1. Introduction
Recently, changes in low clouds in a warmer climate and low cloud feedback have been intensively studied (e.g., Blossey et al. 2013; Bretherton and Blossey 2014; Brient and Bony 2013; Kawai 2012; Webb and Lock 2013; Zhang et al. 2013). In addition, there have been a number of studies examining changes in indices of meteorological factors that can be related to low cloud feedback. For example, Brient and Bony (2013) discussed changes in the vertical gradient of moist static energy (MSE) to explain low cloud changes, while Webb et al. (2015) showed that models with few midlevel clouds or low MSE near the top of the boundary layer (BL) tend to have a large positive tropical cloud feedback.
Wood and Bretherton (2006) showed that estimated inversion strength (EIS), which is a modification of lower-tropospheric stability (LTS; Klein and Hartmann 1993), is a physically more plausible and useful index than LTS for determining low cloud cover (LCC) in the present climate (i.e., larger LCC for stronger EIS). This index has been used in many studies. However, Webb et al. (2013) showed that low cloud feedback is positive despite EIS increases in the future climate in almost all models. Qu et al. (2014) proposed that changes in low cloud cover are determined by a linear combination of changes in EIS and sea surface temperature (SST) in the present climate, and that future changes in low cloud cover can be determined from future changes in EIS and SST. Therefore, LCC can decrease even if EIS increases in the future climate, because SST increases significantly.
In this paper, we develop a new index for LCC, the estimated cloud-top entrainment index (ECTEI), which is a modification of EIS and takes into account a cloud-top entrainment (CTE) criterion, using EIS to estimate the temperature gap at the top of the boundary layer [for a preliminary report on this index, see Kawai (2013)]. We argue here that changes in LCC cannot be fully determined by changes in EIS alone, but can be better determined from changes in both EIS and SST based on ECTEI. The relative contributions of changes in EIS and SST can be deduced if we assume that ECTEI determines LCC. LCC changes related to, for instance, the MSE vertical gradient, SST, or midlevel clouds, can also be better understood using ECTEI. The purpose of this paper is to describe the fundamental ideas and relationships deduced from the basic formula for ECTEI.
2. A new stability index for low cloud cover
a. New stability index
1) Concept
ECTEI was developed from the idea of the corrected gap of low-level moist static energy (CGLMSE) index proposed by Kawai and Teixeira (2010).
2) A coefficient in the index
The coefficient Cqgap, which is the ratio of the qt gap at the inversion and the q difference between the surface and 700 hPa, must be determined to calculate ECTEI. Radiosonde observation data during a field campaign [East Pacific Investigation of Climate (EPIC); Bretherton et al. 2004] off the coast of Peru were used to obtain the coefficient. All 3-hourly sounding profiles during 13 and 25 October 2001 were used and the number of profiles used in the statistics was 96. The vertical sampling interval of the radiosonde soundings is about 10 m.
An inversion height is determined as a lowest height where the absolute value of the vertical gradient of potential temperature is larger than 0.3 K hPa−1. The gap in qt at the inversion is calculated from the specific humidity 100 m below and 150 m above the determined inversion height. Specific humidity near these levels and near 700 hPa are averaged over layers of 50-m thickness to obtain robust values. From this analysis, a value of Cqgap = 0.76 was obtained. The standard deviation of the sample is 0.186, and the standard error of the mean (i.e., the estimated standard deviation of the sample average) is 
b. Observational support
We used the observed cloud fraction derived from shipboard observations, as well as stability indices calculated from reanalysis data, to examine the relationships between LCC and stability indices, including ECTEI. The shipboard observation data for 1957–2002, known as the Extended Edited Cloud Report Archive (EECRA; Hahn and Warren 2009), was used to examine the relationships. Stability indices LTS, EIS, and ECTEI were calculated using the 40-yr ECMWF Re-Analysis (ERA-40) data (Uppala et al. 2005). We regarded the combined cloud cover of stratocumulus, stratus, and sky-obscuring fog as LCC in the present study, as in Koshiro and Shiotani (2014). Figure 1 shows the relationships between the LCC and the stability indices LTS, EIS, and ECTEI calculated simply from both sets of seasonal climatology data, where all the data between 60°N and 60°S for all seasons were used. Figure 1 shows that ECTEI has a very strong correlation with LCC, at least at the same level as EIS, with the correlation coefficients being R = 0.32 for LTS, R = 0.87 for EIS, and R = 0.91 for ECTEI.
Frequencies of occurrence of low stratiform cloud cover (combined cloud cover of stratocumulus, stratus, and sky-obscuring fog) sorted by (a) LTS, (b) EIS, and (c) ECTEI (β = 0.23), based on all 5° × 5° resolution seasonal climatology data (1957–2002). Cloud cover data were obtained from EECRA shipboard observations and stability indices were calculated using ERA-40 data. All the data between 60°N and 60°S for all seasons were used. Linear regression lines and the correlation coefficients are shown.
