1. Introduction
Clouds present the largest modification of energy transfer through the climate system (e.g., Trenberth et al. 2009). Over the pan-Arctic basin, the presence of clouds generally leads to a surplus of energy reaching the surface, where the cloud greenhouse warming effect dominates the shortwave cooling effect (Walsh and Chapman 1998; Shupe and Intrieri 2004; Sedlar et al. 2011). Increased surface reflectivity due to the presence of snow, sea ice, and large solar zenith angles combined with generally lower-atmospheric opacity are the primary reasons for the cloud warming effect outlasting the cloud cooling effect in the Arctic. Outside of the high latitudes, the cooling effect of clouds is the most dominant (e.g., Klein and Hartmann 1993), indicating an important change in sign in the local climate forcing resulting from cloud cover.
The surface warming effect of clouds in the Arctic depends critically on their microphysical composition. The presence of liquid droplets dramatically increases the opacity of both longwave and shortwave radiation (e.g., Stephens 1978a), relative to cloud-free or ice-only clouds. Over the Arctic, clouds consisting of both liquid and ice hydrometeors (mixed-phase clouds) have been observed during the full annual cycle, often observed below 3 km (Shupe 2011); liquid-only clouds are also present, but generally limited in occurrence to sunlit months and at very low levels (<1 km) (Shupe 2011). The radiative warming effect of these clouds over the Arctic has been directly related to the onset of melt and freeze processes of snow and sea ice (Intrieri et al. 2002; Sedlar et al. 2011; Persson 2012). Careful consideration and understanding of Arctic clouds and the processes supporting their frequent occurrence are thus needed to understand cloud-forced changes at the surface, within the atmospheric boundary layer, and thermodynamic changes in the troposphere.
Stratiform cloud formation and persistence are generally the result of an interaction between horizontal advection of thermodynamics and radiative divergence. The advection of heat and moisture into the relatively cool and dry Arctic basin leads to air mass modification (cooling) and formation of condensate on suspended aerosols (e.g., Herman and Goody 1976); the condensate interacts with radiation, and if sufficient cooling continues, clouds generally thrive via buoyancy circulations driven by the cloud layer (e.g., Curry 1986). Liquid droplets tend to be sustained by overturning circulations that produce up- and downdrafts that effectively alter the local supersaturation level (e.g., Paluch and Lenschow 1991). In the presence of ice nuclei, ice crystals also tend to form in locally enhanced supersaturations produced by ascending motions (Shupe et al. 2008b). Coincident ice and liquid hydrometeors are colloidally unstable because of the supersaturation differences between the two phases—the Wegener–Bergeron–Findeisen process; a lack in active mixing and/or source of moisture and aerosols leads to droplet evaporation in order to retain ice supersaturations. However, when in-cloud mixing processes are sustained, clouds generally support their own persistence (Morrison et al. 2012).
Cloud observations from the Arctic have recently been exploited to understand the role of cloud-generated mixing processes in maintaining high cloud fractions (Shupe et al. 2008b, 2013; Solomon et al. 2011; Sedlar and Shupe 2014). These studies have highlighted the importance of liquid droplets in maintaining longwave divergence and sustaining cloud-produced buoyant mixing. However, these studies are based on limited datasets often resembling case studies specific to a particular thermodynamic setting or limited range of settings. A particular uncertainty is how different ranges of cloud liquid water path (LWP) affect the generation and physical characteristics of in-cloud buoyant mixing. Bennartz et al. (2013) report that low liquid-bearing clouds (LWP ranging from 10 to 60 g m−2) are evident in observations across the Arctic basin with occurrence frequencies as large as 30%–40%. In fact, they identify a sufficiently optically thin liquid-bearing cloud layer over Summit, Greenland, during July 2012 as responsible for contributing to the first observed surface melting event at Summit since 1889. Thus, more observational analysis on the similarities and differences of static mixing processes for optically thin and thick liquid-bearing clouds are necessary.
