6 and NDW6.
The 18 models in CMIP3 used for making the boundary condition in both NSW6 and NDW6.

Two datasets with the same initial and boundary conditions mentioned above are prepared with different cloud microphysics schemes, NSW6 and NDW6. Because of a limitation in computational resources, the data length for analysis for NSW6 is one year, whereas the data length for NDW6 is the boreal summer from June to August. For the NSW6 simulation, two integration durations were used: one year (NSW6-ANNUAL) and the boreal summer (NSW6-JJA). The comparison between NSW6-ANNUAL and NSW6-JJA is applied to show the seasonal variation. Then, we compared the results between NSW6-JJA and the boreal summer with NDW6 (NDW6-JJA) to show the differences/similarities between the NSW6 and NDW6 schemes.

c. Analysis method

The aim of this study is to investigate cloud responses to global warming and the associated cloud radiative feedbacks. We analyze numerical results with two cloud microphysics schemes, NSW6 and NDW6, and compare their cloud feedbacks, with those of multiensemble CMIP5/CFMIP2 (http://cfmip.metoffice.com/) simulations using the same cloud radiative kernel platform proposed by Zelinka et al. (2012). Using this radiative kernel, the cloud feedbacks associated with each type of cloud fraction change, based on the International Satellite Cloud Climatology Project (ISCCP; Schiffer and Rossow 1983; Rossow and Schiffer 1999), can be consistently compared with the CMIP5/CFMIP2 GCM results. Note that we use the five models that provide ISCCP output dataset, which are also called CFMIP2 and are the same data used in Zelinka et al. (2013), to make our comparison. The cloud radiative feedbacks in models may be strongly dependent on the radiative transfer model implemented in the model while the cloud radiative kernel calculates the cloud radiative feedbacks directly from the cloud fraction change (Zelinka et al. 2012); thus, this cloud radiative kernel is chosen in this study. How the representation of cloud feedbacks depends on the radiative transfer models implemented in models has been discussed in Zelinka et al. (2012). That study demonstrated a discrepancy in the cloud radiative feedbacks calculated by the partial radiative perturbation (PRP) method (Shell et al. 2008; Satoh et al. 2008) and by their cloud radiative kernel. Zelinka et al. (2012) demonstrated that the cloud radiative feedbacks are sensitive in models containing radiative transfer models that are not the Fu–Liou model, because the cloud radiative kernel is constructed by the Fu–Liou model. The radiative transfer model in NICAM is mstrnX (Sekiguchi and Nakajima 2008), so a discrepancy between the cloud feedbacks calculated by the PRP method and by the cloud radiative kernel can be expected. As the present study focuses on the cloud radiative feedbacks contributed by the cloud fraction change itself and we wish to ignore the effect of the radiative transfer models, we chose the cloud radiative kernel proposed by Zelinka et al. (2012) for use in our analysis. Also, the analysis of the cloud radiative kernel is based on the output of the ISCCP simulator (Klein and Jakob 1999; Webb et al. 2001) implemented in NICAM.

3. Results

a. Cloud distributions

Figures 1a–f show the horizontal visible [cloud optical depth (COD) > 0.3] cloud fraction distributions, which are converted for ISCCP simulator output, for the present climate and the response between the present and warming conditions for the simulations with NSW6-ANNUAL, NSW6-JJA, and NDW6-JJA. The ISCCP D1 visible cloud observations averaged from June 2004 to May 2005 and averaged from June to August (JJA) 2004 are plotted in Figs. 1g and 1h, respectively, as a reference. In the present climate, the global cloud distributions are similar for NSW6-ANNUAL (Fig. 1a), NSW6-JJA (Fig. 1b), and NDW6-JJA (Fig. 1c) but have different magnitudes. Large cloud fractions appear in the northern Pacific, the tropics, and the southern latitudinal belt between 30° and 60°S. Although the patterns of NSW6-ANNUAL, NSW6-JJA, and NDW6-JJA are similar to one another, the magnitudes are different. For example, the cloud fractions in the southern latitudinal belt between 30° and 60°S and at the North Pole (where low clouds are dominant) in NDW6-JJA are larger than those in NSW6-JJA by 20%–30%. The cloud fractions in the tropical region, where high clouds mainly dominate, in NDW6-JJA are 10%–20% smaller than those in NSW6-JJA. Comparing the NSW6-ANNUAL with the one-year ISCCP observation (Fig. 1g), NSW6-ANNUAL underestimates the global mean cloud fraction by about −8% compared with ISCCP. In addition, both NSW6-JJA and NDW6-JJA underestimate the global mean cloud fraction in the present climate when comparing the ISCCP cloud fraction in the boreal summer, as illustrated in Fig. 1h. The seasonal global mean cloud fraction in NSW6 fluctuates within 2%, from 53% to 55%, which is nonnegligible. Although seasonal variation in the cloud fraction is nonnegligible, it does not affect our qualitative analysis of the cloud radiative feedbacks because a similar pattern of cloud distribution remains. The underestimation of cloud fraction in NICAM is similar to the commonly observed bias in the ensemble mean of CMIP3/5 models (Klein et al. 2013) although the negative bias is especially due to underestimating low clouds, which can also be obtained by other NICAM experiments (e.g., Kodama et al. 2012, 2015).

Fig. 1.

Visible cloud fraction distributions in the present climate in (a) NSW6-ANNUAL, (b) NSW6-JJA, and (c) NDW6-JJA, and its response to global warming for (d) NSW6-ANNUAL, (e) NSW6-JJA, and (f) NDW6-JJA. (g) Plots of ISCCP D1 observations averaged from June to August 2004.

Fig. 1.

Visible cloud fraction distributions in the present climate in (a) NSW6-ANNUAL, (b) NSW6-JJA, and (c) NDW6-JJA, and its response to global warming for (d) NSW6-ANNUAL, (e) NSW6-JJA, and (f) NDW6-JJA. (g) Plots of ISCCP D1 observations averaged from June to August 2004.

The cloud responses of the cloud fraction to global warming for NSW6-ANNUAL, NSW6-JJA, and NDW6-JJA are plotted in Figs. 1d, 1e, and 1f, respectively. The response of the cloud fraction geographical distributions to global warming is similar between NSW6-ANNUAL and NSW6-JJA although the magnitude is different, such as the global mean value. Subsequently, if we compare the NSW6-JJA and NDW6-JJA simulations, we find that the response in NDW6-JJA is stronger than that in NSW6-JJA. The cloud responses to global warming in the tropics are regionally different between NSW6 and NDW6. For example, around the Philippine region, the cloud fraction decreases for NSW6-JJA whereas that for NDW6-JJA increases. On the west coast of southern America, a prominent decrease in cloud fraction is observed for NDW6-JJA whereas that for NSW6-JJA increases. These differences in cloud fraction are primarily caused by the high cloud fraction change in this region (figures not shown).