Citation: Journal of Climate 30, 22; 10.1175/JCLI-D-16-0825.1
Although a value of β = 0.23 is used in the calculation of ECTEI, based on Cqgap = 0.76, which is obtained from radiosonde soundings, and k = 0.7 (MacVean and Mason 1990; MacVean 1993), these values have some uncertainties. To examine correlation coefficients between LCC and ECTEI for different values of β, correlation coefficients are obtained by sweeping over a range of values for β (Fig. 2). Note that ECTEI is equivalent to EIS when β = 0 [see Eq. (3)]. Figure 2 shows that the correlation coefficient is highest around β = 0.23. In addition, the 95% confidence interval of β is 0.22–0.24 when we perform a multiple linear regression analysis using LCC as a dependent variable, and EIS and qsurf − q700 as explanatory variables. However, more important and objective information is that the correlation coefficient is relatively high for all values of β shown in Fig. 2. At least, we can say that adding the second term in Eq. (3) to EIS does not reduce the strength of the correlation with LCC, and EIS is not the only option, although EIS has been used as an excellent index for LCC. Although we use β = 0.23, which is derived from the aforementioned concept together with radiosonde observation, in the following discussion, it should be kept in mind that a wide range of β values are consistent with a high correlation with LCC.
Dependency of correlation coefficient between LCC and ECTEI on different value of β [see Eq. (3)]. The LCC data and reanalysis data used to calculate the correlation coefficients are the same as the data used to plot Fig. 1.
Citation: Journal of Climate 30, 22; 10.1175/JCLI-D-16-0825.1
3. Indices for LCC
a. Dependency of indices on sea surface temperature
First, we examined the basic dependency of these indices on SST. A profile of a lower troposphere typical of stratocumulus regions was assumed, which had a dry adiabatic lapse rate of 9.8 K km−1 below the lifting condensation level (LCL), a moist adiabatic lapse rate inside the cloud layer and free atmosphere, a temperature inversion of 4 K, and relative humidity of 80% at the surface and 30% at 700 hPa. See appendix A for details of the profile used. In addition, we assumed that the SST was 1-K higher than the surface air temperature and that β = 0.23. The sensitivities of the results to the profile and the coefficient are discussed later.
Figure 3 shows the dependency of the index differences ECTEI − EIS and LTS − EIS on SST. The mathematical formulations of the index differences, and the qualitative explanation related to the dependency of the differences on SST, are given in appendix B. For a given EIS, LTS is larger (corresponding to more cloud) for a higher SST, and smaller (less cloud) for a lower SST, as is well known. On the other hand, ECTEI is smaller (less cloud) for a higher SST and larger (more cloud) for a lower SST for a given EIS [because of the humidity difference term in Eq. (3)]. Although sensitivity to the profile and the coefficient is discussed in more detail later, sensitivity to relative humidity at 700 hPa only is shown in Fig. 3. The difference between ECTEI and EIS is large for small values of assumed relative humidity at 700 hPa because the second term on the right-hand side of Eq. (3) has a large magnitude. Note that the index difference LTS − EIS does not change with relative humidity at 700 hPa because neither LTS nor EIS takes the humidity profile into account. If ECTEI is a good indicator that is applicable to any time scale of cloud variability, then the dependency of ECTEI − EIS on SST implies that cloud will decrease in the future climate, if EIS does not change or increases by a sufficiently small amount. Climate models suggest that EIS will increase by a small amount with climate warming, but less than LTS (e.g., Qu et al. 2015a; Webb et al. 2013).
Dependency of index differences on SST (K): shown are ECTEI − EIS (left axis; all lines except red line) and LTS − EIS (right axis; red line). A typical profile of the lower troposphere is assumed (see the text for details) with various values of relative humidity at 700 hPa.
Citation: Journal of Climate 30, 22; 10.1175/JCLI-D-16-0825.1
b. LCC dependency on SST
1) Equations
Partial derivatives of ECTEI with respect to SST calculated by differentiating curves in Fig. 3 (left vertical axis). Sensitivities to (a) relative humidity at 700 hPa, (b) temperature gap at the inversion, (c) relative humidity at the surface, and (d) coefficient β are shown. The right vertical axis shows the approximate corresponding partial derivative of LCC with respect to SST by assuming that dLCC/dECTEI = 3.1% K−1. In (d), dLCC/dECTEI = 2.7% and 3.5% K−1 are assumed for β = 0.16 and 0.30, respectively, because different values of β give different slopes in the relationship between LCC and ECTEI based on an analysis using shipboard observation data.
Citation: Journal of Climate 30, 22; 10.1175/JCLI-D-16-0825.1
2) Quantitative discussion
First, as for the total derivative of LCC with respect to EIS in Eq. (6), a value for dLCC/dEIS of 4.7% K−1 was obtained from Fig. 1b using the shipboard observation data, as already shown in Koshiro and Shiotani (2014). This value is not greatly dissimilar to the value of 6% K−1 shown by Wood and Bretherton (2006).
