Abstract

Using long-term (1958–2008) ship-based cloud observations and reanalysis data, interannual variability in the low stratiform cloud (LSC) amount of stratocumulus (Sc), stratus (St), and sky-obscuring fog (FOG) is examined over the summertime North Pacific. The correlation between the LSC amount and the estimated inversion strength is positive but relatively weak, compared with the well-known linear relationship for their seasonal variabilities. This reflects the regional contrast: the correlations are stronger in the southeastern North Pacific (SE NP) and weaker in the northwestern North Pacific (NW NP). Regarding the LSC types, variations in Sc amount are large over the SE NP and correlated with the inferred capping inversion strength. Variations in FOG amount are large over the NW NP and correlated with the inferred surface-based inversion strength. The compensating variations between the Sc and FOG amounts result in an apparent small variation in the total LSC amount in this region. Variations in St amount are small over the whole North Pacific. The increase in the Sc amount over the SE NP is linked to the local cold sea surface temperature (SST) anomalies with a positive feedback, whereas the increase in the FOG amount over the NW NP is related to warm moist advection across the SST front caused by the anticyclonic sea level pressure anomalies over the north-central North Pacific. The former is associated with an El Niño in the preceding winter and the latter with a wavelike teleconnection pattern along the summertime Asian jet.

1. Introduction

Low stratiform clouds (LSCs) are mainly observed over the ocean and have a large negative radiative effect due to their relatively high albedo and cloud-top temperature, which is only slightly below the sea surface temperature (SST). Because of their significant potential impact on Earth’s energy balance, variations in the LSC amount and related climate parameters have been intensively investigated at various time scales. In particular, the empirical seasonal relationship with the inferred strength of the temperature inversion in the lower troposphere is well known. Klein and Hartmann (1993) defined the lower-tropospheric stability (LTS) as the difference in potential temperature between the 700-hPa level and the surface, which showed a positive correlation with the LSC amount. Wood and Bretherton (2006, hereafter WB06) defined the estimated inversion strength (EIS) as a refinement of the LTS and indicated that a single linear relationship can be applied to the LSC amounts and EISs over the subtropical and midlatitude oceans.

LSCs are composed of three types: stratocumulus (Sc), stratus (St), and sky-obscuring fog (FOG). These occur through distinctly different meteorological processes (Norris 1998a,b; Norris and Klein 2000). Koshiro and Shiotani (2014, hereafter KS14) investigated climatological seasonal relationships between the amounts of each LSC type and EISs over the global ocean. They found that the relationships are clearly divided into two regimes at an SST of approximately 16°C. Sc is the only dominant type and its amount is strongly correlated with EIS in the warm SST regime, whereas the St and FOG amounts increase with EIS in the cold SST regime. This suggests that given the same EIS in both warm and cold SST regimes, the inferred inversion contributing to EIS exists at a different level for each SST regime. Layered EIS (newly proposed by KS14 to consider the vertical difference) and air–sea temperature difference (which indicates temperature advection over the ocean) clearly reveal the following features. The general occurrence of cold advection provides favorable conditions for the formation of Sc and capping inversions in the warm SST regime, whereas warm advection causes an increase in the surface-based inversion strength in the cold SST regime, leading to more FOG.

The interannual relationship between the variabilities in the LSC amount and EIS has not been sufficiently investigated, especially in terms of the LSC types. An LSC-type oriented analysis, such as that performed by KS14, can only be performed using synoptic weather observations by human observers on ships, even though various types of satellite observations are also available. Since ship-based observations have the longest period of record for cloud information, it is important to extend the LSC-type oriented analysis to longer time scales. However, for decadal or much longer time scales, the variations in marine low clouds from the ship-based observations are considered to be mostly spurious, although an exact cause for this has not yet been identified. It is possible that subtle changes in observation procedures over time or in the number of nations contributing ship reports could be responsible for the spurious variations (Bajuk and Leovy 1998; Norris 1999, 2005a; Eastman et al. 2011). Since the long-term spurious changes tend to be zonally and globally coherent, recent studies, especially those looking for tropical climate variability, account for these changes by removing the variations averaged over all considered areas (Deser and Phillips 2006; Clement et al. 2009; Deser et al. 2010; Tokinaga et al. 2012; Bellomo et al. 2014, 2015), as recommended by Norris (2005b). In addition, Norris (2000b) emphasized that it is important to corroborate cloud variability with variations in other physically related climate parameters.

On interannual time scales, previous studies have focused on the variability in the total LSC amount and examined its relationship with SST anomalies to obtain a global perspective. Norris and Leovy (1994) showed that interannual anomalies of the LSC amount are negatively correlated with those of SST over the midlatitude and eastern subtropical oceans, especially during summer. They also noted that the negative correlations are strongest over large gradients of SST and LSC amount, implying an association with advection. They discussed the relationship between the LSC amount and advection in terms of correlations with latent and sensible heat flux anomalies.

Spatial and temporal inhomogeneities of ship-based observations have limited the success of global investigations into interannual variability. Therefore, several studies have focused on interannual variations in the LSC amount and related climate parameters over the summertime North Pacific, where and when LSCs are often observed and ship-based observations are abundant. Klein et al. (1995) discussed the summertime interannual covariation among the LSC amount, SST, and sea level pressure (SLP) fields in the eastern North Pacific in terms of advection over the SST gradients with the subtropical anticyclone. Norris et al. (1998) demonstrated that the summertime interannual variability of the LSC amount over the North Pacific is largest in the central and western regions along 35°N and the eastern region near 15°N. The seesaw pattern of the two variations is imprinted in the leading empirical orthogonal function mode and connected with local SST and the basinwide SLP field; it also suggests an association with advection. Park and Leovy (2004) indicated that the weak but significant cloud anomalies related to El Niño–Southern Oscillation (ENSO) are found over the 35°N zone and associated with latitudinal shifts in the storm track and SST gradient, as shown by Norris (2000a). However, advection over the SST gradients by synoptic disturbances modulates the type and amount of LSC (Klein 1997; Norris and Klein 2000; Xu et al. 2005; Norris and Iacobellis 2005; Tanimoto et al. 2009; Tokinaga et al. 2009). This implies that interannual variation in the LSC amount is the superposition of those in each LSC-type amount and is associated with distinctly different meteorological fields.

In this paper, using long-term ship-based observations, we demonstrate interannual variability in the LSC amount and its relation to the inversion strength measured by EIS over the summertime North Pacific, especially in terms of the LSC types. Details of the data used in this study are described in section 2. In section 3, we present the relationship between the LSC amount and the EIS, and those between the LSC-type amounts and the layered EISs proposed by KS14. In section 4, further analyses are performed for two distinct variations in the LSC-type amounts to reveal more physical connections with the related processes highlighted in previous studies. In section 5, we discuss the possibility of links between these variations and ENSO, which is the dominant variability over the Pacific Ocean on an interannual time scale. Section 6 summarizes the study.