The global mean ISCCP cloud fraction, categorized by COD and cloud-top pressure (CTP) for the present climate, and its response to global warming are plotted in Fig. 2. The cloud distributions of the present climate in NSW6-ANNUAL, NSW6-JJA, and NDW6-JJA are plotted in Figs. 2a, 2b, and 2c, respectively, and their responses to global warming are plotted in Figs. 2d, 2e, and 2f, respectively. In addition, the ISCCP D1 observations averaged from June 2004 to May 2005 and averaged from June 2004 to August 2004 are plotted in Figs. 2g and 2h, respectively. We initially focus on the comparison between NSW6-ANNUAL and NSW6-JJA. The global mean ISCCP cloud fraction and the response to global warming of NSW6-ANNUAL and NSW6-JJA are similar. In the present climate, large cloud fractions of high, thin clouds and small cloud fractions of low clouds appear in both NSW6-ANNUAL and NSW6-JJA compared with the ISCCP observations. High clouds in NSW6 and NDW6 are ~3.5% and ~2% more than ISCCP D1 observations, respectively, and low clouds in NSW6 and NDW6 are ~10% and ~9% less than ISCCP D1 observations, respectively. Global warming leads to a significant increase in high, thin clouds. Seasonal variations in high cloud fractions in ISCCP are small compared with NSW6-ANNUAL and NSW6-JJA. As systematic biases in GCMs appear over short integration times within a few months (Williams et al. 2013) and we do not observe any large seasonal variation between NSW6-ANNAUL and NSW6-JJA, the 3-month averaged cloud fractions likely represent characteristic cloud behavior in both NSW6 and NDW6. In the following analysis, we use NSW6-JJA and NDW6-JJA to compare different cloud microphysics schemes.

Fig. 2.

Global mean cloud distribution in the ISCCP diagram. (a)–(c) The ISCCP distribution for NSW6-ANNUAL, NSW6-JJA, and NDW6-JJA, respectively. (d)–(f) As in (a)–(c), but showing the response of the ISCCP distribution to global warming. The values plotted are global mean cloud fractions of each cloud type. The ISCCP D1 observations (g) averaged from June 2004 to May 2005 and (h) averaged from June 2004 to August 2004. Units are %.

Fig. 2.

Global mean cloud distribution in the ISCCP diagram. (a)–(c) The ISCCP distribution for NSW6-ANNUAL, NSW6-JJA, and NDW6-JJA, respectively. (d)–(f) As in (a)–(c), but showing the response of the ISCCP distribution to global warming. The values plotted are global mean cloud fractions of each cloud type. The ISCCP D1 observations (g) averaged from June 2004 to May 2005 and (h) averaged from June 2004 to August 2004. Units are %.

Upon investigation of the global mean ISCCP distribution in the present climate between NSW6-JJA and NDW6-JJA, as shown in Fig. 2, we first note that the overall structures are similar. We found that the uppermost clouds, including cirrus clouds (high and thin clouds), generally increase, whereas low clouds generally decrease following global warming for both NSW6-JJA and NDW6-JJA. This prominent increase in cirrus clouds has been noted in other experiments with NICAM (Satoh et al. 2012; Tsushima et al. 2014; Noda et al. 2014). Although cirrus clouds prominently increase when the atmosphere warms, detailed cloud response behaviors are found to be different between NSW6-JJA and NDW6-JJA. In NSW6-JJA, the increase in the cirrus cloud fraction is mainly dominated by the increase in the uppermost and thinnest clouds; however, for NDW6-JJA, there is an increase in both the uppermost cloud fraction and in the cloud fractions with CODs between 0.3 and 3.6 and CTPs from 310 to 180 hPa. Note that the midlevel clouds are more sensitive in NDW6-JJA than in NSW6-JJA; however, we mainly focus on upper clouds in the following discussion.

b. Cloud feedbacks

The global mean SW, LW, and net cloud feedbacks for NSW6-ANNUAL, NSW6-JJA, NDW6-JJA, and the CMIP5/CFMIP2 average are plotted in Fig. 3. The total SW, LW, and net cloud feedbacks and the contribution of each type of clouds are also shown. Here, clouds are categorized by the ISCCP classification of CTP (high clouds, CTP ≤ 440 hPa; middle clouds, 440 hPa ≤ CTP < 680 hPa; low clouds, 680 hPa ≤ CTP < 1000 hPa) and COD (thin clouds, 0.3 ≤ COD < 3.6; medium clouds, 3.6 ≤ COD < 23; thick clouds, 23 ≤ COD). The seasonal variation in SW cloud feedbacks can be obtained by comparing the SW cloud feedbacks for both NSW6-ANNUAL and NSW6-JJA. This seasonal variation in SW cloud feedbacks comes from high and low clouds and thin to medium clouds. The seasonal variation in SW cloud feedbacks can be found when comparing the SW cloud feedbacks for both NSW6-ANNUAL and NDW6-JJA. This seasonal variation in SW cloud feedbacks comes from high and low clouds and thin to medium clouds. In contrast, the seasonal variation in LW cloud feedbacks is less pronounced than the SW cloud feedbacks, which confirms that LW cloud feedbacks can be analyzed by using the JJA datasets. Note that the behaviors of LW cloud feedbacks are similar in CMIP3/CFMIP1 and CMIP5/CFMIP2.

Fig. 3.

The global mean SW, LW, and net cloud feedbacks corresponding to all clouds and each type of cloud when clouds are classified by CTP and COD. The results of NSW6-ANNUAL (triangles), NSW6-JJA (stars), NDW6-JJA (circles), and CMIP5/CFMIP2 average (diamonds) are plotted. Note the small dots indicate each member of the CMIP5/CFMIP2 models (by courtesy of Dr. Zelinka).

Fig. 3.

The global mean SW, LW, and net cloud feedbacks corresponding to all clouds and each type of cloud when clouds are classified by CTP and COD. The results of NSW6-ANNUAL (triangles), NSW6-JJA (stars), NDW6-JJA (circles), and CMIP5/CFMIP2 average (diamonds) are plotted. Note the small dots indicate each member of the CMIP5/CFMIP2 models (by courtesy of Dr. Zelinka).

Our discussion will focus on the LW cloud feedbacks, which have smaller seasonal variation and can be analyzed with NICAM from the shorter JJA dataset for both NSW6-JJA and NDW6-JJA, because in this study we are interested in a comparison of the cloud behavior of NSW6 and NDW6. Figure 3 shows that LW cloud feedbacks for both NSW6 and NDW6 are within the uncertainty of the CMIP5/CFMIP2 model ensemble. An exception is for thin clouds, for which the LW cloud feedbacks in NSW6 and NDW6 are larger than the CMIP5/CFMIP2model ensembles. Changes in thin clouds contribute to only a small fraction of the LW cloud feedbacks in CMIP3/CFMIP1 models (Zelinka et al. 2012).