The EIS slope, (∂LCC/∂EIS)SST, approximately corresponds to dLCC/dECTEI from Eq. (8), and this total derivative can be estimated from our analysis in Fig. 1c. This EIS slope was calculated to be 3.1% K−1 from the shipboard observations. Qu et al. (2015b) calculated EIS slopes using satellite-based cloud data together with ERA-Interim data. The obtained EIS slopes are 2.6% K−1 using the International Satellite Cloud Climatology Project (ISCCP) data, 2.4% K−1 using the Pathfinder Atmospheres–Extended (PATMOS-x) data, 3.5% K−1 using the Multiangle Imaging SpectroRadiometer (MISR) data, and 3.3% K−1 using the Moderate Resolution Imaging Spectroradiometer (MODIS) data. Seethala et al. (2015) derived an EIS slope, where temperature advection is also constant, of about 3.4% and 3.9% K−1 based on the ISCCP and PATMOS-x observations, respectively. Our EIS slope is consistent with the slopes from these previous studies whose targets were areas of frequent subtropical stratocumulus occurrence. In addition, Qu et al. (2014) obtained ensemble mean EIS slopes of 1.1% and 1.5% K−1 (the intermodel standard deviations are 1.3 and 1.7) using data from phase 3 (CMIP3; Meehl et al. 2007) and phase 5 of the of the Coupled Model Intercomparison Project (CMIP5; Taylor et al. 2012) model simulations, respectively.
The SST slope, (∂LCC/∂SST)EIS, can be approximately calculated from Eq. (9)—that is, as a product of a total derivative dLCC/dECTEI and a partial derivative (∂ECTEI/∂SST)EIS. The partial derivative can be obtained from Fig. 4 (left vertical axis), and it is roughly −0.29 K K−1 for a typical SST of 295 K and relative humidity at 700 hPa of 30%. Therefore, the SST slope is −0.90% K−1 using the same value for dLCC/dECTEI based on the observation data as in the discussion above. Qu et al. (2015b) found SST slopes of −2.5% K−1 based on ISCCP data, −2.2% K−1 based on PATMOS-x data, −1.4% K−1 based on MISR data, and −1.0% K−1 based on MODIS data. Seethala et al. (2015) calculated SST slopes of about −1.7% and −1.3% K−1 based on ISCCP and PATMOS-x observations, respectively. Our SST slope is again consistent with the slopes reported in these studies. Furthermore, we obtained the SST slopes directly from the EECRA shipboard observation data that were used to plot Fig. 1. The EIS was calculated using ERA-40 data, and SST data were taken from the Hadley Centre Sea Ice and Sea Surface Temperature dataset (HadISST) version 1.1 (Rayner et al. 2003), as in Koshiro and Shiotani (2014). Figure 5 shows that low stratiform cloud cover clearly has dependency on SST for given values of EIS. The SST slopes are from −1.1% to −0.6% K−1 for each EIS bin shown in Fig. 5. These SST slopes obtained directly from the shipboard observation data are also consistent with the SST slope estimated from the formula of ECTEI (−0.90% K−1 at SST = 295 K). As a reference, Qu et al. (2014) obtained SST slopes of −1.1% and −1.2% K−1 (the intermodel standard deviations are 1.1 and 1.0) from the CMIP3 and CMIP5 model simulations, respectively. All of these derivatives are summarized in Table 1.
Scatter diagrams showing relationships between SST and low stratiform cloud cover (combined cloud cover of stratocumulus, stratus, and sky-obscuring fog) for each EIS bin: (top) 0–1-, 1–2-, and 2–3-K bins and (bottom) 3–4-, 4–5-, and 5–6-K bins are shown, based on all 5° × 5° seasonal climatology data (1957–2002). Cloud cover data were obtained from EECRA shipboard observations and EIS was calculated using ERA-40 data. SST data were taken from HadISST version 1.1. All the data between 60°N and 60°S for all seasons were used. Linear regression lines and the correlation coefficients are shown.
Citation: Journal of Climate 30, 22; 10.1175/JCLI-D-16-0825.1
Total derivatives of LCC with respect to EIS, and EIS and SST slopes (% K−1). From the left, derivatives are from ECTEI in our study, analysis using EECRA data in our study, Wood and Bretherton (2006), Qu et al. (2015b) (based on ISCCP, PATMOS-x, MISR, and MODIS observations; these are denoted in parentheses by IS, PA, MI, and MO, respectively, in the table), and Seethala et al. (2015) (based on ISCCP, and PATMOS-x). For partial derivatives from Seethala et al. (2015), temperature advection was also held constant (the derivative values indicated with asterisks are values not exactly the same definition as others). The ensemble mean derivatives of CMIP3 and CMIP5 models by Qu et al. (2014) are also shown in the rightmost column as a reference.
Approximate values of (∂LCC/∂SST)EIS (i.e., SST slopes) are shown on the right vertical axis in Fig. 4 by assuming that dLCC/dECTEI = 3.1% K−1 for reference. In addition, Figs. 4a–d show the sensitivities of the SST slope to relative humidity at 700 hPa, the temperature gap at the inversion, relative humidity at the surface, and the coefficient β, respectively. Figure 4a shows that when relative humidity at 700 hPa is small, the SST slope has a large negative value. It is also clear that the sensitivity of the SST slope to the assumed magnitude of the temperature gap at the inversion (i.e., the vertical temperature profile) is small (Fig. 4b). Figure 4c shows that the magnitude of the sensitivity of the SST slope to the relative humidity at the surface is similar to its sensitivity to relative humidity at 700 hPa. Figure 4c shows that the SST slope has a large negative value for large relative humidity at the surface, in contrast to the sensitivity to relative humidity at 700 hPa. This characteristic can be attributed to the fact that the second term on the right-hand side of Eq. (3) has a large magnitude for a large relative humidity at the surface. The magnitude of the sensitivity of the SST slope to the coefficient β is also similar to its sensitivity to relative humidity at 700 hPa (Fig. 4d), where the values β = 0.16 and 0.30 correspond to Cqgap = 0.53 and 1.0 respectively, if k is constant as 0.70.
c. Changes in LCC and vertical difference of MSE
The second term of ECTEI in Eq. (3) includes the vertical difference of the specific humidity, and this inspired us to connect this term with the vertical difference of MSE.