2. Data

The cloud amount was calculated from the ship-based data of the Extended Edited Cloud Report Archive (EECRA; Hahn and Warren 2009; Eastman et al. 2011), which is a quality-controlled compilation of individual synoptic surface cloud observations, including a low cloud condition code CL and a low cloud amount, with some original extensions. Similar to KS14, we constructed 5° × 5° (in this study, the longitudinal width of the grid boxes poleward of 50° was increased to maintain approximately equal areas in the boxes; cf. the so-called “5c” grid boxes in the EECRA documentation) seasonal means of the Sc (CL = 4, 5, and 8), St (CL = 6), FOG (CL = 11), and LSC (Sc + St + FOG; CL = 4, 5, 6, 8, and 11) amounts for individual years from 1958 to 2008. Only the ocean grid boxes were considered. The seasons were defined as follows: December–February (DJF), March–May (MAM), June–August (JJA), and September–November (SON). Since surface observers have difficulty identifying cloudiness under conditions of insufficient illumination (Hahn et al. 1995), only daytime (including twilight) observations were used to construct the seasonal means; however, this may not greatly differ from the true daily average because the maxima and minima in the low cloud amount over the ocean typically occur within this period, as pointed out by KS14.

The summertime (i.e., JJA) North Pacific is one of the best target areas for investigating interannual variabilities in the amounts of the LSC types since the LSC amounts are climatologically large (Fig. 1a) and all three types of LSC are often observed there in that season (see Fig. 4 in KS14). In KS14, although grid boxes of the less than 100 observations that contributed to the average were discarded to reduce the sampling error, the climatological seasonal means of the LSC amount are available over most ocean areas between 60°N and 60°S (Fig. 1a). On the other hand, to investigate the interannual variability, sufficient ship-based observations are required for each season of an individual year. The target areas are therefore limited. The summertime North Pacific is again the best target area because ship-based observations are abundant (Fig. 1b). The latitudinal area between 25° and 55°N of the North Pacific was considered in this study. In this area, almost all grid boxes are available for individual years throughout the data period (Fig. 1c). Here, to reduce the sampling error below 5%, we set the minimum number of observations to 25 to form an average cloud amount for a particular season (Warren et al. 1988; Eastman et al. 2011).

Fig. 1.

(a) Climatological mean LSC amount for the JJA season over the Pacific Ocean. Contour interval is 10%. Values greater than 20% are lightly shaded and values greater than 50% are heavily shaded. The 5° × 5° grid boxes (the longitudinal width of the grid boxes poleward of 50° is increased to maintain approximately equal areas in the boxes) with less than 100 observations contributing to the average are omitted. (b) Total number of ship-based observations in the JJA season from 1958 to 2008 for each grid box over the Pacific Ocean. (c) Number of individual JJA seasons from 1958 to 2008 with ≥25 ship-based observations for each grid box over the Pacific Ocean. Black rectangle designates the North Pacific region considered in this study (25°–55°N, 140°E–120°W).

Fig. 1.

(a) Climatological mean LSC amount for the JJA season over the Pacific Ocean. Contour interval is 10%. Values greater than 20% are lightly shaded and values greater than 50% are heavily shaded. The 5° × 5° grid boxes (the longitudinal width of the grid boxes poleward of 50° is increased to maintain approximately equal areas in the boxes) with less than 100 observations contributing to the average are omitted. (b) Total number of ship-based observations in the JJA season from 1958 to 2008 for each grid box over the Pacific Ocean. (c) Number of individual JJA seasons from 1958 to 2008 with ≥25 ship-based observations for each grid box over the Pacific Ocean. Black rectangle designates the North Pacific region considered in this study (25°–55°N, 140°E–120°W).

The 5° × 5° seasonal means of related meteorological fields were constructed from 1.25° × 1.25° monthly products of the Japanese 55-year Reanalysis (JRA-55; Kobayashi et al. 2015) data, that is, air temperatures, zonal and meridional winds, vertical pressure velocities, geopotential heights on the standard pressure levels, SLP, surface (2 m) air temperature (SAT), surface (10 m) zonal and meridional winds, and upward surface latent and sensible heat fluxes. The 5° × 5° seasonal means of surface scalar wind speed were based on 1.25° × 1.25° 6-hourly snapshot data of the surface zonal and meridional winds. JRA-55 is one of the few reanalyses that consistently cover the period of the ship-based observations, from 1958 to 2008.

The conventional (i.e., bulk) EIS was calculated from the following formula given by WB06:

 
formula

where θ700 and θsfc are the potential temperatures at the 700-hPa level and the surface, respectively, is the moist adiabatic lapse rate at 850 hPa, z700 is the height of the 700-hPa level, and zLCL is the lifting condensation level. In this study, monthly means of the SLP, SAT, and 700-hPa temperature data were used for this calculation. The value of was calculated using the mean of SAT and the 700-hPa temperature, and z700 was calculated assuming an exponential decrease in pressure with height for a single scale height; also, zLCL was calculated from the dry adiabat and the lifting condensation temperature estimated by Bolton (1980), with the surface relative humidity fixed at 0.8 (following the assumption of WB06; the sensitivity of EIS to changes in surface relative humidity is quite small). Moreover, we calculated the layered EISs, as proposed by KS14. Although the layered EISs in KS14 were defined for three divided layers , in this study, and are employed as measures of the capping inversion strength and surface-based inversion strength, respectively, to facilitate a physical understanding:

 
formula

where

 
formula

is the EIS for the layer between 700 and 925 hPa, and

 
formula

is the EIS for the layer between 925 hPa and the surface; θ925 is the potential temperature at 925 hPa, and z925 is the height of the 925-hPa level, which was calculated from an equation similar to that for z700, as described above. If zLCL surpassed z925, z925 was replaced by zLCL in Eqs. (3) and (4); however, this occurrence was rare in this calculation.

The SST data for the same period were taken from the Centennial In Situ Observation-Based Estimates of the Variability of SST and Marine Meteorological Variables (COBE) products (Ishii et al. 2005), which were used as the boundary condition data for the JRA-55. The 5° × 5° monthly zonal and meridional gradients and gridbox averages of SST were calculated from 1° × 1° monthly means of the COBE SST, following the linear regression method described by Klein et al. (1995). Only the open ocean 5° × 5° grid boxes (i.e., those in which all 25 1° × 1° grid boxes were ocean) were considered. The SST gradients were used to estimate advection by surface winds over varying SSTs, the so-called SST advection: , where represents the surface zonal and meridional winds, and represents the zonal and meridional SST gradients (e.g., Klein et al. 1995; Norris and Iacobellis 2005; Mansbach and Norris 2007). A positive value corresponds to warm advection, whereas a negative value corresponds to cold advection. The calculation of SST advection was also based on the 5° × 5° monthly means. The 5° × 5° seasonal means were constructed from the calculated 5° × 5° monthly means.

As mentioned in the introduction, previous studies considered that long-term variations at greater than decadal time scales in marine low clouds based on ship-based observations are mostly spurious (e.g., Bajuk and Leovy 1998; Norris 1999, 2005a; Eastman et al. 2011). To remove the longer time-scale variability, first detrended anomalies were constructed by subtracting the seasonal climatology and least squares linear trend throughout the data period from the seasonal means for individual years. Then, a four-pole Butterworth recursive filter was applied to the detrended anomalies, with a half-power cutoff period of seven years. The obtained low-frequency components were subtracted from the detrended anomalies. In this study, we examine this interannual (i.e., time scales of less than several years) variability for all datasets.