To address the implications of limited LWP on cloud-generated mixing, long-term observations of cloud and thermodynamic properties are examined in this study. Nearly a decade of observations from Barrow, Alaska, are compared with one full annual cycle of observations from a sea ice drifting station in the Beaufort Sea north of Alaska. Methods are applied to analyze the thermodynamic structure of in-cloud mixing and how it relates to the amount, if any, of cloud liquid present. Furthermore, the results are analyzed with specific focus on how in-cloud mixing is impacted by frequently observed in-cloud stable layers in the upper portion of these clouds (Sedlar and Tjernström 2009; Sedlar et al. 2012; Shupe et al. 2013; Sedlar and Shupe 2014).
2. Observations and analysis methods
a. Thermodynamic profiles
Long-term surface and atmospheric monitoring has been the overarching goal of the Atmospheric Radiation Measurement Program (ARM)–North Slope of Alaska (NSA) super site at Barrow since 1998. Radiosoundings provide accurate profiles of the thermodynamic state of the full troposphere, often with temperature uncertainties documented in the sensor specification sheet at 0.5°C. Equivalent potential temperature θe, a conserved variable in moist adiabatic motions, and specific humidity q profiles are analyzed to quantify the lower-tropospheric thermodynamic structure. The focus of this study is on low-level cloud and thermodynamic properties at Barrow from 2002 to 2010, where the low level is defined as below 1.5 km AGL. During this time, a total of 4527 radiosounding profiles are analyzed in this study. Prior to 2005, only one radiosounding was launched daily; two profiles were made daily thereafter.
To complement observations from the Barrow super site, thermodynamic profiles from the Surface Heat Budget of the Arctic (SHEBA; Uttal et al. 2002) are also examined. The SHEBA experiment was essentially a mobile super site observing detailed surface, boundary layer, and atmospheric conditions over one full annual cycle between 1997 and 1998 within the Arctic pack ice. SHEBA observations are analyzed for comparison with Barrow observations under the assumption that thermodynamic and cloud properties may differ between a remote sea ice observatory and Barrow that may have a more continental footprint. Radiosoundings were released twice daily, yielding a total of 801 radiosounding profiles for analysis.
b. Cloud observations
Similar surface-based instrumentation to observe cloud properties was deployed at Barrow and SHEBA (see Table 1). The cloud boundaries (base and top heights) and number of layers for both locations were derived from a combination of Millimeter Cloud Radar (MMCR; Moran et al. 1998) reflectivity and either ceilometer and/or lidar backscatter ratios following the Active Remote Sensing of Clouds (ARSCL) retrieval algorithm (Clothiaux et al. 2000). The ARSCL retrieved datasets provide a best-estimate time series of cloud-base and cloud-top heights following various processing chains of the input sensor data (Clothiaux et al. 2000). The Barrow time series were downloaded from the ARM Data Archive (http://www.archive.arm.gov/). The remote sensors retrieve cloud properties on relatively fast frequencies ranging from 10 to 30 s, and the final cloud boundary datasets are interpolated to 10-s temporal resolution.
Surface-based remote sensing instrumentation descriptions and respective derived measurements of cloud observations.
To accommodate the sparse temporal resolution of the radiosounding profiles, median cloud properties (base and top heights, number of layers, and cloud LWP) are derived within a time window of 10 min following each radiosounding launch. A microwave radiometer (MWR) estimates LWP using radiometric retrievals just above the transmission window near 31.6 GHz (Guiraud et al. 1979). LWP is a column-integrated quantity; during profiles of multiple cloud layers, it is impossible to know with certainty the presence and distribution of liquid hydrometeors. Therefore, single-layer-only clouds are the focus of this study. However, retrievals of LWP from the MWR are not without uncertainty and potential bias (e.g., Marchand et al. 2003), which typically are reported to be 25 g m−2 (Westwater et al. 2001). While mixed-phase clouds are often the most frequently observed cloud type over the annual cycle in the Arctic (e.g., Shupe 2011), LWP uncertainty from the MWR retrievals poses a problem for identifying cloud layers that may in reality contain liquid when the LWP value is within the uncertainty range. Therefore, in this study, single-layer clouds are separated by LWP classes: 1) having LWP ranging from 0 to 50 g m−2 and 2) having LWP ranging from 75 to 300 g m−2. While an inherent uncertainty remains whether or not clouds with LWP < 25 g m−2 contain droplets, it is still possible to confidently separate optically thin LWP clouds (0–50 g m−2) from optically thicker clouds (75–300 g m−2). Therefore, it must be noted that clouds in the optically thin LWP range may be entirely free of liquid (ice only); however, this study is concerned with the impacts of limited LWP on cloud thermodynamic structure and radiative mixing. Thus, differences between optically thick LWP clouds with optically thin, or liquid-free, clouds can and will be quantified here.