Equation (15) suggests that the change in LCC is affected by a change in the vertical difference of MSE between 700 hPa and the surface, δ700−surfMSE. The coefficient of the second term on the right-hand side, referred to as the δ700−surfMSE slope, is 0.71% (kJ kg−1)−1 for β = 0.23 and dLCC/dECTEI = 3.1% K−1. These values are similar to the slope of the relationship between the low cloud frequency and the vertical difference of MSE between the inversion and the surface found by Kubar et al. (2011) of 0.74%–2.78% (kJ kg−1)−1, although these slopes do not need to be equivalent because the definitions of the vertical difference are not the same. The important point is that our discussion can provide some possible interpretation of the contribution of the change in δ700−surfMSE to the change in LCC.
4. Discussion
a. Low cloud cover and the new stability index
In the present study, a new stability index, ECTEI, in which a term using the specific humidity difference between 700 hPa and the surface is added to EIS to take into account a CTE criterion, is proposed. It is worth noting that van der Dussen et al. (2015) showed that the specific humidity difference between 700 hPa and the surface affects the liquid water path (LWP) of stratocumulus using a large-eddy simulation (LES) model. In their simulation, LWP is large for a given LTS when a magnitude of the specific humidity difference between 700 hPa and the surface is small (i.e., the relative humidity at 700 hPa is large). Although their result is related to LWP and our study is related to LCC, their results seem to support the importance of the vertical moisture difference for marine boundary layer clouds.
In section 3, we demonstrated that the SST slope in the present climate obtained by, for example, Qu et al. (2014) and Seethala et al. (2015) and the δ700−surfMSE slope calculated by Kubar et al. (2011) can be obtained mathematically, via some approximations, using ECTEI. The essential concept is that the second term (the humidity difference term) of ECTEI is highly correlated with SST and surface MSE via the specific humidity at the surface.
Actually, several studies have found that stratiform low clouds are controlled by meteorological indices other than EIS and SST, including subsidence velocity and horizontal temperature advection (e.g., Myers and Norris 2015, 2016; Seethala et al. 2015). For instance, Myers and Norris (2013) showed that enhanced subsidence promotes reduced LCC for the same value of inversion strength based on satellite cloud data and reanalysis data. Therefore, we cannot conclude that LCC variations are controlled solely by a single index such as EIS or ECTEI. However, the stability indices EIS and LTS are definitely good indices for LCC and have been used in many studies to understand LCC variations. The new ECTEI gives some additional insights for LCC variations by including a humidity difference term. In addition, ECTEI can be useful for interpreting the factors controlling low cloud feedbacks, as described below.
b. Low cloud feedback and the new stability index
Qu et al. (2014) showed that future changes in LCC can be largely explained by the combination of EIS change and SST change, that is, using Eq. (4). Qu et al. (2015b) and Myers and Norris (2016) suggest that EIS and SST are the most important factors for low cloud feedback, although other factors, including subsidence velocity and horizontal temperature advection, can exert some influence. Brient and Bony (2013) suggested that a decrease in low-level clouds in a warmer climate is caused by an increase in the vertical gradient of MSE, based on a framework of an MSE budget. We consider it useful to interpret low cloud feedback using ECTEI because the concepts described in section 3 suggest a unified physical interpretation of the relationships between low cloud feedback, EIS, SST, and vertical gradients in MSE. However, we should consider the possibility that the EIS and SST slopes, as well as the δ700−surfMSE slope obtained in the present study and based on the present climate, may need to be modified to some extent for a discussion related to cloud feedback, as the relationship between LCC and ECTEI might change to some degree under a future climate.
c. Further insights into low cloud feedback
It is possible that a discussion focused on ECTEI could provide additional insights into low cloud feedback if the relationship between changes in LCC and ECTEI can also be applied to future changes. First, Fig. 4 shows that the magnitude of an SST slope, (∂LCC/∂SST)EIS, is larger for a higher SST. This means that a given SST increase will cause a greater reduction in LCC in a future climate, for example, in the subtropics compared with midlatitudes, because SST is higher in the subtropics.
Second, Fig. 4 shows that the SST slope can be affected by relative humidity at 700 hPa (drier RH700 corresponds to larger magnitude of SST slope). This implies that the cloud reduction is greater for low RH700 models for a given SST and a given SST increase. The sensitivity of the SST slope to RH700 is substantial and the slope varies by a factor of 2 for a range of RH700 = 20%–50% (−0.6% K−1 for RH700 = 50% and −1.0% K−1 for RH700 = 20% at SST = 295 K, where a dLCC/dECTEI value of 3.1% K−1 is used). In addition, for comparison, the sensitivity to RH700 is comparable to the dependency of the SST slope on SST itself (−0.4% K−1 for SST = 280 K and −1.1% K−1 for SST = 300 K, where the same value of dLCC/dECTEI is used).