3. Interannual variations in the LSC amount and EIS

a. Relationship between the total LSC amount and bulk EIS

First, we examine the relationship between the LSC amount and the EIS. Figure 2 displays a frequency distribution obtained by classifying each 5° × 5° JJA interannual anomaly into 0.2-K intervals of the EIS and 1% intervals of the LSC amount over the North Pacific area designated in Fig. 1. Although a positive correlation is found, the correlation coefficient of 0.50 is rather weak compared with the quite strong correlation for their seasonal variabilities, with correlation coefficients of approximately 0.9 (WB06; KS14). It is also noted that the interannual standard deviations of the LSC amount and the EIS are 4.2% and 0.62 K, respectively, which are both approximately half of those for the seasonal variabilities in this area (not shown).

Fig. 2.

Frequency of occurrence for EIS and LSC amount intervals for 5° × 5° JJA interannual anomalies over the North Pacific. Dashed line represents one std dev for each axis. Correlation coefficient is shown at the upper-left corner.

Fig. 2.

Frequency of occurrence for EIS and LSC amount intervals for 5° × 5° JJA interannual anomalies over the North Pacific. Dashed line represents one std dev for each axis. Correlation coefficient is shown at the upper-left corner.

Figure 3 shows the geographical distribution of the correlation coefficients between the interannual variations in the LSC amount and the EIS. This indicates that the interannual relationship between the LSC amount and the EIS is generally positive but spatially inhomogeneous. The correlations are stronger in the southeastern region (greater than 0.7 at the peak) and weaker in the northwestern region (near zero at the peak) over the North Pacific, although rather noisy patterns are also observed. The rather weak positive correlation over the entire North Pacific shown in Fig. 2 reflects this regional contrast.

Fig. 3.

Coefficients of local linear correlation between JJA interannual anomalies in LSC amount and EIS over the North Pacific. Contour interval is 0.1. Positive correlations are drawn with solid lines, and zero correlation is thick. Shading indicates correlation coefficients greater than 0.6 (dark) and less than 0.4 (light).

Fig. 3.

Coefficients of local linear correlation between JJA interannual anomalies in LSC amount and EIS over the North Pacific. Contour interval is 0.1. Positive correlations are drawn with solid lines, and zero correlation is thick. Shading indicates correlation coefficients greater than 0.6 (dark) and less than 0.4 (light).

The contrast in the correlation coefficients between the southeastern and northwestern regions in Fig. 3 can be explained by the difference between interannual variances of the LSC amount and the EIS. Figure 4 displays JJA climatological means and interannual standard deviations for the LSC amount and the EIS over the North Pacific. In the southeastern region of the North Pacific, both the LSC amount and the EIS have large interannual variances to the west off the climatological maxima; this may be responsible for their strong correlation, as seen in Fig. 3. However, the variance of the LSC amount is relatively small in the northwestern region of the North Pacific, whereas that of the EIS is large. Their correlations may thus be weak (Fig. 3). This contrasts starkly with the general coincidence of the JJA climatological distributions of the LSC amount and the EIS over the whole North Pacific (contours in Fig. 4). The small interannual variations in the LSC amount over the climatological maxima were demonstrated by Norris et al. (1998). They found that the variability is largest along 35°N in the central and western North Pacific, as also seen in Fig. 4a. In our study, regions equatorward of 25°N, including another maximum they indicated (peaking at 15°N in the eastern North Pacific), are not considered due to an insufficient number of ship observations. The coastal grid boxes are also omitted. However, our result is consistent with their indication that the interannual variability in the LSC amount is large along the edge of the climatological maxima. Nevertheless, we found that the geographical distribution of the interannual variances in the LSC amount does not match those in the EIS, especially over the northwestern region of the North Pacific.

Fig. 4.

JJA climatological means (contours) and interannual std devs (colors) for (a) LSC amount and (b) EIS over the North Pacific. Contour intervals are 5% and 1 K, respectively; negative contours are dashed. Color intervals are as indicated on the right-hand side of each panel.

Fig. 4.

JJA climatological means (contours) and interannual std devs (colors) for (a) LSC amount and (b) EIS over the North Pacific. Contour intervals are 5% and 1 K, respectively; negative contours are dashed. Color intervals are as indicated on the right-hand side of each panel.

b. Relationships between the LSC-type amounts and layered EISs

Next, we divide the LSC amount into three types, and the EIS into two layers, as demonstrated in Fig. 5. This makes it possible to reveal clearer relationships between the LSCs and the inferred temperature inversions in the lower troposphere. Large variances in the Sc amount coincide with those in in the southeastern region of the North Pacific (Figs. 5a,d). These are responsible for the large variances in the total LSC amount and bulk EIS (Figs. 4a,b) because the other types have small variations in this area (Figs. 5b,c,e). On the other hand, large variances in the FOG amount coincide with those in in the northwestern region of the North Pacific (Figs. 5c,e), where the variances in the total LSC amount are quite small (Fig. 4a) despite the large variances in the EIS (Fig. 4b). In this region, the Sc amount also shows relatively large variances (Fig. 5a). These features suggest that the small variation in the total LSC amount may be a result of compensating variations of the Sc and FOG amounts. As KS14 indicated for seasonal variations, given the same interannual variance of EIS in both the southeastern and northwestern regions of the North Pacific, the variability in the inferred inversion contributing to the EIS at a different level is associated with a different LSC type.

Fig. 5.

As in Fig. 4, but for (a) Sc amount, (b) St amount, (c) FOG amount, (d) , and (e) . Thick solid and dashed rectangles designate the SE NP and NW NP regions, respectively, discussed in detail later in this paper.

Fig. 5.

As in Fig. 4, but for (a) Sc amount, (b) St amount, (c) FOG amount, (d) , and (e) . Thick solid and dashed rectangles designate the SE NP and NW NP regions, respectively, discussed in detail later in this paper.

We also found that interannual variations in the St amount are relatively small, and that significant St maxima cannot be found over the whole of the North Pacific (Fig. 5b). Although the reasons for these features are not clear, we do not discuss them further in this paper. In the coastal area off the North American continent, variations in are quite large (Fig. 5e). However, no variation in any LSC-type amount coincides with this feature. In this area, is climatologically large (contours in Fig. 5e), and “no low cloud” (CL = 0) is frequently reported as a low cloud condition; this might be associated with the warm advection of dry air from land (Norris 1998b; KS14).

As described above, Fig. 5 reveals two major variations in the amounts of the LSC types over the North Pacific for the JJA interannual variabilities: one is for the Sc amount in the southeastern region, and the other is for the FOG amount in the northwestern region. Here we define the former region as the southeastern North Pacific (SE NP; 25°–35°N, 130°–160°W) and the latter as the northwestern North Pacific (NW NP; 40°–55°N, 155°E–170°W). Table 1 lists the correlation coefficients between the JJA interannual anomalies of the total LSC and LSC-type amounts and those of the bulk and layered EISs for the two regions. In the SE NP, the correlation between the total LSC amount and the bulk EIS is clearly dominated by those between the Sc amount and ; this indicates that stronger capping inversion coincides well with a greater Sc amount.

Table 1.