It is important to note that the MMCR is very sensitive to hydrometeor size. Larger ice crystals falling through multiple, embedded liquid cloud layers may occur (e.g., Rambukkange et al. 2011; Verlinde et al. 2013), in which case the ARSCL cloud boundary dataset may only report a single cloud layer. From these data streams, it is impossible to characterize if or how frequently multiple liquid cloud layers embedded within precipitation may have impacted these observations.
c. Cloudy fraction around radiosounding profiles
Figure 1 highlights that cloud fractions around radiosounding profiles have generally large occurrence frequencies greater than 80%. High cloud fractions are canonical in the Arctic (e.g., Curry et al. 1996; Wang and Key 2005). Fractions above 80% agree well with those reported by Shupe et al. (2011), suggesting that cloud analysis around radiosounding times is representative of the general conditions at Barrow and SHEBA.
RFD of cloud fraction observed following radiosounding launches at NSA–Barrow (magenta) and SHEBA (blue). “Cloudy any sensor” refers to the either the MMCR or ceilometer–lidar combination sensing cloud overhead. Single layer is the relative frequency of single-layer clouds, further segregated into single-layer clouds with a top below 1.5 km AGL; single-layer clouds with tops below 1.5 km are also segregated into two LWP ranges: optically thin (0–50 g m−2) and optically thick (75–300 g m−2).
Citation: Journal of Applied Meteorology and Climatology 53, 12; 10.1175/JAMC-D-14-0065.1
The majority of these observed clouds are single-layer clouds, while approximately 27%–30% are single layers with a cloud top below 1.5 km AGL (Fig. 1). Of these, optically thin clouds (LWP 0–50 g m−2) are more frequent relative to optically thick clouds (LWP 75–300 g m−2) at both Barrow (14% to 10%, respectively) and SHEBA (12% to 4%, respectively). These results suggest we are examining a substantially large subsample (16%–24%) of all the cloudy conditions over both Barrow and SHEBA. These single-layer, low clouds contain small to ample amounts of liquid condensate, in agreement with long-term observations from Barrow (Shupe 2011; Shupe et al. 2011; Sedlar et al. 2012). Hence, the radiative impact of these clouds can potentially result in significant alterations to the surface energy budget (e.g., Sedlar et al. 2011). Optically thin clouds at SHEBA were over 3 times as frequent as thick clouds (Fig. 1); however, the total number of single-layer clouds around radiosounding times is rather low (94 in total between the two LWP ranges). Therefore, the optically thick LWP cloud subclass from SHEBA is excluded in subsequent analyses based on the small sample size (20 observations).