Third, the midlevel cloud fraction should have some correlation with RH700, and MSE700 should have some correlation with q700. Therefore, the SST slope is possibly sensitive to the midlevel cloud fraction and MSE700 (models that have a smaller midlevel cloud fraction or smaller MSE700 tend to have a larger magnitude SST slope) because the SST slope is sensitive to RH700 and q700. Hence, the relationships between low cloud feedback and RH700, MSE700, and midlevel clouds (Webb et al. 2015) among CMIP multimodel simulations may be partly explained through the relationship between in LCC and ECTEI.
The second and the third points above are somewhat speculative and are based on the assumption that climate models have relationships between LCC and ECTEI similar to that in observations. However, Qu et al. (2014) showed that the SST and EIS slopes vary widely among models and this implies that changes in LCC and ECTEI also vary widely. Therefore, it will be necessary in future work to examine in detail the relationships between LCC and ECTEI changes in CMIP models if we are to use ECTEI to understand low cloud feedback in models.
d. Physical meaning of the CTE criterion and ECTEI
The criterion related to cloud-top entrainment [inequality shown in Eq. (1)], which is used as a base in the present study, was originally developed as a pure CTE instability criterion (e.g., Randall 1980). Subsequently, it has been recognized that the criterion, where k ≈ 0.7, is a break-up criterion of low clouds as a result of a balance among several physical processes, such as a production process by radiative cooling and dissipation processes by precipitation production and ventilation effect of boundary layer humidity due to cumulus transport (e.g., Yamaguchi and Randall 2008; Lock 2009; Wood 2012). On the other hand, it is true that the CTE mechanism still plays an important role as a dissipation term of cloud layers and the criterion corresponds to the relative importance of CTE among several competing processes. Although discussion of low cloud feedback based on ECTEI implies that the magnitude of positive low cloud feedback might be attributed to increased importance of dissipation terms (including activated CTE and the cumulus ventilation effect, relative to the production terms, in a future climate), more detailed investigations are necessary for the physical interpretation. Whatever the case, the discussion presented in this study suggests the possibility of integrating various hypotheses related to low cloud feedback based on different indices using one simple index for marine low cloud cover.
e. Constant in the CTE criterion
The present study has used the factor k = 0.70 in the CTE criterion in Eq. (1), which was derived by MacVean and Mason (1990) and validated by numerical simulations undertaken by MacVean (1993). However, the original form of k, which is proposed as in Eq. (13) of MacVean and Mason (1990), has a temperature dependency because the formulation involves temperature derivatives of saturation specific humidity. Figure 6a shows the dependency of factor k on SST, under the assumption of an atmospheric profile commonly used in the present study (see appendix A). The range of k is 0.61–0.74 for the SST range of 275–300 K. Figure 6c shows that the magnitude of a partial derivative of ECTEI with respect to SST is smaller for a temperature-dependent k than for constant k = 0.70. The magnitude of the partial derivative of ECTEI with respect to SST for a temperature-dependent k is smaller by 30% at SST = 293 K than for the constant k, although the value of the temperature-dependent k is almost equal to 0.70 at the SST (Fig. 6a). Figure 6b helps us understand this characteristic. Even though the difference ECTEI − EIS is the same for two cases of k at SST = 293 K, the gradient of the curve of ECTEI − EIS is smaller for temperature-dependent k than for constant k. Figure 6c shows that the SST slope can be smaller than the value discussed in the present study, if the temperature-dependent k of MacVean and Mason (1990) is more realistic than constant k. It is also worth noting that the sensitivity test on β shown in Fig. 4d can be regarded as the sensitivity to different values of the constant k; β = 0.16 and 0.30 correspond to k = 0.79 and 0.61, respectively, if Cqgap is constant as 0.76. We can understand from this sensitivity examination that a smaller (larger) constant k gives a larger (smaller) SST slope. The value of k has some uncertainty and additionally the value can depend on cloud regimes. The results shown in Figs. 4d and 6 can give some information for the use of a value of k different from a constant value 0.70. In addition, different values of k can be considered conceivable given that a relatively wide range of values for β, corresponding to a substantial range of k, give high correlation between LCC and ECTEI (Fig. 2). For reference, when the temperature-dependent k is used (the reference pressure is set to a mandatory level of 925 hPa to calculate k from ERA-40 data), the correlation between LCC and ECTEI is also 0.91, which is equivalent to the case of constant k = 0.70 (for the temperature-dependent k, a derivative dLCC/dECTEI = 3.4% K−1).
(a) Dependency of factor k in the CTE criterion on SST. The form is proposed by MacVean and Mason (1990) in their Eq. (13), and an atmospheric profile commonly used in the present study is assumed (see appendix A) to obtain the SST dependency (light blue line). The reference pressure is set to 950 hPa. A constant k (=0.70) is also shown (blue line). (b) ECTEI − EIS and (c) partial derivatives of ECTEI with respect to SST for the two cases of factor k.