Correlation coefficients of JJA interannual anomalies of LSC and LSC-type amounts with those of EIS and layered EISs over the SE NP and NW NP. The absolute values greater than 0.50 are in boldface. The parentheses indicate that the correlation coefficient is not statistically significant at the 99% confidence level.

Correlation coefficients of JJA interannual anomalies of LSC and LSC-type amounts with those of EIS and layered EISs over the SE NP and NW NP. The absolute values greater than 0.50 are in boldface. The parentheses indicate that the correlation coefficient is not statistically significant at the 99% confidence level.
Correlation coefficients of JJA interannual anomalies of LSC and LSC-type amounts with those of EIS and layered EISs over the SE NP and NW NP. The absolute values greater than 0.50 are in boldface. The parentheses indicate that the correlation coefficient is not statistically significant at the 99% confidence level.

In the NW NP, FOG amounts are strongly correlated with , indicating that the stronger surface-based inversion coincides well with more FOG. However, Sc amounts are negatively correlated with . The compensating variations of the Sc and FOG amounts with result in a weak correlation between the total LSC amount and the bulk EIS. On the other hand, unlike for the SE NP, no correlation can be observed between the Sc amount and . In this region, the interannual variability of is relatively small but the climatological values are large (Fig. 5d) because the continents are considerably warmer than the surrounding oceans and a monsoon-like circulation develops with subsidence over the colder oceans (Klein and Hartmann 1993; Kawai et al. 2015). Therefore, if the anomalies are negative, the Sc amount anomalies increase due to the climatologically large .

4. Association with large-scale meteorological processes

To further examine the JJA interannual anomalies for the Sc amount averaged over the SE NP and the FOG amount averaged over the NW NP, we focus on the regional mean time series shown in Fig. 6. As mentioned in the previous section, the Sc amount is strongly correlated with in the SE NP (Fig. 6a), whereas the FOG amount is strongly correlated with and negatively correlated with the Sc amount in the NW NP (Fig. 6b). The correlation coefficients between the dominant LSC-type amounts and the corresponding layered EISs are greater than those for the 5° × 5° gridded data shown in Table 1. As displayed below Fig. 6b, no correlation is observed between the SE NP Sc amount and the NW NP FOG amount, with a correlation coefficient of almost zero. This suggests that the two major interannual variations are independent.

Fig. 6.

(a) Time series of JJA interannual anomalies averaged over the SE NP: Sc amount (black solid) and (black dashed). Correlation coefficient of Sc amount with is also shown. (b) As in (a), but for the NW NP: FOG amount (black solid), Sc amount (gray solid), and (black dashed). Correlation coefficients of FOG and Sc amounts with are also shown. Open and closed circles indicate positive and negative values absolutely greater than the 0.67 std dev, respectively, for SE NP Sc amount and NW NP FOG amount. Correlation coefficient between SE NP Sc amount and NW NP FOG amount is displayed below (b).

Fig. 6.

(a) Time series of JJA interannual anomalies averaged over the SE NP: Sc amount (black solid) and (black dashed). Correlation coefficient of Sc amount with is also shown. (b) As in (a), but for the NW NP: FOG amount (black solid), Sc amount (gray solid), and (black dashed). Correlation coefficients of FOG and Sc amounts with are also shown. Open and closed circles indicate positive and negative values absolutely greater than the 0.67 std dev, respectively, for SE NP Sc amount and NW NP FOG amount. Correlation coefficient between SE NP Sc amount and NW NP FOG amount is displayed below (b).

a. Local processes in the SE NP and NW NP

We also calculated the regional mean time series (similar to Fig. 6) for related meteorological parameters and examined the correlations with the dominant LSC-type amount and the corresponding layered EIS for each region. Table 2 lists the correlation coefficients. In the SE NP, SST is negatively correlated with both the Sc amount and . SAT is also negatively correlated; it basically follows the SST due to heat transfer from the ocean to the atmosphere in this region where cold advection steadily occurs as a result of trade winds.

Table 2.

Correlation coefficients of regional mean interannual anomalies of various meteorological parameters with those of Sc amount and for SE NP, and those of FOG amount and for NW NP. The parentheses indicate that the correlation coefficient is not statistically significant at the 95% confidence level. Boldface indicates 99.9% significance.

Correlation coefficients of regional mean interannual anomalies of various meteorological parameters with those of Sc amount and  for SE NP, and those of FOG amount and  for NW NP. The parentheses indicate that the correlation coefficient is not statistically significant at the 95% confidence level. Boldface indicates 99.9% significance.
Correlation coefficients of regional mean interannual anomalies of various meteorological parameters with those of Sc amount and  for SE NP, and those of FOG amount and  for NW NP. The parentheses indicate that the correlation coefficient is not statistically significant at the 95% confidence level. Boldface indicates 99.9% significance.

Negative correlations with SST advection indicate that enhanced cold advection coincides with large Sc amounts and values. Considering the correlations with each component of the SST advection, the magnitude of the surface wind vector is most important. The correlation coefficients are comparable with those for surface scalar wind speed. It is suggested that the strength of the steady trade winds is closely tied to interannual variations in the Sc amount and the capping inversion strength. On the other hand, no correlations are observed with the surface latent and sensible heat fluxes. These results are consistent with Klein et al. (1995), who demonstrated similar relationships at an ocean weather station over the northeastern Pacific.

It is interesting to note that the parameters related to subsidence, such as SLP, 500-hPa vertical pressure velocity ω500, and 700-hPa air temperature, have clear correlations (correlation coefficients of approximately 0.5–0.6) with , whereas their correlations with the Sc amount are generally weak. Myers and Norris (2013) showed that enhanced subsidence reduces the Sc amount for the same value of inversion strength by pushing down the top of the marine boundary layer (MBL), although enhanced subsidence is typically associated with stronger capping inversion, with stronger inversions favoring greater LSC amounts. Their results consistently support our findings.

In contrast to the case for the SE NP, the FOG amount and in the NW NP have a positive correlation with air–sea temperature difference. SST advection is also positively correlated. Therefore, enhanced warm advection coincides with large FOG amounts and values. Negative correlations with upward surface latent and sensible heat fluxes are consistent with warm advection over the cold ocean. Moreover, warm advection often leads to a slight ascent and warm conditions throughout the lower troposphere; negative correlations with ω500 and positive correlations with the 700-hPa temperature are also consistent. Whereas all the components of SST advection show significantly positive correlations, the relative direction between the wind vector and the SST gradient shows higher correlation coefficients (greater than 0.5) with both the FOG amount and . On the other hand, surface scalar wind speed is not correlated with these factors. These results suggest that wind direction is important for the interannual variabilities of the FOG amount and the surface-based inversion in this region. This is in contrast with the much less importance of wind direction for the Sc amount over the SE NP due to the steadiness of the trade winds.

b. Basin-scale variations over the North Pacific

To reveal the association with variations in these meteorological parameters more widely, we demonstrate the composite difference fields between the means of positive and negative years over the whole North Pacific. For this, we used the Sc amount in Fig. 6a and the FOG amount in Fig. 6b as the reference time series. The composites are made if the anomaly absolutely exceeds the 0.67 standard deviation, which means the highest and lowest 25% of cases for the normal distribution. As a result, 11 positive and 14 negative years for the SE NP Sc amount and 11 positive and 11 negative years for the NW NP FOG amount are used, as shown using open and closed circles in Fig. 6. These composite difference fields show the emphasized geographical distributions of the interannual anomalies of the related meteorological parameters when the Sc or FOG amount anomalies averaged in each reference area are significantly positive (and also mean that the distributions with the opposite signs appear when they are significantly negative). In this context, the composite difference values are also called anomalies.