The distributions of observed LWP as a function of cloud-base temperature are shown in Fig. 2. These histograms for the longer time series at Barrow reveal several insights into the cloud regimes observed. First, the distribution of LWP for the lower LWP subclass (Fig. 2a) indicates local maxima around 10–15 g m−2, well within the LWP retrieval uncertainty; it is therefore impossible to determine whether liquid is in reality present for these clouds. This LWP distribution peak also corresponds to cloud-base temperatures colder than −15°C, which may 1) be related to the decrease in vapor pressure at these cooler temperatures following the Clausius–Clapeyron relation or 2) be an indication of more ice-dominated, or ice-only, clouds. A secondary peak in the histogram is generally found for LWPs > 25 g m−2 and warmer temperatures (Fig. 2a). Despite the scatter at SHEBA, the lower LWP clouds here also have a tendency to group with either base temperatures around −20° and −5°C (Fig. 2c). However, the larger LWP subclass at Barrow generally shows a wide range of observed LWPs, generally for cloud-base temperatures between −12° and 0°C (Fig. 2b). Analysis of the annual cycle of these LWP subclasses at Barrow reveals that the lower LWP subclass has a relatively homogenous frequency occurrence during all months, while the larger subclass occurs primarily from May to November (not shown). Thus, the large annual cycle variation in the atmospheric thermodynamic setting correlates with the observed amount of cloud condensate for these single-layer, low-level clouds.
Relative frequency 2D distributions (color shading) of observed cloud LWP (g m−2) as a function of cloud-base temperature (°C) for (a) Barrow LWP from 0 to 50 g m−2, (b) Barrow LWP from 75 to 300 g m−2, and (c) SHEBA LWP from 0 to 50 g m−2; SHEBA observations are shown as a scatterplot (black stars).
Citation: Journal of Applied Meteorology and Climatology 53, 12; 10.1175/JAMC-D-14-0065.1
d. Profile normalization
3. Results
a. Cloud geometric boundaries
Relative frequency distributions (RFDs) of cloud thickness and cloud-base height are presented in Fig. 3. The majority of these clouds have both thicknesses (Fig. 3a) and base heights under 600 m (Fig. 3b). Cloud-base heights are often lower at SHEBA compared to Barrow (Fig. 3b), most frequently below 200 m. While cloud-base RFDs are similar for both LWP ranges at Barrow, the RFDs of thickness have a bimodal distribution. Optically thin and thick clouds share a thickness distribution maximum between 75 and 125 m (Fig. 3a, red and blue). The secondary thickness maxima, however, is larger (approximately 525 vs 300 m), and the distribution is more positively skewed for the larger LWP clouds. Nearly 90% of Barrow clouds with LWP < 50 g m−2 have a thickness below 425 m as compared with only 44% for LWP > 75 g m−2. These cloud thickness distributions tend to correlate with the cloud-base temperature regimes, and hence the dependence of water vapor on temperature and resulting condensate, observed in Fig. 2.
RFDs of (a) cloud thickness and (b) cloud-base heights. Data are grouped by LWP ranges for Barrow (red and blue) and SHEBA (black). RFD bin sizes are 50 m, centered on the interval.
Citation: Journal of Applied Meteorology and Climatology 53, 12; 10.1175/JAMC-D-14-0065.1
Arctic mixed-phase cloud condensate generally increases with height near adiabatically (e.g., Curry 1986; Shupe et al. 2001; McFarquhar et al. 2011). Thus, larger LWPs can simply be a caveat of thicker clouds, as the RFDs for Barrow cloud thickness suggest (Fig. 3a). However, LWP is defined as the column integral of cloud liquid water content (LWC) over the cloud thickness—thicker clouds must contain more condensate than thinner clouds in order for LWP to increase. The distribution of cloud thickness over the sea ice at SHEBA (black) reveals generally thicker clouds (300–600 m) than were seen for the same LWP range at Barrow (blue). Near-adiabatic condensate production over a thicker cloud layer thus cannot be concluded as the only factor impacting cloud LWP. These results hint toward additional mechanisms contributing to the cloud condensate, which will be described in the subsequent sections.
b. Cloud layer thermodynamic profile
Vertical distributions of thermodynamic stability exhibit statistically similar behavior among the two locations and LWP subsamples (Fig. 4a). Median stabilities are near neutral (~0) in the lower half of the cloud and increasingly stable (2–7 K) above. Interquartile ranges also increase, indicating larger Δθe variability with increasing height. The stability distributions agree very well with in-cloud analyses over the central Arctic during autumn 2008 (Shupe et al. 2013; Sedlar and Shupe 2014; Sotiropoulou et al. 2014).