Citation: Journal of Climate 30, 22; 10.1175/JCLI-D-16-0825.1
Although influences from different value of k are discussed in this section, we also consider it necessary to discuss the influence of the liquid water term on the CTE criterion if we are to consider the criterion in more detail. The more accurate representation of the CTE criterion should include a dependence on the availability of liquid water at the top of the cloud layer (e.g., Wood 2012; MacVean and Mason 1990; Duynkerke 1993). Although this issue is not discussed in the present study, it would be also an interesting research topic in the context of low cloud feedback because the liquid water content can change in the future climate.
5. Summary
This paper reports the development of a new index for LCC, ECTEI, which is based on the EIS index and takes into account a CTE criterion. Analysis of shipboard cloud observation data showed that this new ECTEI is strongly correlated with LCC. We have proposed a concept based on the ECTEI that can comprehensively integrate various previous findings on indices related to LCC. We have calculated the slopes of various indices using ECTEI and compared them with those reported in previous observational and model simulation studies, and have demonstrated that the EIS and SST slopes under the present climate, studied by Qu et al. (2014), can be deduced mathematically from ECTEI. Furthermore, the slope of the vertical gradient of MSE under the present climate, shown by Kubar et al. (2011), can be also explained using the concept of ECTEI.
There is a possibility that a change in ECTEI can explain a change in LCC in a future climate if the relationship persists, and hence low cloud feedback can be interpreted using ECTEI. This study proposes a possible explanation for simulated results in CMIP models indicating that low clouds are decreased in the future climate with increased SST in spite of increases in EIS. Physical interpretations of cloud feedback that have been proposed previously, based on the vertical gradients of MSE, EIS, and SST, can be comprehensively integrated using ECTEI, at least qualitatively.
It is also possible that a discussion centered on ECTEI could provide additional insights. For example, cloud reduction in the future climate may be greater in the subtropics than in midlatitudes for the same temperature increase, because the SST slope, (∂LCC/∂SST)EIS, is more negative in high-SST regions (Fig. 4). In addition, our results imply that cloud reduction is greater for low-RH700 models for a given SST and a given SST increase. A more detailed investigation of low cloud feedback based on ECTEI will be the subject of a future study. In addition, we plan to examine the relationships between changes in LCC and ECTEI in CMIP models, in both present-day and climate change experiments.
Acknowledgments
The EECRA data were obtained from the Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, U.S. Department of Energy. The ERA-40 and ERA-Interim data used in this study were provided by ECMWF. EPIC data were downloaded from the EPIC Stratocumulus Integrated Dataset website. This research was partly supported by the “Program for Risk Information on Climate Change” (SOUSEI), the TOUGOU Program, and “Social Implementation Program on Climate Change Adaptation Technology” of the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan. Mark Webb was supported by the Joint UK BEIS/Defra Met Office Hadley Centre Climate Program (GA01101). We thank Dr. Chris Bretherton for insightful comments. We acknowledge Dr. Stephen Klein and Dr. Robert Wood as reviewers and one anonymous reviewer for their detailed constructive and insightful comments. The figures were edited by Rikako Matsumoto.
APPENDIX A
Assumed Profile of a Lower Troposphere
This appendix describes in detail the lower-tropospheric profiles typical for stratocumulus regions that are used in the present study to examine the dependencies of stability indices on SST. First, we need to provide the temperature difference between the surface air temperature and SST. For instance, de Szoeke et al. (2010) showed that the observed surface air temperature was lower than the SST by 0.5–1 K in the southeastern Pacific during Variability of the American Monsoon Systems (VAMOS) Ocean–Cloud–Atmosphere–Land Study (VOCALS) regional experiment in 2008 (Wood et al. 2011). To obtain a rough global value, the temperature difference was calculated from the European Centre for Medium-Range Weather Forecasts (ECMWF) interim reanalysis (ERA-Interim) data (Dee et al. 2011). The data over the oceans between 60°S and 60°N during 1979 and 2008 give an average difference from −1.0 to −1.2 K depending on the month. Therefore, it is assumed in our study that the surface air temperature is 1-K lower than the SST as a rough estimate. However, note that although for simplicity the average value over the oceans is used, the temperature difference does of course vary regionally and seasonally; some regions such as the fog regions over the midlatitude oceans in summer have positive air temperature minus SST differences, while other regions such as cold-air outbreak regions over wintertime midlatitude oceans have much larger negative temperature differences. The given surface pressure is 1013 hPa.
It is assumed that the profile has a dry adiabatic lapse rate of 9.8 K km−1 below the LCL and a moist adiabatic lapse rate inside the cloud layer and free atmosphere. The standard profile used in the present study has a temperature inversion of 4 K, and relative humidity of 80% at the surface and 30% at 700 hPa (sensitivities of the results to the given profile are discussed in the text). The total water content below the inversion is constant and equal to the specific humidity at the surface, and the relative humidity above the inversion is the same as at 700 hPa. The moist adiabatic lapse rate at each height is calculated using the temperature and pressure of each height, and the LCL is also determined uniquely from the given temperature and humidity profiles. To determine the inversion height, the thickness of cloud layer (the distance between the LCL and inversion height) is assumed to be 300 m. Figure A1 shows vertical profiles of temperature and specific humidity for the assumed standard atmospheric structures for SST = 280 and 300 K as examples.