Figure 7 shows the composite difference fields for the SE NP regional mean time series of the Sc amount interannual anomalies. It is apparent that the Sc amount (Fig. 7a) and (Fig. 7b) anomalies are positively large in the southeastern region of the North Pacific, with maxima in the reference area. Correspondingly, as shown in Table 2, negative SST anomalies occur with a similar distribution (Fig. 7d). Although not shown, air temperature anomalies at the 700-hPa level (i.e., in the free troposphere) are relatively small. Therefore, the relatively lower SST directly contributes to intensify the capping inversion owing to the climatologically well-mixed MBL. Decreases in upward latent heat fluxes from the sea surface due to the colder SST (Fig. 7e) prevent the MBL from deepening and mixing with the free troposphere, leading to an increase in the Sc amount (Bretherton et al. 2013; Qu et al. 2014; Kamae et al. 2016). Since higher-level clouds are much less common in this region, the larger Sc amounts lead to greater reflection of the solar radiation incoming to the surface. Thus, there exists a positive feedback between the Sc amount and the SST, as mentioned in previous studies (e.g., Norris and Leovy 1994; Norris et al. 1998; Clement et al. 2009).

Fig. 7.

JJA composite difference fields between the means of positive and negative years absolutely greater than the 0.67 std dev for the SE NP (thick solid rectangle) regional mean time series of Sc amount interannual anomalies: (a) Sc amount, (b) , (c) SLP (contours) and surface wind velocity (arrows), (d) SST, (e) upward surface latent heat flux, and (f) SST advection. Negative contours are dashed and the zero contour is thick. Shading indicates statistically significant positive (light) and negative (dark) differences at the 95% confidence level based on Welch’s t test. Thick arrow also indicates the 95% significance.

Fig. 7.

JJA composite difference fields between the means of positive and negative years absolutely greater than the 0.67 std dev for the SE NP (thick solid rectangle) regional mean time series of Sc amount interannual anomalies: (a) Sc amount, (b) , (c) SLP (contours) and surface wind velocity (arrows), (d) SST, (e) upward surface latent heat flux, and (f) SST advection. Negative contours are dashed and the zero contour is thick. Shading indicates statistically significant positive (light) and negative (dark) differences at the 95% confidence level based on Welch’s t test. Thick arrow also indicates the 95% significance.

Simultaneously, high SLP anomalies occur to the north of the SE NP (Fig. 7c), where a climatological maximum in SLP (i.e., the North Pacific high) is observed. This is consistent with the work of Miyasaka and Nakamura (2005), who used numerical experiments to demonstrate that the land–sea thermal contrast associated with radiative cooling over the Sc deck is the most important forcing for the climatological formation of the Northern Hemisphere summertime subtropical highs. However, the high SLP anomalies do not work as a positive feedback contributor to the Sc amount at this time scale. The increase in the accompanying trade wind speed (Fig. 7c) does not lead to a decrease in SST (Fig. 7d) because, owing to the cold SST anomalies themselves, the anomalies of the upward surface latent heat flux are negative (Fig. 7e). Moreover, as shown in Table 2, strong correlations are not observed between the anomalies of the Sc amount and the parameters related to the subsidence closely tied to the North Pacific high.

SST advection anomalies are generally negative but insignificant in the SE NP, except for in the southern region (Fig. 7f). The cold SST anomalies (Fig. 7d) reduce (and enhance) the SST gradients to the north (and south) of the peak. This is consistent with the positive anomalies of upward surface latent (Fig. 7e) and sensible (not shown) heat fluxes in the south of the SE NP. Previous studies also suggested that these fluxes are enhanced on the downwind side of the SST anomalies (Klein et al. 1995; Mauger and Norris 2010). Therefore, no correlations can be found with these fluxes in Table 2; that is, in determining the regional mean, the positive anomalies in the south cancel out the negative anomalies in the north.

Figure 8 shows similar composite difference fields to Fig. 7 for the NW NP regional mean time series of the FOG amount interannual anomalies. The FOG amount (Fig. 8a) and (Fig. 8b) anomalies are obviously positive and large in the northwestern region of the North Pacific, with maxima in the reference area. The anomalies of the Sc amount show a similar distribution to those of the FOG amount, with negative values (not shown). As shown in Fig. 8c, anticyclonic anomalies are widely observed in the north-central region of the North Pacific. This leads to the southerly wind anomalies around 45°N in the south of the reference area, where the climatological oceanic frontal zone exists. The positive anomalies of SST advection in this region (Fig. 8f) clearly show enhanced warm advection, as also indicated by the negative anomalies of upward surface latent (Fig. 8e) and sensible (not shown) heat fluxes. The near-surface cooling of the advected warm air due to heat transfer to the cold ocean results in the intensification of the surface-based inversion (Fig. 8b) and the increase in the FOG amount (Fig. 8a). This has been noted in many previous studies (e.g., Norris and Klein 2000; Norris and Iacobellis 2005; Tanimoto et al. 2009; Tokinaga et al. 2009; Tokinaga and Xie 2009; Kawai et al. 2015). Moreover, significant zonal easterly wind anomalies around 30°N (Fig. 8c) cause the positive anomalies of the upward surface latent heat flux in this region (Fig. 8e). It is suggested that the anticyclonic transport of more humid air contributes to the increase in the FOG amount in the NW NP.

Fig. 8.

As in Fig. 7, but for the NW NP (thick dashed rectangle) regional mean time series of FOG amount interannual anomalies: (a) FOG amount, (b) , (c) SLP (contours) and surface wind velocity (arrows), (d) SST, (e) upward surface latent heat flux, and (f) SST advection.

Fig. 8.

As in Fig. 7, but for the NW NP (thick dashed rectangle) regional mean time series of FOG amount interannual anomalies: (a) FOG amount, (b) , (c) SLP (contours) and surface wind velocity (arrows), (d) SST, (e) upward surface latent heat flux, and (f) SST advection.

Positive SST anomalies around 40°N (Fig. 8d) are also assumed to be caused by the easterly wind anomalies around 30°N (Fig. 8c) via Ekman transport. The positive SST anomalies lead to strengthening of the SST gradient around 45°N, which tends to intensify the warm advection over the cold ocean, although the statistical significance is small (not shown). The spatial scale of the SST front formed by western boundary currents such as the Kuroshio is quite small (approximately 100–200 km; Xie 2004). Based on the 1° × 1° SST data, the insignificant SST gradient anomalies may be due to the lower representation of the SST front.

Consequently, the two major variations of the LSC-type amount contrast markedly. The local SST anomalies are essential to the variation in the Sc amount over the SE NP; a positive feedback is suggested. The SLP anomalies indicating the modulation of the North Pacific high are also observed, but they have no feedback relationship with the variation in the Sc amount. On the other hand, the basinwide SLP anomalies are essential to the variation in the FOG amount over the NW NP. The SST anomalies are also observed but are dependent upon the SLP anomalies.