In-cloud statistical box-and-whisker distributions of (a) Δθe (K) and (b) Δq (g kg−1) estimated following Eq. (2). Cloud layer is normalized by the base (zn = 0) and top (zn = 1) heights, interpolated to a common height grid of Δzn = 0.1. Box widths represent the interquartile range with median values denoted by the vertical line; whiskers span the 5th–95th percentiles. Cloud LWP ranges are specified by the color key in the bottom of both panels.
Citation: Journal of Applied Meteorology and Climatology 53, 12; 10.1175/JAMC-D-14-0065.1
Variations in specific humidity with height, following Eq. (2), above cloud base are less comparable between Barrow and SHEBA and across LWP ranges (Fig. 4b). Median Δq across clouds with LWP < 50 g m−2 are generally around zero (Fig. 4b, blue and black), with an increasing frequency of a positive Δq in the upper cloud layer at SHEBA. Across optically thicker clouds at Barrow, Δq are most frequently negative (Fig. 4b, red). Despite the increasing stability with height, q remains unchanged or tends to increase slightly in the upper half of clouds with LWP below 50 g m−2. Considering the stability changes with in-cloud height, the lower LWP subclass of clouds often is observed with increased moisture coinciding with increasing temperature. A potential source of moisture aloft for cloud lifetime (e.g., Sedlar et al. 2012) is therefore present near the cloud top for a large portion of these clouds.
Median Δθe for the two Barrow LWP subclasses are significantly different (99% level; two-sided Wilcoxon rank-sum confidence test) for zn = 0.3–0.8 (Fig. 4a, red and blue). The transition toward increasing stability with height is even more pronounced for the low-LWP subclass at SHEBA (Fig. 4a, black). These profiles suggest that static mixed layers tend to span a larger portion of the cloud for LWPs > 75 g m−2 than for LWPs < 50 g m−2. These larger portions of the cloud also physically correspond to a deeper in-cloud mixed layer considering the distribution differences in the positively skewed cloud thickness RFDs between the two LWP subclasses (Fig. 3a). Statistical profiles in Fig. 4 also indicate a uniquely common phenomenon of Arctic cloud top penetrating and persisting within a stable temperature inversion structure (Sedlar and Tjernström 2009; Sedlar et al. 2012), frequently coinciding with increased q with height (moisture inversion) (Sedlar et al. 2012). Despite differences in the in-cloud mixed layer depth and the presence of moisture inversions between the two LWP subclasses, both cases are often subject to an in-cloud stable layer. Large-eddy simulations of in-cloud stable layers have been shown to contain very little cloud condensate, but enough to support weak radiative divergence resulting in droplet condensation (Solomon et al. 2011); a condition necessary to maintain condensation in this stable layer, however, is a moisture inversion located within the stable layer (Solomon et al. 2011; Sedlar et al. 2012).
c. Cloud properties and in-cloud static stability
Estimates of the vertical static mixing depth within the cloud are made following Eq. (2). The normalized height where Δθe becomes >1.0 K from cloud base is considered the upper bound of the statically mixed layer. A value of 1 K is used as the radiosounding temperature uncertainty is listed in the sensor specifications sheet as <0.5 K.
Figure 5 illustrates a positive relationship between cloud thickness and depth of in-cloud mixed layer. The ratio between mixed layer depth and cloud thickness also reveals a dependence on cloud LWP. At Barrow, static mixed layer depths for the low LWP subclass are frequently less than 200 m, with considerable scatter in corresponding cloud thicknesses (Fig. 5b); the median relationship indicates the mixed layer depth is approximately half of the cloud physical thickness (Fig. 5b, red box-and-whisker plot). As LWP increases, so does the physical mixing depth (Fig. 5c). In a statistical sense, however, the median mixed layer depth is only modestly larger than half (0.6) the cloud thickness (Fig. 5c, red box-and-whisker plot). However, the scatterplot shows for LWP > 75 g m−2 that the in-cloud mixed layer generally spans entire cloud depth when cloud thickness < 300 m. Geometrically thinner clouds are thus less likely to be statically mixed across the full cloud depth when LWP is below 50 g m−2. Over the sea ice at SHEBA, increases in cloud thickness generally translate to an in-cloud mixed layer depth only slightly larger than half the cloud thickness (Fig. 5a).