Vertical profiles of (a) temperature (K) and (b) specific humidity (g kg−1) for the assumed standard atmospheric structures used in the present study. Two cases for SST = 280 K (blue) and 300 K (red) are shown. See appendix A for details.
Citation: Journal of Climate 30, 22; 10.1175/JCLI-D-16-0825.1
APPENDIX B
Formulation of Indices for LCC and Dependency of the Differences on SST
REFERENCES
Betts, A. K., and R. Boers, 1990: A cloudiness transition in a marine boundary layer. J. Atmos. Sci., 47, 1480–1497, doi:10.1175/1520-0469(1990)047<1480:ACTIAM>2.0.CO;2.
Blossey, P. N., and Coauthors, 2013: Marine low cloud sensitivity to an idealized climate change: The CGILS LES intercomparison. J. Adv. Model. Earth Syst., 5, 234–258, doi:10.1002/jame.20025.
Bretherton, C. S., and P. N. Blossey, 2014: Low cloud reduction in a greenhouse-warmed climate: Results from Lagrangian LES of a subtropical marine cloudiness transition. J. Adv. Model. Earth Syst., 6, 91–114, doi:10.1002/2013MS000250.
Bretherton, C. S., and Coauthors, 2004: The EPIC 2001 stratocumulus study. Bull. Amer. Meteor. Soc., 85, 967–977, doi:10.1175/BAMS-85-7-967.
Brient, F., and S. Bony, 2013: Interpretation of the positive low-cloud feedback predicted by a climate model under global warming. Climate Dyn., 40, 2415–2431, doi:10.1007/s00382-011-1279-7.
Deardorff, J. W., 1980: Cloud top entrainment instability. J. Atmos. Sci., 37, 131–147, doi:10.1175/1520-0469(1980)037<0131:CTEI>2.0.CO;2.
Dee, D. P., and Coauthors, 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137, 553–597, doi:10.1002/qj.828.
de Szoeke, S. P., C. W. Fairall, D. E. Wolfe, L. Bariteau, and P. Zuidema, 2010: Surface flux observations on the southeastern tropical Pacific Ocean and attribution of SST errors in coupled ocean–atmosphere models. J. Climate, 23, 4152–4174, doi:10.1175/2010JCLI3411.1.
Duynkerke, P. G., 1993: The stability of cloud top with regard to entrainment: Amendment of the theory of cloud-top entrainment instability. J. Atmos. Sci., 50, 495–502, doi:10.1175/1520-0469(1993)050<0495:TSOCTW>2.0.CO;2.
Hahn, C. J., and S. G. Warren, 2009: Extended edited synoptic cloud reports from ships and land stations over the globe, 1952–1996 (2009 update). NDP-026C, Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, doi:10.3334/CDIAC/cli.ndp026c.
Kawai, H., 2012: Examples of mechanisms for negative cloud feedback of stratocumulus and stratus in cloud parameterizations. SOLA, 8, 150–154, doi:10.2151/sola.2012-037.
Kawai, H., 2013: Improvement of a stratocumulus scheme for mid-latitude marine low clouds. CAS/JSC WGNE Research Activities in Atmospheric and Oceanic Modelling/WMO, 43, 4.03–4.04, https://www.wcrp-climate.org/WGNE/BlueBook/2013/individual-articles/04_Kawai_Hideaki_ResActWGNE2013b.pdf.
Kawai, H., and J. Teixeira, 2010: Probability density functions of liquid water path and cloud amount of marine boundary layer clouds: Geographical and seasonal variations and controlling meteorological factors. J. Climate, 23, 2079–2092, doi:10.1175/2009JCLI3070.1.
Klein, S. A., and D. L. Hartmann, 1993: The seasonal cycle of low stratiform clouds. J. Climate, 6, 1587–1606, doi:10.1175/1520-0442(1993)006<1587:TSCOLS>2.0.CO;2.
Koshiro, T., and M. Shiotani, 2014: Relationship between low stratiform cloud amount and estimated inversion strength in the lower troposphere over the global ocean in terms of cloud types. J. Meteor. Soc. Japan, 92, 107–120, doi:10.2151/jmsj.2014-107.
Kubar, T. L., D. E. Waliser, and J. L. Li, 2011: Boundary layer and cloud structure controls on tropical low cloud cover using A-Train satellite data and ECMWF analyses. J. Climate, 24, 194–215, doi:10.1175/2010JCLI3702.1.
Kuo, H.-C., and W. H. Schubert, 1988: Stability of cloud-topped boundary layers. Quart. J. Roy. Meteor. Soc., 114, 887–916, doi:10.1002/qj.49711448204.
Lock, A. P., 2009: Factors influencing cloud area at the capping inversion for shallow cumulus clouds. Quart. J. Roy. Meteor. Soc., 135, 941–952, doi:10.1002/qj.424.