Figure 4 displays the global mean LW cloud feedbacks for each category of the ISCCP matrix for NSW6-JJA (Fig. 4a) and NDW6-JJA (Fig. 4b). These results demonstrate that the LW cloud feedbacks are mostly attributed to cirrus clouds (high and thin clouds) in both NSW6-JJA and NDW6-JJA, although nonnegligible contributions from the other cloud categories are apparent (e.g., low and medium clouds for NDW6-JJA and high and medium to thick clouds for NSW6-JJA and NDW6-JJA). In more detail, this large cirrus cloud contribution to LW cloud feedbacks is dominated by high-top clouds with CTP < 180 hPa for both NSW6-JJA and NDW6-JJA. For NSW6-JJA (Fig. 4a), clouds with CTP < 180 hPa and COD < 3.6 contribute to positive LW cloud feedbacks, whereas clouds with 180 hPa < CTP contribute to negative LW cloud feedbacks. For NDW6-JJA (Fig. 4b), clouds with CTP < 310 hPa contribute to positive LW cloud feedbacks, whereas clouds with CTP < 180 hPa and 1.3 < COD < 3.6 contribute slightly negative LW cloud feedbacks. This difference is consistent with the difference in cloud fraction response to a warming atmosphere, as shown in Figs. 2e and 2f. High cloud properties simulated by 7- and 14-km simulations in NICAM and their responses to global warming have been examined by Noda et al. (2014). They found that model-resolvable high cloud sizes are more appropriately simulated by NICAM as the horizontal resolution increases. They also found that, under the global warming, the increase in high clouds with size smaller than 50 km has strongest impact on the increase in LW cloud radiative forcing, which may lead the cloud feedbacks in NICAM are larger than conventional GCMs.

Fig. 4.

LW cloud feedbacks on ISCCP diagrams estimated from the cloud radiative kernel, for (a) NSW6-JJA and (b) NDW6-JJA.

Fig. 4.

LW cloud feedbacks on ISCCP diagrams estimated from the cloud radiative kernel, for (a) NSW6-JJA and (b) NDW6-JJA.

Some studies have estimated the cloud feedbacks from observational data. Mauritsen and Stevens (2014) argued that the LW cloud feedbacks should be negative by analyzing regression between LWCRF and SST due to the “iris effect.” This iris effect, which argued the cloud coverage should decrease under the global warming condition, is not observed in the results of NICAM. Noda et al. (2014) showed that the coverage of high clouds increases under the global warming condition. On the other hand, Zhou et al. (2014) showed that the LW cloud feedbacks attributed by cirrus cloud 0.20 ± 0.21 W m−2 K by analyzing CALIPSO observations and indicated that the LW cloud feedbacks attributed by cirrus cloud are underestimated by most of conventional GCM. Results shown in this study are consistent with the later study. Furthermore, Noda et al. (2016) revealed that, under the global warming, the emissivity change in small high clouds is important, and to simulate small high cloud is necessary for getting a better understanding of cloud feedbacks.

In Fig. 4, both NSW6-JJA and NDW6-JJA showed large negative LW cloud radiative feedbacks attributed to low-medium clouds (COD between 306 and 23 and CTP lower than 680 hPa). This pattern is related to the large cloud fraction decrease in this cloud category in both NSW6 and NDW6 (Figs. 2e and 2f). Moreover, both NSW6 and NDW6 have negative biases in the cloud fraction of this category compared to ISCCP observations (Fig. 1). The low to medium clouds in models are strongly related to the boundary layer conditions and are complex. Some studies have revealed, by using conventional models such as CMIP3 and CMIP5, that there is a relationship between the negative bias of the low clouds and high clouds and demonstrated that models with fewer low clouds may allow more high clouds to exist: these models are then more sensitive to global warming (Fasullo and Trenberth 2012; Sherwood et al. 2014). In NICAM, Noda et al. (2012) conducted a sensitivity study on how the turbulence scheme modulated the low and high clouds by using 14-km NICAM simulations. They found that the cloud fraction of low and high clouds is sensitive to the state (saturated or unsaturated) of the grid cells. When the condition of the grid cell is assumed to be saturated (unsaturated), the number of low clouds decreases (increases) and the number of high clouds increases (decreases) in NICAM. They also indicated that the formation of high clouds cannot be simply explained by a change in the vertical transportation of water vapor, although there is a prominent increase (decrease) in high clouds (low clouds) in the experiment in which the grid cell is assumed to be saturated. The issue of how the boundary schemes affect cloud formation is beyond the scope of the present study and should be explored in the near future.

Figure 5 shows the zonal mean distributions of cirrus cloud (high and thin cloud) change and LW cloud feedbacks, which are attributed to cirrus clouds, for NSW6-JJA and NDW6-JJA (Figs. 5a and 5b, respectively). Note that we extracted out the LW cloud feedbacks, which are calculated by the cloud radiative kernel, attributed by cirrus cloud changes shown in Fig. 4, and plotted the zonal mean distributions of the cirrus LW cloud feedbacks of NSW6 and NDW6 in Figs. 5a and 5b, respectively. For NSW6-JJA, the cirrus cloud fraction significantly increases in the tropics and, as a result, the LW cloud feedbacks attributed to cirrus clouds are mostly dominant in the tropics. In contrast to NSW6-JJA, although the NDW6-JJA cirrus cloud fraction changes mainly in the tropics, appreciable changes occur in the midlatitudes. This means that the LW cloud feedbacks from cirrus clouds dominate in the tropics, although they are nonnegligible in the midlatitudes. Table 2 summarizes the values of the cirrus cloud fraction change and cirrus clouds contributing to the LW cloud feedbacks within the tropics (30°N–30°S) and outside of that area. Our results indicate that the characteristics of the cirrus cloud fraction change for NSW6-JJA are different from those of NDW6-JJA, suggesting that the cloud responses to global warming are different between the simulations with NSW6-JJA and NDW6-JJA. We explore the mechanism of the difference between the two schemes in the next section.

Fig. 5.

The zonal mean distribution of cirrus cloud changes (red lines) and LW cloud feedbacks contributed by cirrus clouds changes (blue lines) for (a) NSW6-JJA and (b) NDW6-JJA.

Fig. 5.

The zonal mean distribution of cirrus cloud changes (red lines) and LW cloud feedbacks contributed by cirrus clouds changes (blue lines) for (a) NSW6-JJA and (b) NDW6-JJA.

Table 2.

List of cloud fraction changes in cirrus clouds and LW cloud feedback contributed by cirrus clouds within the tropical region (30°N–30°S) and outside of the tropical region for NSW6-JJA and NDW6-JJA. The units for cloud fraction change and LW cloud feedback are % and W m−2 K−1, respectively.

List of cloud fraction changes in cirrus clouds and LW cloud feedback contributed by cirrus clouds within the tropical region (30°N–30°S) and outside of the tropical region for NSW6-JJA and NDW6-JJA. The units for cloud fraction change and LW cloud feedback are % and W m−2 K−1, respectively.
List of cloud fraction changes in cirrus clouds and LW cloud feedback contributed by cirrus clouds within the tropical region (30°N–30°S) and outside of the tropical region for NSW6-JJA and NDW6-JJA. The units for cloud fraction change and LW cloud feedback are % and W m−2 K−1, respectively.