5. Discussion

What causes the characteristic interannual variations in SST and SLP, which are closely linked with those in the summertime SE NP Sc amount and NW NP FOG amount? In this section, we discuss the possibility of a relationship with ENSO, which is the dominant variability over the Pacific Ocean on an interannual time scale. The relationship of the LSC amount over the North Pacific with ENSO has been less intensively investigated because strong correlations cannot be found between them in this region (Weare 2000; Park and Leovy 2004; Zhu et al. 2007; Eastman et al. 2011).

Table 3 lists the correlation coefficients of the time series of the Sc and FOG amounts with those of the seasonal mean Niño-3 SST indices from the Japan Meteorological Agency (JMA). The index is defined as the 5-month running mean SST deviation (which is the difference between the monthly mean SST and the climatological reference based on a sliding 30-yr period) for the Niño-3 region (5°N–5°S, 90°–150°W). There is no correlation between the SE NP Sc amount and the Niño-3 SST index for the same JJA season. This is consistent with the correlation maps presented in previous studies (Park and Leovy 2004; Eastman et al. 2011). However, the JJA SE NP Sc amount is correlated with the Niño-3 SST index for the preceding DJF season, with a correlation coefficient of 0.57. This suggests that the wintertime El Niño is associated with the increase in the Sc amount over the SE NP in the following summer. On the other hand, for the JJA FOG amount over the NW NP, significant correlations cannot be observed with the Niño-3 SST index for both the same JJA season and the preceding DJF season, although the correlation tends to be negative for the same JJA season.

Table 3.

Correlation coefficients of JJA interannual anomalies of Sc amount averaged over the SE NP or FOG amount averaged over the NW NP with the JJA or preceding DJF mean Niño-3 SST indices. Boldface indicates statistical significance at the 99% confidence level.

Correlation coefficients of JJA interannual anomalies of Sc amount averaged over the SE NP or FOG amount averaged over the NW NP with the JJA or preceding DJF mean Niño-3 SST indices. Boldface indicates statistical significance at the 99% confidence level.
Correlation coefficients of JJA interannual anomalies of Sc amount averaged over the SE NP or FOG amount averaged over the NW NP with the JJA or preceding DJF mean Niño-3 SST indices. Boldface indicates statistical significance at the 99% confidence level.

Figure 9 shows the SST interannual anomalies regressed onto the normalized JJA SE NP Sc amount over the tropics and the Northern Hemisphere. In the preceding DJF season, El Niño obviously occurs in the eastern equatorial Pacific (Fig. 9a). Cold SST anomalies simultaneously occur in the surrounding regions, which are well known as the typical SST distribution associated with ENSO (e.g., Alexander et al. 2002). Although the warm SST pattern of El Niño gradually decays from DJF to JJA (Figs. 9a–c), the cold SST anomalies are still observed and enhanced around the reference area through JJA (Fig. 9c). This implies that El Niño–related cold SST anomalies already existing in DJF are maintained and intensified by interaction with the Sc amount over the SE NP, as described in the previous section.

Fig. 9.

Tropical and Northern Hemisphere (90°N–30°S) fields of SST regressed upon the SE NP (thick solid rectangle) mean interannual anomalies of JJA Sc amount normalized by the std dev for the (a) preceding DJF, (b) preceding MAM, and (c) JJA seasons. Negative contours are dashed. Shading indicates positive (light) and negative (dark) values significant at the 95% confidence level.

Fig. 9.

Tropical and Northern Hemisphere (90°N–30°S) fields of SST regressed upon the SE NP (thick solid rectangle) mean interannual anomalies of JJA Sc amount normalized by the std dev for the (a) preceding DJF, (b) preceding MAM, and (c) JJA seasons. Negative contours are dashed. Shading indicates positive (light) and negative (dark) values significant at the 95% confidence level.

The JJA FOG amount over the NW NP does not have a clear relationship with ENSO. This begs the question: is there any possibility of a link with other climate modes? Figure 10 shows similar anomaly maps to Fig. 9c but for SLP, 200-hPa geopotential height z200, and 200-hPa meridional wind υ200 regressed onto the normalized JJA NW NP FOG amount. The high SLP anomalies in and to the east of the reference area (Fig. 10a), which are also shown in Fig. 8c, extend to the upper troposphere (Fig. 10b); the z200 and υ200 anomalies show a zonally propagating wave train pattern to the west of the reference area (Figs. 10b,c). This suggests that the interannual variation in the FOG amount over the NW NP is linked to the wavelike teleconnection pattern known as the Silk Road pattern (Enomoto et al. 2003; Enomoto 2004; Kosaka et al. 2009) or the circumglobal teleconnection pattern (Ding and Wang 2005). The stationary Rossby waves propagating along the Asian jet can amplify and break down near the northwestern Pacific, where the zonal wind speed is relatively weak. This can lead to high SLP anomalies in the north-central North Pacific. Ding and Wang (2005) also showed that the wave train pattern has weak association with La Niña through the Indian summer monsoon activity, which is consistent with the less significant but negative correlation between the NW NP FOG amount and the Niño-3 SST index in JJA (Table 3).

Fig. 10.

Tropical and Northern Hemisphere (90°N–30°S) fields of (a) SLP, (b) z200, and (c) υ200 regressed upon the NW NP (thick dashed rectangle) mean interannual anomalies of JJA FOG amount normalized by the std dev for the JJA season. Negative contours are dashed. Shading indicates positive (light) and negative (dark) values significant at the 95% confidence level.

Fig. 10.

Tropical and Northern Hemisphere (90°N–30°S) fields of (a) SLP, (b) z200, and (c) υ200 regressed upon the NW NP (thick dashed rectangle) mean interannual anomalies of JJA FOG amount normalized by the std dev for the JJA season. Negative contours are dashed. Shading indicates positive (light) and negative (dark) values significant at the 95% confidence level.

6. Summary

Using long-term ship-based cloud observations, we have investigated the summertime interannual variability in the LSC amount and its relation to the inversion strength measured by EIS over the North Pacific. The interannual correlation between the LSC amount and EIS is positive but relatively weak, compared with the well-known strong linear relationship for the seasonal variabilities. This reflects the regional contrast: the correlations are stronger in the SE NP and weaker in the NW NP. The interannual variations in the LSC amount are large in the SE NP, whereas those are small in the NW NP. However, the interannual variations in EIS are large in both regions.

In terms of the LSC types (i.e., Sc, St, and FOG), we found two major variations: one is in the Sc amount over the SE NP and the other in the FOG amount over the NW NP. From analyses using layered EISs proposed by KS14, the former is correlated with the inferred inversion strength between the 700- and 925-hPa levels (corresponding to the capping inversion) and the latter is correlated with that between the 925-hPa level and the surface (corresponding to the surface-based inversion). On the other hand, variations in the St amount are small over the whole North Pacific. It should be emphasized that it is difficult to elucidate the physical relationship between the total LSC amount and the bulk EIS at this time scale, especially over the NW NP. Their correlation is apparently weak due to the compensating variations of the Sc and FOG amounts. Our study clearly indicates the importance of LSC type. The usefulness of the layered EISs is also demonstrated in terms of distinguishing between the capping inversion and the surface-based inversion, and in identifying the corresponding LSC type.