MacVean, M. K., 1993: A numerical investigation of the criterion for cloud-top entrainment instability. J. Atmos. Sci., 50, 2481–2495, doi:10.1175/1520-0469(1993)050<2481:ANIOTC>2.0.CO;2.
MacVean, M. K., and P. J. Mason, 1990: Cloud-top entrainment instability through small-scale mixing and its parameterization in numerical models. J. Atmos. Sci., 47, 1012–1030, doi:10.1175/1520-0469(1990)047<1012:CTEITS>2.0.CO;2.
Meehl, G. A., C. Covey, T. Delworth, M. Latif, B. McAvaney, J. F. B. Mitchell, R. J. Stouffer, and K. E. Taylor, 2007: The WCRP CMIP3 multimodel dataset: A new era in climatic change research. Bull. Amer. Meteor. Soc., 88, 1383–1394, doi:10.1175/BAMS-88-9-1383.
Myers, T. A., and J. R. Norris, 2013: Observational evidence that enhanced subsidence reduces subtropical marine boundary layer cloudiness. J. Climate, 26, 7507–7524, doi:10.1175/JCLI-D-12-00736.1.
Myers, T. A., and J. R. Norris, 2015: On the relationships between subtropical clouds and meteorology in observations and CMIP3 and CMIP5 models. J. Climate, 28, 2945–2967, doi:10.1175/JCLI-D-14-00475.1.
Myers, T. A., and J. R. Norris, 2016: Reducing the uncertainty in subtropical cloud feedback. Geophys. Res. Lett., 43, 2144–2148, doi:10.1002/2015GL067416.
Qu, X., A. Hall, S. A. Klein, and P. M. Caldwell, 2014: On the spread of changes in marine low cloud cover in climate model simulations of the 21st century. Climate Dyn., 42, 2603–2626, doi:10.1007/s00382-013-1945-z.
Qu, X., A. Hall, S. A. Klein, and P. M. Caldwell, 2015a: The strength of the tropical inversion and its response to climate change in 18 CMIP5 models. Climate Dyn., 45, 375–396, doi:10.1007/s00382-014-2441-9.
Qu, X., A. Hall, S. A. Klein, and A. M. DeAngelis, 2015b: Positive tropical marine low-cloud cover feedback inferred from cloud-controlling factors. Geophys. Res. Lett., 42, 7767–7775, doi:10.1002/2015GL065627.
Randall, D. A., 1980: Conditional instability of the first kind upside-down. J. Atmos. Sci., 37, 125–130, doi:10.1175/1520-0469(1980)037<0125:CIOTFK>2.0.CO;2.
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:10.1029/2002JD002670.
Seethala, C., J. R. Norris, and T. A. Myers, 2015: How has subtropical stratocumulus and associated meteorology changed since the 1980s? J. Climate, 28, 8396–8410, doi:10.1175/JCLI-D-15-0120.1.
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.
Uppala, S. M., and Coauthors, 2005: The ERA-40 Re-Analysis. Quart. J. Roy. Meteor. Soc., 131, 2961–3012, doi:10.1256/qj.04.176.
van der Dussen, J. J., S. R. de Roode, S. D. Gesso, and A. P. Siebesma, 2015: An LES model study of the influence of the free tropospheric thermodynamic conditions on the stratocumulus response to a climate perturbation. J. Adv. Model. Earth Syst., 7, 670–691, doi:10.1002/2014MS000380.
Webb, M. J., and A. P. Lock, 2013: Coupling between subtropical cloud feedback and the local hydrological cycle in a climate model. Climate Dyn., 41, 1923–1939, doi:10.1007/s00382-012-1608-5.
Webb, M. J., F. H. Lambert, and J. M. Gregory, 2013: Origins of differences in climate sensitivity, forcing and feedback in climate models. Climate Dyn., 40, 677–707, doi:10.1007/s00382-012-1336-x.
Webb, M. J., and Coauthors, 2015: The impact of parametrized convection on cloud feedback. Philos. Trans. Roy. Soc., 373A, 20140414, doi:10.1098/rsta.2014.0414.
Wood, R., 2012: Stratocumulus clouds. Mon. Wea. Rev., 140, 2373–2423, doi:10.1175/MWR-D-11-00121.1.
Wood, R., and C. S. Bretherton, 2006: On the relationship between stratiform low cloud cover and lower-tropospheric stability. J. Climate, 19, 6425–6432, doi:10.1175/JCLI3988.1.
Wood, R., and Coauthors, 2011: The VAMOS Ocean-Cloud-Atmosphere-Land Study Regional Experiment (VOCALS-REx): Goals, platforms, and field operations. Atmos. Chem. Phys., 11, 627–654, doi:10.5194/acp-11-627-2011.
Yamaguchi, T., and D. A. Randall, 2008: Large-eddy simulation of evaporatively driven entrainment in cloud-topped mixed layers. J. Atmos. Sci., 65, 1481–1504, doi:10.1175/2007JAS2438.1.
Zhang, M., and Coauthors, 2013: CGILS: Results from the first phase of an international project to understand the physical mechanisms of low cloud feedbacks in single column models. J. Adv. Model. Earth Syst., 5, 826–842, doi:10.1002/2013MS000246.