4. Discussion

In the previous section, we demonstrated that large LW cloud feedbacks for both NSW6 and NDW6 are due to the increase in cirrus clouds in the tropics. Tsushima et al. (2014) found that the strength of the cloud radiative feedbacks is strongly correlated with the tropical high cloud fraction in the present climate. The tropical high cloud fraction is thought to be affected by detailed ice cloud microphysical processes. Here, we analyze the tropical cirrus cloud microphysics characteristics in NSW6 and NDW6. In particular, we focus on the distribution of IWP and ice optical depth (), and their relation to the effective radius of ice particles (especially for snow and cloud ice) and the emissivity.

We use the probability density function [PDF; ] that satisfies

 
formula

and the weighted distribution of IWP, which is

 
formula

that represents

 
formula

to evaluate the ice cloud fraction and the total amount of ice matters in each IWP bin simulated by NSW6-JJA and NDW6-JJA. The results are compared with IWP collected by CloudSat (the 2B-CWC-RO data; Austin and Stephens 2001; Austin et al. 2009). Note that the IWP data collected by CloudSat are averaged from 2007 to 2011 as a reference climatological state. Figures 6a and 6b show the PDF and the weighted distribution of the PDF, respectively, and the responses of the PDF and the weighted distribution of PDF to global warming are plotted in Figs. 6c and 6d, respectively. NDW6 generally reproduces a better PDF and weighted PDF distribution than NSW6, as illustrated in Figs. 6a and 6b. This result is consistent with that of Seiki et al. (2015a). They revealed that the cross section of ice water content in NSW6 is similar to NDW6, whereas the ice content in NSW6 is half of that in NDW6. We speculate that double-moment bulk schemes generally produce more ice hydrometeors than single-moment bulk schemes do. Milbrandt and Yau (2005) and Shipway and Hill (2012) demonstrated that single-moment bulk schemes are likely to lose the liquid water path rapidly through precipitation processes; thus, subsequent ice-cloud processes do not develop sufficiently when using single-moment schemes. The superiority of double-moment schemes has been argued previously by Igel et al. (2015).

Fig. 6.

The (a) unweighted and (b) weighted PDF of IWP in the tropical region for NSW6-JJA, NDW6-JJA, and observations (CloudSat). Also shown are the response to atmospheric warming of the (c) unweighted and (d) weighted PDF of IWP for NSW6-JJA and NDW6-JJA.

Fig. 6.

The (a) unweighted and (b) weighted PDF of IWP in the tropical region for NSW6-JJA, NDW6-JJA, and observations (CloudSat). Also shown are the response to atmospheric warming of the (c) unweighted and (d) weighted PDF of IWP for NSW6-JJA and NDW6-JJA.

We now focus on the cirrus cloud range in the IWP diagrams. As a reference, the upper boundary of IWP for cirrus clouds in the ISCCP category is marked as black dotted lines in Figs. 6a–d; moreover, the upper bound of IWP is 90 g m−2, which corresponds approximately to a COD value of 3.6 assuming that the typical cloud effective radius is 40 μm, as shown in Eq. (1) below:

 
formula

where is the optical depth of ice clouds and ρi and are the ice particle density and ice particle effective radius, respectively. Note that in NDW6-JJA is calculated using IWP and , which are prognostically evaluated by the double-moment cloud microphysics scheme (Seiki et al. 2014), whereas in NSW6 is fixed as 40 μm. Note that the of ice matter in the analysis of LW cloud feedbacks using the cloud radiative kernel is fixed at 30 μm and does not change due to the global warming, but it is still a useful tool for analyzing cloud radiative feedbacks quantitatively. Vertical profiles of the ice water content and , simulated by NDW6, were statistically evaluated by comparison to the CALIPSO and CloudSat satellite observations, and were found to be slightly underestimated for cirrus clouds (Seiki et al. 2015b).

According to Tsushima et al. (2014), the strong LW cloud feedbacks for NSW6-JJA and NDW6-JJA are expected to be induced by relatively larger high cloud fractions. Note that both NSW6 and NDW6 simulated more high clouds than the ISCCP observations (Fig. 2). As shown in Figs. 6a and 6b, the simulations with both NSW6 and NDW6 produce more cloud fraction in the IWP cirrus cloud range and less IWP in the entire range of IWP compared with the CloudSat observation, although NDW6 simulations are in relatively better agreement with the CloudSat observation. Thus, our results are consistent with the findings of Tsushima et al. (2014). Concerning the global warming response, our results show that the cirrus cloud fraction increases in the tropics for both NSW6 and NDW6 (Fig. 6c).

The larger LW cloud feedback in our results (see Fig. 3) may be related to the larger cirrus cloud fraction in the present climate and the increase in cirrus cloud fraction resulting from global warming. Compared with the CloudSat observations, both NSW6 and NDW6 overestimate the occurrence of cirrus, as illustrated in Fig. 6a, where the IWP is less than 90 g m−2: the increase in cirrus clouds can be mainly attributed to thin clouds with IWP less than 30 g m−2 whereas there is a slight decrease in the IWP range from 30 to 90 g m−2. Comparing Figs. 6b and 6d, the increase in cirrus clouds is sensitive to changes in the amount of IWP. Berry and Mace (2014) found that cirrus clouds with an IWP of less than 200 g m−2 play an important role in LW heating at the top of the atmosphere, and the LW heating due to cirrus clouds is mostly attributed to cirrus clouds with a small IWP range from 5 to 60 g m−2. Hence, we examine the LW radiative effect of ice clouds as a function of IWP.

We first demonstrated how the simulation of in NSW6 differs from that in NDW6 as a function of IWP. The relationship between IWP and is given in Eq. (1). From this equation, of NSW6 is expected to be different from that of NDW6 at the same IWP since in NSW6 is fixed as 40 μm, whereas is predicted at every time step in NDW6. In the following analysis, to focus on the characteristics of cirrus clouds, we exclude graupel from the column values of ice microphysical parameters and only cloud ice and snow are used.

We examine how changes as a function of IWP using Eq. (1): the results are plotted in Fig. 7. Figure 7a shows the relation between and IWP for NSW6-JJA and NDW6-JJA in the present climate and under a warming climate. Note that the relationship between and IWP for NSW6 does not change when the atmosphere warms, because in NSW6 is fixed; in contrast, for NDW6, this relationship changes in the atmosphere because is predicted. Figure 7b shows the response of the IWP– relationship in NDW6-JJA. In Fig. 7a, NSW6 clearly produces a larger optical depth than NDW6 at the same IWP, and the difference between NSW6 and NDW6 becomes larger as IWP increases. The contribution of snow is nonnegligible in NICAM. Figure 7b indicates that decreases at the same IWP when the atmosphere warms. This finding suggests that becomes larger when the atmosphere warms. Our results also show that the global mean becomes larger when the atmosphere warms (figures not shown).

Fig. 7.