The two major variations occur independently of each other and are caused by distinctly different mechanisms. The interannual variation in the Sc amount anomalies over the SE NP is closely linked to that in the local SST anomalies; the increase in the Sc amount corresponds to the colder SST. This suggests that there is a positive feedback between these factors. Moreover, we identified the possibility for cold SST anomalies to be triggered by an El Niño in the preceding winter. On the other hand, the interannual variation in the FOG amount anomalies in the NW NP is caused by the SLP anomalies at the north-central region of the North Pacific. The surface anticyclonic anomalies lead to southerly warm advection of moist air across the climatologically strong SST gradients, resulting in more FOG. The SLP anomalies are associated with a wavelike extratropical teleconnection pattern along the Asian jet in the same summer season.

It is beyond the scope of this paper to examine the more detailed mechanisms involved in the effect of the preceding ENSO on the SST anomalies in the SE NP and the influence of the extratropical atmospheric wave train on the SLP anomalies in the north-central North Pacific. They would individually deserve further investigation in future studies. While this study focused on interannual variability in the LSC-type amounts over the North Pacific, the North Atlantic is another potential target area to be investigated because the LSC amounts are climatologically large and all three types of LSC often occur there. In addition, since ship-based observations are abundant over these oceans, statistically careful treatments, such as those used in the previous studies mentioned in the introduction (e.g., Norris 2005b), might reveal longer time-scale variabilities. Decadal variabilities and long-term trends of the LSC-type amounts and the related meteorological parameters would be challenging topics for future investigation.

Acknowledgments

The authors wish to thank Masami Nonaka, Hisashi Nakamura, and Youichi Tanimoto for helpful discussions on this work. Thanks are also due to Hiroki Tokinaga, Yu Kosaka, Tamaki Yasuda, Shuhei Maeda, Masato Sugi, Tomoaki Ose, and Hideaki Kawai for their insightful comments. The authors are grateful to Steve Klein for editing this article and to three anonymous reviewers for their constructive suggestions for improvement. The EECRA data are obtained from the Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, U.S. Department of Energy (http://cdiac.ornl.gov/epubs/ndp/ndp026c/ndp026c.html). The JRA-55 (http://jra.kishou.go.jp/JRA-55/index_en.html) and COBE SST data (http://ds.data.jma.go.jp/tcc/tcc/products/elnino/cobesst/cobe-sst.html) are both provided from the projects carried out by JMA. The Niño-3 SST index data are also obtained from the website of JMA El Niño monitoring indices (http://ds.data.jma.go.jp/tcc/tcc/products/elnino/index/). This study was supported in part by the Japanese Ministry of Education, Culture, Sports, Science and Technology (MEXT) through Grant-in-Aid for Scientific Research on Innovative Areas #2205 JP22106009 and by the Japan Society for the Promotion of Science (JSPS) through Grant-in-Aid for Young Scientists (A) JP26701004. Data analysis and visualization were performed using the GFD-Dennou Ruby libraries (http://ruby.gfd-dennou.org/).