Scatterplot relationship between in-cloud mixed layer depth (m; see text for determination of mixed layer depth) and cloud thickness (m) for (a) SHEBA LWP of 0–50 g m−2, (b) Barrow LWP of 0–50 g m−2, and (c) Barrow LWP of 75–300 g m−2. Box-and-whisker distributions showing the median, interquartile ranges, and 10th–90th percentiles overlay the data points. Note that the box-and-whisker distributions are located at the median values of both in-cloud mixed layer depth and cloud thickness. The 1:1 and 1:2 relationship lines (gray) are shown as references.
Citation: Journal of Applied Meteorology and Climatology 53, 12; 10.1175/JAMC-D-14-0065.1
The degree of static mixing may be a key factor in determining the depth of the in-cloud mixed layer. To test this, a bulk mixed layer temperature gradient is estimated from the temperature and layer depth differences between the mixed layer top and cloud base. The observed relationship between bulk temperature gradients and in-cloud mixed layer depths is shown in Figs. 6a and 6b. Clearly, the deepest mixed layers are observed when LWP is larger (e.g., Fig. 6b). Interestingly, mixed layer temperature gradients are approximately equal to the moist adiabatic value, or even slightly superadiabatic, for the larger half of mixed layer depths and for all LWP subclasses. These temperature gradients likely contribute to the associated deeper cloud mixed layers. For the shallower mixed layers, the layer temperature gradients are more frequently subadiabatic for the lower LWP subclass at both Barrow and SHEBA (Fig. 6a). In some cases, the supposed mixed layer is in fact stable, but these observations are generally for very thin mixed layers. Closer inspection shows the larger of these stable layers (>0.02°C m−1) has LWPs below 20 g m−2, well within the instrument uncertainty and potentially void of liquid altogether. Distributions of subadiabatic temperature gradients are not observed for LWPs ranging from 75 to 300 g m−2 at Barrow (Fig. 6b). A correlation between LWP, in-cloud mixing depth, and the sign and magnitude of the mixed layer temperature gradient is therefore present in these observations.
Relative frequency 2D distributions of in-cloud mixed layer depth (m) as a function of bulk in-cloud mixed layer temperature gradient (°C m−1) for (a) lower LWP subclasses at Barrow (color shading) and SHEBA (stars) and (b) the larger LWP subclass at Barrow. (c) The relationship between in-cloud stable layer temperature increases (°C) and bulk in-cloud mixed layer temperature gradients for Barrow (0–50 g m−2, blue dots; 75–300 g m−2, red circles) and SHEBA (0–50 g m−2; stars). The black vertical line in each panel is the moist adiabatic lapse rate value.
Citation: Journal of Applied Meteorology and Climatology 53, 12; 10.1175/JAMC-D-14-0065.1
Figure 6c shows the observed relationship between in-cloud stable layer temperature increases and bulk cloud mixed layer temperature gradients; the in-cloud stable layer is the layer between the top of the mixed layer and cloud top. A stable layer ΔT = 0 indicates a fully mixed cloud layer with no stable layer present. Despite the larger LWP subclass at Barrow with more frequent cloud mixed layers near the moist adiabatic lapse rate, subsequent temperature inversion strengths generally match those observed for the lower LWP subclasses at Barrow and SHEBA. While the amount of cloud LWP correlates with the potential mixed layer temperature gradient (Figs. 6a,b), there is no distinct correlation with the subsequent in-cloud stable layer temperature increase.