The relationship between IWP and . (a) The results of each analysis. NSW6 is shown in red. NDW6-JJA, which is the result of cloud ice and snow for the present climate, is shown in purple. NDW6-JJA, the results from cloud ice and snow in the warming climate, is shown in yellow. NDW6-JJA, which is the result for snow in the present climate, is shown in blue. Finally, NDW6-JJA, the results from snow during the warming climate, is shown in green. (b) The response to global warming for NDW6-JJA. The change resulting from cloud ice and snow is plotted in blue and that attributed to snow is plotted in green. Note that the black vertical dotted-and-dashed line indicates the upper limit of cirrus clouds.

Fig. 7.

The relationship between IWP and . (a) The results of each analysis. NSW6 is shown in red. NDW6-JJA, which is the result of cloud ice and snow for the present climate, is shown in purple. NDW6-JJA, the results from cloud ice and snow in the warming climate, is shown in yellow. NDW6-JJA, which is the result for snow in the present climate, is shown in blue. Finally, NDW6-JJA, the results from snow during the warming climate, is shown in green. (b) The response to global warming for NDW6-JJA. The change resulting from cloud ice and snow is plotted in blue and that attributed to snow is plotted in green. Note that the black vertical dotted-and-dashed line indicates the upper limit of cirrus clouds.

To understand the relationship between the changes in and the LWCRF, we use a diagnostic method for cirrus cloud emissivity, as proposed by Fu and Liou (1993):

 
formula

where is the emissivity, a = 0.79, τi is given as a function of IWP, and is as given in Eq. (1). Figure 8 illustrates how varies as a function of IWP. As in Fig. 7, does not change in NSW6 in response to a warming atmosphere at the same IWP because is fixed. In contrast, changes in NDW6, because is predicted. Similarly to Fig. 7, both the cloud ice and snow relationship between and and that between and of snow in NDW6 shown here reveal the contribution of snow in NICAM. The contribution from snow appears to be important. The response of to the warming atmosphere in NDW6 is plotted in Fig. 8b. In Fig. 8a, ε simulated with NDW6 is smaller than that with NSW6 at the same IWP in the region where IWP is larger than 10 g m−2. The emissivity almost saturates as ε = 0.99 at IWP = 150 g m−2 for NSW6 and IWP = 450 g m−2 for NDW6. As shown in Fig. 8b, ε for NDW6 decreases when the atmosphere warms. As before, this is because of the increase in in response to the warming atmosphere for NDW6.

Fig. 8.

As in Fig. 7, but for the relation between IWP and ε.

Fig. 8.

As in Fig. 7, but for the relation between IWP and ε.

The reason why becomes larger when the atmosphere warms is unclear: it may be related to a change in the tropical circulation under warming conditions. The tropical circulation has been reported to become weaker as the atmosphere warms (Vecchi and Soden 2007). It can be speculated that, if the tropical circulation in NICAM becomes weaker under warming conditions (Satoh et al. 2012, 2015), ice nucleation becomes less active and collisional growth occurs more efficiently among ice particles with slower vertical flow. Consequently, the ice particles grow larger and becomes larger. This issue is beyond the scope of the present analysis and may be explored in detail in the near future.

When discussing the and values of cirrus in the tropics, Figs. 7 and 8 demonstrate that, of the different hydrometeors, snow plays an important role in contributing to and . Note that this large contribution is related to the definition of the snow category in NDW6. In NDW6, snow is categorized as aggregates and originates from aggregation between cloud ice particles (Seiki and Nakajima 2014). As the definition of snow categories is different between models, the quantity of precipitating hydrometeors such as snow is still uncertain in models.

We determined that the NDW6 simulation exhibits a change in the COD to global warming at the same IWP because of the change of . Subsequently, we estimate how LWCRF changes when changes. Assuming a single layer radiative transfer model and neglecting atmospheric absorption, LWCRF can be roughly estimated as follows (Liou 2002):

 
formula

where Ts and Tc are the surface temperature and the cloud-top temperature, respectively, η is the cloud fraction, and σ is the Stefan–Boltzmann constant, W m−2 K−4. To evaluate how ε, the cloud fraction, and the surface temperature response to global warming affect the LWCRF response to global warming, the change in LWCRF can be deconstructed as

 
formula

For the NDW6 simulation, the values of η and ε of cirrus clouds in the present climate are 18.96% and 0.49, respectively, and the response to global warming is 1.54% and −0.03, respectively. The value of Ts in the present climate and its response to global warming are approximately estimated to be 300 K and 4 K, respectively. We incorporate the fixed anvil temperature hypothesis (FAT; Hartmann and Larson 2002), in which Tc is assumed to be unchanged under the global warming condition. With these values, it is found that the portions of attributed to the changes in η, ε, and Ts are estimated as 2.7, −2.08, and 2.28 W m−2, respectively. These values imply that the impact of the change in ε of cirrus clouds is comparable to the impact of the change in η, but with an opposite sign. These results indicate that the change of LWCRF is mainly dominated by the change of Ts. Although the change of LWCRF due to the change in ε, under the global warming condition, is cancelled by the effect of cloud fraction change the effect of ε is nonnegligible and causes negative feedback of LWCRF. We found that effective radius prediction of ice hydrometeors is important for a more accurate evaluation of LWCRF in the climate projection. In addition, the contribution of the precipitating ice hydrometeor, which is categorized as snow in this study, to the LW radiative effect is found to be nonnegligible. This finding is informative for the climate modeling community using conventional GCMs, because with a few exceptions, most GCMs do not predict precipitating solid hydrometeors (e.g., Gettelman et al. 2008, 2010; Morrison and Gettelman 2008; Salzmann et al. 2010).

5. Concluding remarks

This study examined the cloud feedbacks simulated by NICAM with two different cloud microphysics schemes, namely, the single-moment bulk cloud microphysics scheme NSW6 and the double-moment cloud microphysics scheme NDW6. This study is the first attempt to investigate how clouds respond to a warming atmosphere and, as a result, how cloud feedbacks change when single- and double-moment cloud microphysics schemes are implemented in the global nonhydrostatic model with explicit cloud processes. In calculating cloud feedbacks, we used the radiative kernel proposed by Zelinka et al. (2012) to facilitate consistent comparison of our results with the CMIP5/CFMIP2 results. Our main results are as follows: 1) NICAM shows a large response in cirrus clouds to the warming atmosphere with both cloud microphysics schemes, NSW6 and NDW6, and 2) the large LW cloud feedbacks in NICAM can be mainly attributed to an increase in cirrus clouds in the tropics (30°S–30°N), although the increase of cirrus in NDW6 extends to midlatitudes.