REFERENCES

REFERENCES
Alexander
,
M. A.
,
I.
Bladé
,
M.
Newman
,
J. R.
Lanzante
,
N.-C.
Lau
, and
J. D.
Scott
,
2002
:
The atmospheric bridge: The influence of ENSO teleconnections on air–sea interaction over the global oceans
.
J. Climate
,
15
,
2205
2231
, doi:.
Bajuk
,
L. J.
, and
C. B.
Leovy
,
1998
:
Are there real interdecadal variations in marine low clouds?
J. Climate
,
11
,
2910
2921
, doi:.
Bellomo
,
K.
,
A. C.
Clement
,
J. R.
Norris
, and
B. J.
Soden
,
2014
:
Observational and model estimates of cloud amount feedback over the Indian and Pacific Oceans
.
J. Climate
,
27
,
925
940
, doi:.
Bellomo
,
K.
,
A. C.
Clement
,
T.
Mauritsen
,
G.
Rädel
, and
B.
Stevens
,
2015
:
The influence of cloud feedbacks on equatorial Atlantic variability
.
J. Climate
,
28
,
2725
2744
, doi:.
Bolton
,
D.
,
1980
:
The computation of equivalent potential temperature
.
Mon. Wea. Rev.
,
108
,
1046
1053
, doi:.
Bretherton
,
C. S.
,
P. N.
Blossey
, and
C. R.
Jones
,
2013
:
Mechanisms of marine low cloud sensitivity to idealized climate perturbations: A single-LES exploration extending the CGILS cases
.
J. Adv. Model. Earth Syst.
,
5
,
316
337
, doi:.
Clement
,
A. C.
,
R.
Burgman
, and
J. R.
Norris
,
2009
:
Observational and model evidence for positive low-level cloud feedback
.
Science
,
325
,
460
464
, doi:.
Deser
,
C.
, and
A. S.
Phillips
,
2006
:
Simulation of the 1976/77 climate transition over the North Pacific: Sensitivity to tropical forcing
.
J. Climate
,
19
,
6170
6180
, doi:.
Deser
,
C.
,
A. S.
Phillips
, and
M. A.
Alexander
,
2010
:
Twentieth century tropical sea surface temperature trends revisited
.
Geophys. Res. Lett.
,
37
,
L10701
, doi:.
Ding
,
Q.
, and
B.
Wang
,
2005
:
Circumglobal teleconnection in the Northern Hemisphere summer
.
J. Climate
,
18
,
3483
3505
, doi:.
Eastman
,
R.
,
S. G.
Warren
, and
C. J.
Hahn
,
2011
:
Variations in cloud cover and cloud types over the ocean from surface observations, 1954–2008
.
J. Climate
,
24
,
5914
5934
, doi:.
Enomoto
,
T.
,
2004
:
Interannual variability of the Bonin high associated with the propagation of Rossby waves along the Asian jet
.
J. Meteor. Soc. Japan
,
82
,
1019
1034
, doi:.
Enomoto
,
T.
,
B. J.
Hoskins
, and
Y.
Matsuda
,
2003
:
The formation mechanism of the Bonin high in August
.
Quart. J. Roy. Meteor. Soc.
,
129
,
157
178
, doi:.
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). CDIAC Numerical Data Package NDP-026C, Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, 79 pp., doi:.
Hahn
,
C. J.
,
S. G.
Warren
, and
J.
London
,
1995
:
The effect of moonlight on observation of cloud cover at night, and application to cloud climatology
.
J. Climate
,
8
,
1429
1446
, doi:.
Ishii
,
M.
,
A.
Shouji
,
S.
Sugimoto
, and
T.
Matsumoto
,
2005
:
Objective analyses of sea-surface temperature and marine meteorological variables for the 20th century using ICOADS and the Kobe Collection
.
Int. J. Climatol.
,
25
,
865
879
, doi:.
Kamae
,
Y.
,
T.
Ogura
,
H.
Shiogama
, and
M.
Watanabe
,
2016
:
Recent progress toward reducing the uncertainty in tropical low cloud feedback and climate sensitivity: A review
.
Geosci. Lett.
,
3
,
17
, doi:.
Kawai
,
H.
,
S.
Yabu
,
Y.
Hagihara
,
T.
Koshiro
, and
H.
Okamoto
,
2015
:
Characteristics of the cloud top heights of marine boundary layer clouds and the frequency of marine fog over mid-latitudes
.
J. Meteor. Soc. Japan
,
93
,
613
628
, doi:.
Klein
,
S. A.
,
1997
:
Synoptic variability of low-cloud properties and meteorological parameters in the subtropical trade wind boundary layer
.
J. Climate
,
10
,
2018
2039
, doi:.
Klein
,
S. A.
, and
D. L.
Hartmann
,
1993
:
The seasonal cycle of low stratiform clouds
.
J. Climate
,
6
,
1587
1606
, doi:.
Klein
,
S. A.
,
D. L.
Hartmann
, and
J. R.
Norris
,
1995
:
On the relationships among low-cloud structure, sea surface temperature, and atmospheric circulation in the summertime northeast Pacific
.
J. Climate
,
8
,
1140
1155
, doi:.
Kobayashi
,
S.
, and Coauthors
,
2015
:
The JRA-55 reanalysis: General specifications and basic characteristics
.
J. Meteor. Soc. Japan
,
93
,
5
48
, doi:.
Kosaka
,
Y.
,
H.
Nakamura
,
M.
Watanabe
, and
M.
Kimoto
,
2009
:
Analysis on the dynamics of a wave-like teleconnection pattern along the summertime Asian jet based on a reanalysis dataset and climate model simulations
.
J. Meteor. Soc. Japan
,
87
,
561
580
, doi:.
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:.
Mansbach
,
D. K.
, and
J. R.
Norris
,
2007
:
Low-level cloud variability over the equatorial cold tongue in observations and models
.
J. Climate
,
20
,
1555
1570
, doi:.
Mauger
,
G. S.
, and
J. R.
Norris
,
2010
:
Assessing the impact of meteorological history on subtropical cloud fraction
.
J. Climate
,
23
,
2926
2940
, doi:.
Miyasaka
,
T.
, and
H.
Nakamura
,
2005
:
Structure and formation mechanisms of the Northern Hemisphere summertime subtropical highs
.
J. Climate
,
18
,
5046
5065
, doi:.
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:.
Norris
,
J. R.
,
1998a
:
Low cloud type over the ocean from surface observations. Part I: Relationship to surface meteorology and the vertical distribution of temperature and moisture
.
J. Climate
,
11
,
369
382
, doi:.
Norris
,
J. R.
,
1998b
:
Low cloud type over the ocean from surface observations. Part II: Geographical and seasonal variations
.
J. Climate
,
11
,
383
403
, doi:.
Norris
,
J. R.
,
1999
:
On trends and possible artifacts in global ocean cloud cover between 1952 and 1995
.
J. Climate
,
12
,
1864
1870
, doi:.
Norris
,
J. R.
,
2000a
:
Interannual and interdecadal variability in the storm track, cloudiness, and sea surface temperature over the summertime North Pacific
.
J. Climate
,
13
,
422
430
, doi:.
Norris
,
J. R.
,
2000b
:
What can cloud observations tell us about climate variability?
Space Sci. Rev.
,
94
,
375
380
, doi:.
Norris
,
J. R.
,
2005a
:
Multidecadal changes in near-global cloud cover and estimated cloud cover radiative forcing
.
J. Geophys. Res.
,
110
,
D08206
, doi:.
Norris
,
J. R.
,
2005b
:
Trends in upper-level cloud cover and surface divergence over the tropical Indo-Pacific Ocean between 1952 and 1997
.
J. Geophys. Res.
,
110
,
D21110
, doi:.
Norris
,
J. R.
, and
C. B.
Leovy
,
1994
:
Interannual variability in stratiform cloudiness and sea surface temperature
.
J. Climate
,
7
,
1915
1925
, doi:.
Norris
,
J. R.
, and
S. A.
Klein
,
2000
:
Low cloud type over the ocean from surface observations. Part III: Relationship to vertical motion and the regional surface synoptic environment
.
J. Climate
,
13
,
245
256
, doi:.
Norris
,
J. R.
, and
S. F.
Iacobellis
,
2005
:
North Pacific cloud feedbacks inferred from synoptic-scale dynamic and thermodynamic relationships
.
J. Climate
,
18
,
4862
4878
, doi:.
Norris
,
J. R.
,
Y.
Zhang
, and
J. M.
Wallace
,
1998
:
Role of low clouds in summertime atmosphere–ocean interactions over the North Pacific
.
J. Climate
,
11
,
2482
2490
, doi:.
Park
,
S.
, and
C. B.
Leovy
,
2004
:
Marine low-cloud anomalies associated with ENSO
.
J. Climate
,
17
,
3448
3469
, doi:.
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:.
Tanimoto
,
Y.
,
S.-P.
Xie
,
K.
Kai
,
H.
Okajima
,
H.
Tokinaga
,
T.
Murayama
,
M.
Nonaka
, and
H.
Nakamura
,
2009
:
Observations of marine atmospheric boundary layer transitions across the summer Kuroshio Extension
.
J. Climate
,
22
,
1360
1374
, doi:.
Tokinaga
,
H.
, and
S.-P.
Xie
,
2009
:
Ocean tidal cooling effect on summer sea fog over the Okhotsk Sea
.
J. Geophys. Res.
,
114
,
D14102
, doi:.
Tokinaga
,
H.
,
Y.
Tanimoto
,
S.-P.
Xie
,
T.
Sampe
,
H.
Tomita
, and
H.
Ichikawa
,
2009
:
Ocean frontal effects on the vertical development of clouds over the western North Pacific: In situ and satellite observations
.
J. Climate
,
22
,
4241
4260
, doi:.
Tokinaga
,
H.
,
S.-P.
Xie
,
A.
Timmermann
,
S.
McGregor
,
T.
Ogata
,
H.
Kubota
, and
Y. M.
Okumura
,
2012
:
Regional patterns of tropical Indo-Pacific climate change: Evidence of the Walker circulation weakening
.
J. Climate
,
25
,
1689
1710
, doi:.
Warren
,
S. G.
,
C. J.
Hahn
,
J.
London
,
R. M.
Chervin
, and
R. L.
Jenne
,
1988
: Global distribution of total cloud cover and cloud type amounts over the ocean. NCAR Tech. Note NCAR/TN-317+STR, 42 pp., doi:.
Weare
,
B. C.
,
2000
:
Near-global observations of low clouds
.
J. Climate
,
13
,
1255
1268
, doi:.
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:.
Xie
,
S.-P.
,
2004
:
Satellite observations of cool ocean–atmosphere interaction
.
Bull. Amer. Meteor. Soc.
,
85
,
195
208
, doi:.
Xu
,
H.
,
S.-P.
Xie
, and
Y.
Wang
,
2005
:
Subseasonal variability of the southeast Pacific stratus cloud deck
.
J. Climate
,
18
,
131
142
, doi:.
Zhu
,
P.
,
J. J.
Hack
,
J. T.
Kiehl
, and
C. S.
Bretherton
,
2007
:
Climate sensitivity of tropical and subtropical marine low cloud amount to ENSO and global warming due to doubled CO2
.
J. Geophys. Res.
,
112
,
D17108
, doi:.

Footnotes

© 2017 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).