To reveal the characteristic behaviors of simulated cirrus clouds by NICAM, we further investigated the microphysical and optical properties of cirrus clouds. The simulation with NDW6 shows a prominent improvement in cloud ice simulation, although both simulations (NSW6 and NDW6) underestimate IWP in the tropics compared with observations. We deconvoluted the factors contributing to LWCRF responses to global warming in the tropics and found that due to cirrus clouds mainly originates from a cloud fraction change but is partly reduced by the decrease in emissivity through increased Rei. The fixed Rei assumed in NSW6 possibly causes systematic positive biases in estimating LW feedbacks under the global warming condition. We suggest that accurate estimations of the ice effective radius and optical properties (e.g., emissivity) including precipitating hydrometeors are issues that should be addressed.

Acknowledgments

Our thanks go to three anonymous reviewers for their constructive comments. We thank Dr. Mark D. Zelinka for providing cloud radiative kernel and CMIP5/CFMIP5 analysis results. This work is supported by the Program for Risk Information on Climate Change (SOUSEI) and Strategic Program for Innovative Research (SPIRE) Field 3 funded by Ministry of Education, Culture, Sports, Science, and Technology, Japan (MEXT). The NICAM simulations of NDW6 are performed on the Earth Simulator at Japan Agency for Marine-Earth Science and Technology (JAMSTEC), and the NICAM simulations of NSW6 are performed on the K computer at the RIKEN/AICS (Proposal number: hp120279, hp130010, and hp140219).

APPENDIX

Season Variation Estimation

In this appendix, we address the seasonal variation of the cloud fraction change and the LW cloud radiative feedbacks in NICAM by using the 20-yr AMIP-type experiment (Kodama et al. 2015). Figure A1 plots the seasonal global mean visible cloud fraction for NSW6 and NDW6, and compares the results with the 20-yr NICAM-AMIP experiment (Kodama et al. 2015) to address the seasonal variation in visible cloud fraction. In this figure, the cloud fraction in NSW6 in the present climate fluctuates between 53% and 54%, while the 20-yr mean cloud fraction in NICAM-AMIP experiment fluctuates between 53% and 56% with the standard deviation around 1%. Figure A2 compares the seasonal variation of the LW cloud radiative feedbacks performed by NSW6, NDW6, and the 20-yr NICAM-AMIP experiment. NICAM-AMIP data are divided into four (boreal) seasons: spring [March–May (MAM)], summer [June–August (JJA)], autumn [September–November (SON)], and winter [December–February (of the next year; DJF)]. The annual variation of each season fluctuates from 0.2 to 0.3 W m−2 K−1 in NICAM-AMIP. Also, the seasonal variation of cloud feedbacks, which can be also treated as the seasonal variation of cloud responses to the global warming, is around 2%, as shown in Fig. A1. Since that the annual variation of cloud feedbacks is in the range from 0.2 to 0.3, one may expect similar annual variation may also be found in NSW6 and NDW6 when the long-period integration is available for these two experiments.

Fig. A1.

The visible cloud fraction in each season of NSW6 (blue) and NICAM-AMIP (green) and in boreal summer of NDW6 (red). The standard deviation of each season in NICAM-AMIP are shown as vertical bars and the small green dots are the result of each year in NICAM-AMIP.

Fig. A1.

The visible cloud fraction in each season of NSW6 (blue) and NICAM-AMIP (green) and in boreal summer of NDW6 (red). The standard deviation of each season in NICAM-AMIP are shown as vertical bars and the small green dots are the result of each year in NICAM-AMIP.

Fig. A2.

As in Fig. A1, but for the LW cloud feedbacks attributed by all clouds.

Fig. A2.

As in Fig. A1, but for the LW cloud feedbacks attributed by all clouds.

REFERENCES

REFERENCES
Austin
,
R. T.
, and
G. L.
Stephens
,
2001
:
Retrieval of stratus cloud microphysical parameters using millimeter-wave radar and visible optical depth in preparation for CloudSat: 1. Algorithm formulation
.
J. Geophys. Res.
,
106
,
28 233
28 242
, doi:.
Austin
,
R. T.
,
A. J.
Heymsfield
, and
G. L.
Stephens
,
2009
:
Retrieval of ice cloud microphysical parameters using the CloudSat millimeter-wave radar and temperature
.
J. Geophys. Res.
,
114
,
D00A23
, doi:.
Berry
,
E.
, and
G. G.
Mace
,
2014
:
Cloud properties and radiative effects of the Asian summer monsoon derived from A-Train data
.
J. Geophys. Res. Atmos.
,
119
,
9492
9508
, doi:.
Bretherton
,
C. S.
,
2015
:
Insights into low-latitude cloud feedbacks from high-resolution models
.
Philos. Trans. Roy. Soc. London
,
373A
, 20140415, doi:.
Collins
,
W. D.
, and
M.
Satoh
,
2009
: Simulating global clouds, past, present, and future. Clouds in the Perturbed Climate Systems: Their Relationship to Energy Balance, Atmospheric Dynamics, and Precipitation, J. Heintzenberg. and T. J. Charlson. Eds., MIT Press, 469–486.
Dufresne
,
J.-L.
, and
S.
Bony
,
2008
:
An assessment of the primary sources of spread of global warming estimates from coupled atmosphere–ocean models
.
J. Climate
,
21
,
5135
5144
, doi:.
Fasullo
,
J. T.
, and
K. E.
Trenberth
,
2012
:
A less cloudy future: The role of subtropical subsidence in climate sensitivity
.
Science
,
338
,
792
794
, doi:.
Fu
,
Q.
, and
K. N.
Liou
,
1993
:
Parameterization of radiative properties of cirrus clouds
.
J. Atmos. Sci.
,
50
,
2008
2025
, doi:.
Gettelman
,
A.
,
H.
Morrison
, and
S. J.
Ghan
,
2008
:
A new two-moment bulk stratiform cloud microphysics scheme in the Community Atmospheric Model, version 3 (CAM3). Part II: Single-column and global results
.
J. Climate
,
21
,
3660
3679
, doi:.
Gettelman
,
A.
, and Coauthors
,
2010
:
Global simulations of ice nucleation and ice supersaturation with an improved cloud scheme in the Community Atmosphere Model
.
J. Geophys. Res.
,
115
,
D18216
, doi:.
Hartmann
,
D. L.
, and
K.
Larson
,
2002
:
An important constraint on tropical cloud–climate feedback
.
Geophys. Res. Lett.
,
29
,
1951
, doi:.
Hartmann
,
D. L.
,
M. E.
Ockert-Bell
, and
M. L.
Michelsen
,
1992
:
The effect of cloud type on Earth’s energy balance: Global analysis
.
J. Climate
,
5
,
1281
1304
, doi:.
Hashino
,
T.
,
M.
Satoh
,
Y.
Hagihara
,
T.
Kubota
,
T.
Matsui
,
T.
Nasuno
, and
H.
Okamoto
,
2013
:
Evaluating cloud microphysics from NICAM against CloudSat and CALIPSO
.
J. Geophys. Res.
,
118
,
7273
7292
, doi:.
Iga
,
S.
,
H.
Tomita
,
Y.
Tsushima
, and
M.
Satoh
,
2011
:
Sensitivity of Hadley circulation of physical parameters and resolution through changing upper-tropospheric ice clouds using a global cloud-system resolving model
.
J. Climate
,
24
,
2666
2679
, doi:.
Igel
,
A. L.
,
M. R.
Igel
, and
S. C.
van den Heever
,
2015
:
Make it double? Sobering results from simulations using single-moment microphysics schemes
.
J. Atmos. Sci.
,
72
,
910
925
, doi:.
Klein
,
S. A.
, and
C.
Jakob
,
1999
:
Validation and sensitivities of frontal clouds simulated by ECMWF model
.
Mon. Wea. Rev.
,
127
,
2514
2531
, doi:.
Klein
,
S. A.
,
Y.
Zhang
,
M. D.
Zelinka
,
R.
Pincus
,
J.
Boyle
, and
P. J.
Gleckler
,
2013
:
Are climate model simulations of clouds improving? An evaluation using the ISCCP simulator
.
J. Geophys. Res. Atmos.
,
118
,
1329
1342
, doi:.
Kodama
,
C.
,
A. T.
Noda
, and
M.
Satoh
,
2012
:
An assessment of the cloud signals simulated by NICAM using ISCCP, CALIPSO, and CloudSat satellite simulators
.
J. Geophys. Res.
,
117
,
D12210
, doi:.
Kodama
,
C.
, and Coauthors
,
2015
:
A 20-year climatology of a NICAM AMIP-type simulation
.
J. Meteor. Soc. Japan
,
93
,
393
424
, doi:.
Li
,
J.-L. F.
, and Coauthors
,
2012
:
An observationally based evaluation of cloud ice water in CMIP3 and CMIP5 GCMs and contemporary reanalyses using contemporary satellite data
.
J. Geophys. Res.
,
117
,
D16105
, doi:.
Li
,
J.-L. F.
,
D. E.
Waliser
,
G.
Stephens
,
S.
Lee
,
T.
L’Ecuyer
,
S.
Kato
,
N.
Loeb
, and
H.-Y.
Ma
,
2013
:
Characterizing and understanding radiation budget biases in CMIP3/CMIP5 GCMs, contemporary GCM, and reanalysis
.
J. Geophys. Res. Atmos.
,
118
,
8166
8184
, doi:.
Liou
,
K. N.
,
2002
: An Introduction to Atmospheric Radiation. 2nd ed. Academic Press, 583 pp.
Mauritsen
,
T.
, and
B.
Stevens
,
2014
:
Missing iris effect as a possible cause of muted hydrological change and high climate sensitivity in models
.
Nat. Geosci.
,
8
,
346
351
, doi:.
Milbrandt
,
J. A.
, and
M. K.
Yau
,
2005
:
A multimoment bulk microphysics parameterization. Part I: Analysis of the role of the spectral shape parameter
.
J. Atmos. Sci.
,
62
,
3051
3064
, doi:.
Mizuta
,
R.
,
Y.
Adachi
,
S.
Yukimoto
, and
S.
Kusunoki
,
2008
: Estimation of the future distribution of sea surface temperature and sea ice using the CMIP3 multi-model ensemble mean. MRI Tech. Rep. 56, 28 pp., doi:.
Morrison
,
H.
, and
A.
Gettelman
,
2008
:
A new two-moment bulk stratiform cloud microphysics scheme in the Community Atmospheric Model, version 3 (CAM3). Part I: Description and numerical tests
.
J. Climate
,
21
,
3642
3659
, doi:.
Noda
,
A. T.
,
K.
Oouchi
,
M.
Satoh
, and
H.
Tomita
,
2012
:
Quantitative assessment of diurnal variation of tropical convection simulated by a global nonhydrostatic model without cumulus parameterization
.
J. Climate
,
25
,
5119
5134
, doi:.
Noda
,
A. T.
,
M.
Satoh
,
Y.
Yamada
,
C.
Kodama
, and
T.
Seiki
,
2014
:
Responses of tropical and subtropical high-cloud statistics to global warming
.
J. Climate
,
27
,
7753
7768
, doi:.
Noda
,
A. T.
,
M.
Satoh
,
Y.
Yamada
,
C.
Kodama
,
T.
Miyakawa
, and
T.
Seiki
,
2015
:
Cold and warm rain simulated using a global nonhydrostatic model without cumulus parameterization, and their responses to global warming
.
J. Meteor. Soc. Japan
,
93
,
181
197
, doi:.
Noda
,
A. T.
,
T.
Seiki
,
M.
Satoh
, and
Y.
Yamada
,
2016
:
High cloud size dependency in the applicability of the fixed anvil temperature hypothesis using global nonhydrostatic simulations
.
Geophys. Res. Lett.
,
43
,
2307
2314
, doi:.
Rayner
,
N. A.
,
D. E.
Parker
,
E. B.
Horton
,
C. K.
Folland
,
L. V.
Alexander
,
D. P.
Rowell
,
E. C.
Kent
, and
A.
Kaplan
,
2003
:
Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century
.
J. Geophys. Res.
,
108
,
4407
, doi:.
Reynolds
,
R. W.
,
N. A.
Rayner
,
T. M.
Smith
,
D. C.
Stokes
, and
W. Q.
Wang
,
2002
:
An improved in situ and satellite SST analysis for climate
.
J. Climate
,
15
,
1609
1625
, doi:.
Roh
,
W.
, and
M.
Satoh
,
2014
:
Evaluation of precipitating hydrometeor parameterizations in a single-moment bulk microphysics scheme for deep convective systems over the tropical central Pacific
.
J. Atmos. Sci.
,
71
,
2654
2673
, doi:.
Rossow
,
W. B.
, and
R. A.
Schiffer
,
1999
:
Advances in understanding clouds from ISCCP
.
Bull. Amer. Meteor. Soc.
,
80
,
2261
2287
, doi:.
Salzmann
,
M.
,
Y.
Ming
,
J.-C.
Golaz
,
P. A.
Ginoux
,
H.
Morrison
,
A.
Gettelman
,
M.
Krämer
, and
L. J.
Donner
,
2010
:
Two-moment bulk stratiform cloud microphysics in the GFDL AM3 GCM: Description, evaluation, and sensitivity tests
.
Atmos. Chem. Phys.
,
10
,
8037
8064
, doi:.
Satoh
,
M.
,
T.
Matsuno
,
H.
Tomita
,
H.
Miura
,
T.
Nasuno
, and
S.
Iga
,
2008
:
Nonhydrostatic Icosahedral Atmospheric Model (NICAM) for global cloud resolving simulations
.
J. Comput. Phys.
,
227
,
3486
3514
, doi:.
Satoh
,
M.
,
T.
Inoue
, and
H.
Miura
,
2010
:
Evaluations of cloud properties of global and local cloud system resolving models using CALIPSO and CloudSat simulators
.
J. Geophys. Res.
,
115
,
D00H14
, doi:.
Satoh
,
M.
,
S.
Iga
,
H.
Tomita
,
Y.
Tsushima
, and
A. T.
Noda
,
2012
:
Response to upper clouds in global warming experiments obtained using a global nonhydrostatic model with explicit cloud processes
.
J. Climate
,
25
,
2178
2191
, doi:.
Satoh
,
M.
, and Coauthors
,
2014
:
The Non-hydrostatic Icosahedral Atmospheric Model: Description and development
.
Prog. Earth Planet. Sci.
,
1
(
18
), doi:.
Satoh
,
M.
,
Y.
Yamada
,
M.
Sugi
,
C.
Kodama
, and
A. T.
Noda
,
2015
:
Constraint on future change in global frequency of tropical cyclones due to global warming
.
J. Meteor. Soc. Japan
,
93
,
489
500
, doi:.
Schiffer
,
R. A.
, and
W. B.
Rossow
,
1983
:
The International Satellite Cloud Climatology Project (ISCCP): The first project of the World Climate Research Programme
.
Bull. Amer. Meteor. Soc.
,
64
,
779
784
.
Seiki
,
T.
, and
T.
Nakajima
,
2014
:
Aerosol effects of the condensation process on a convective cloud simulation
.
J. Atmos. Sci.
,
71
,
833
853
, doi:.
Seiki
,
T.
,
M.
Satoh
,
H.
Tomita
, and
T.
Nakajima
,
2014
:
Simultaneous evaluation of ice cloud microphysics and non-sphericity of the cloud optical properties using hydrometeor video sonde and radiometer sonde in-situ observations
.
J. Geophys. Res. Atmos.
,
119
,
6681
6701
, doi:.
Seiki
,
T.
,
C.
Kodama
,
A. T.
Noda
, and
M.
Satoh
,
2015a
:
Improvements in global cloud-system-resolving simulations by using a double-moment bulk cloud microphysics scheme
.
J. Climate
,
28
,
2405
2419
, doi:.
Seiki
,
T.
,
C.
Kodama
,
M.
Satoh
,
T.
Hashino
,
Y.
Hagihara
, and
H.
Okamoto
,
2015b
:
Vertical grid spacing necessary for simulating tropical cirrus clouds with a high-resolution atmospheric general circulation model
.
Geophys. Res. Lett.
,
42
,
4150
4157
, doi:.
Sekiguchi
,
M.
, and
T.
Nakajima
,
2008
:
A k-distribution-based radiation code and its computational optimization for an atmospheric general model
.
J. Quant. Spectrosc. Ra.
,
109
,
2779
2793
, doi:.
Shell
,
K. M.
,
J. T.
Kiehl
, and
C. A.
Shields
,
2008
:
Using the radiative kernel technique to calculate climate feedbacks in NCAR’s Community Atmospheric Model
.
J. Climate
,
21
,
2269
2282
, doi:.
Sherwood
,
S. C.
,
S.
Bony
, and
J.-L.
Dufresne
,
2014
:
Spread in model climate sensitivity traced to atmospheric convective mixing
.
Nature
,
505
,
37
42
, doi:.
Shipway
,
B. J.
, and
A. A.
Hill
,
2012
:
Diagnosis of systematic differences between multiple parametrizations of warm rain microphysics using a kinematic framework
.
Quart. J. Roy. Meteor. Soc.
,
138
,
2196
2211
, doi:.
Soden
,
B. J.
, and
I. M.
Held
,
2006
:
An assessment of climate feedbacks in coupled ocean–atmosphere models
.
J. Climate
,
19
,
3354
3360
, doi:.
Soden
,
B. J.
,
I. M.
Held
,
R.
Colman
,
K. M.
Shell
,
J. T.
Kiehl
, and
C. A.
Shields
,
2008
:
Quantifying climate feedbacks using radiative kernels
.
J. Climate
,
21
,
3504
3520
, doi:.
Tomita
,
H.
,
2008
:
New microphysical schemes with five and six categories by diagnostic generation of cloud ice
.
J. Meteor. Soc. Japan
,
86A
,
121
142
, doi:.
Tomita
,
H.
, and
M.
Satoh
,
2004
:
A new dynamical framework of nonhydrostatic global model using the icosahedral grid
.
Fluid Dyn. Res.
,
34
,
357
400
, doi:.
Tsushima
,
Y.
,
S.
Iga
,
H.
Tomita
,
M.
Satoh
,
A. T.
Noda
, and
M. J.
Webb
,
2014
:
High cloud increase in a perturbed SST experiment with a global nonhydrostatic model including explicit convective process
.
J. Adv. Model. Earth Syst.
,
6
,
571
585
, doi:.
Vecchi
,
G. A.
, and
B. J.
Soden
,
2007
:
Global warming and the weakening of the tropical circulation
.
J. Climate
,
20
,
4316
4340
, doi:.
Vial
,
J.
,
J.-L.
Dufresne
, and
S.
Bony
,
2013
:
On the interpretation of inter-model spread in CMIP5 climate sensitivity estimates
.
Climate Dyn.
,
41
,
3339
3362
, doi:.
Waliser
,
D. E.
, and Coauthors
,
2009
:
Cloud ice: A climate model challenge with signs and expectations of progress
.
J. Geophys. Res.
,
114
,
D00A21
, doi:.
Webb
,
M.
,
C.
Senior
,
S.
Bony
, and
J. J.
Morcrette
,
2001
:
Combining and ERBE and ISCCP data to assess clouds in the Hadley Centre, ECMWF and LMD atmospheric climate models
.
Climate Dyn.
,
17
,
905
922
, doi:.
Williams
,
K. D.
, and Coauthors
,
2013
:
The Transpose-AMIP II experiment and its application to the understanding of Southern Ocean cloud biases in climate models
.
J. Climate
,
26
,
3258
3274
, doi:.
Yamada
,
Y.
,
K.
Oouchi
,
M.
Satoh
,
H.
Tomita
, and
W.
Yanase
,
2010
:
Projection of changes in tropical cyclone activity and cloud height due to greenhouse warming: Global cloud-system-resolving approach
.
Geophys. Res. Lett.
,
37
,
L07709
, doi:.
Zelinka
,
M. D.
,
S. A.
Klein
, and
D. L.
Hartmann
,
2012
:
Computing and partitioning cloud feedbacks using cloud property histograms. Part I: Cloud radiative kernels
.
J. Climate
,
25
,
3715
3735
, doi:.
Zelinka
,
M. D.
,
S. A.
Klein
,
K. E.
Taylor
,
T.
Andrews
,
M. J.
Webb
,
J. M.
Gregory
, and
P. M.
Forster
,
2013
:
Contributions of different cloud types to feedbacks and rapid adjustments in CMIP5
.
J. Climate
,
26
,
5007
5027
, doi:.
Zhou
,
C.
,
A. E.
Dessler
,
M. D.
Zelinka
,
P.
Yang
, and
T.
Wang
,
2014
:
Cirrus feedback on interannual climate fluctuations
.
Geophys. Res. Lett.
,
41
,
9166
9173
, doi:.

Footnotes

*

Current affiliation: Japan Agency for Marine-Earth Science and Technology, Yokohama, Japan.