Figure 5 suggests that, depending on LWP and cloud thickness, a substantial portion of the cloud layer (often 30%–50%) resides within a thermodynamically stable layer. Here, the thermodynamic properties of this in-cloud stable layer are examined. Histogram statistics of the in-cloud stable layer temperature change (temperature inversion strength) show a positive relationship with stable layer depth (Figs. 7a,b). Linear least squares robust fits were applied to observations (lines in Figs. 7a,b). With fit correlations of r2 = 0.84, 0.90, and 0.98 for SHEBA and Barrow lower LWP subclasses and Barrow larger LWP subclass, respectively, a deeper stable layer is generally a more stably stratified layer. Yet, as was shown in Fig. 6, no obvious correlation is observed between the in-cloud mixed layer temperature gradient and stable layer temperature increases; cloud-driven mixing in the presence of cloud liquid does not appear to completely erode the stable layer for either LWP subclass. Horizontal advection characterizing the thermodynamic setting near the cloud top appears to be an important mechanism influencing the stable portion of the cloud (e.g., Solomon et al. 2011; Shupe et al. 2013; Sedlar and Shupe 2014).
Relative frequency 2D distributions of in-cloud thermodynamic stable layer depth (m) and (a),(b) ΔT (°C) and (c),(d) Δq (g kg−1) across this layer, showing the observed relationships for the (left) LWP 0–50 g m−2 subclasses at Barrow (color shading) and SHEBA (stars) and the (right) LWP 75–300 g m−2 subclass at Barrow. Lines (red is Barrow; dashed black is SHEBA) indicate a linear least squares robust fit to the observations in (a) and (b), with correlation coefficients squared between observations and fit in parentheses.
Citation: Journal of Applied Meteorology and Climatology 53, 12; 10.1175/JAMC-D-14-0065.1
Changes in the specific humidity q within this stable layer are also observed (Figs. 7c,d). In-cloud stable layer depth and Δq across this layer reveal a more scattered relationship with both increases and decreases in q; because of the scatter, linear regressions were meaningless. Barrow histogram probabilities and the SHEBA scatterplot of Δq indicate a higher frequency of q increases within the thermally stable portion of the cloud for the lower LWP subclasses (Fig. 7c vs Fig. 7d). Of the lower LWP subclasses at Barrow and SHEBA having a change in q above the in-cloud mixed layer, the relative occurrence of a q inversion is 66% and 78%, respectively. When compared with 48% at Barrow for LWP > 75 g m−2, these results indicate that q inversions are more frequently connected with temperature inversions when the clouds are optically thinner. It is important to note that these humidity changes are estimated over the full stable layer below cloud top. Given the median range of stable layer depth is often 100–200 m, geometrically thinner inversions not spanning the full layer depth may be present and missed in these statistics. The more frequent presence of q increases within the stable cloud portion for optically thinner clouds are in better agreement with the frequently observed q inversions reported in Sedlar et al. (2012). These increases in q further highlight the role of horizontal advection in modifying the thermal structure of these clouds.
d. Cloud LWP impact on emissivity
Figure 8a shows the relationship between in-cloud mixed layer depth and cloud LWP. Considering the MWR uncertainty in LWP retrievals, only a slight increase in cloud mixed layer depth correlates with increasing LWP for the 0–50 g m−2 LWP subclass (blue and black); more modest increases in mixed layer depth are observed for the larger LWP subclass (Fig. 8a, red). However, even for these increased LWPs, there is still a regime of relatively thin cloud mixed layers that do not increase with LWP. Following the ϵ dependence on LWP [Eq. (3)], a broad correlation is observed where increased cloud condensate promotes larger emissivities, enhanced radiative divergence, and potentially increased cloud mixing and mixing depths.
Observed relationship between cloud LWP (g m−2) and the depth of the (a) in-cloud mixed layer (m) and (b) in-cloud stable layer (m). Blue and red circles represent the 0–50 and 75–300 (g m−2) LWP subclasses at Barrow, respectively, and the SHEBA 0–50 (g m−2) LWP subclass is represented by black stars. Box-and-whisker plots (median, interquartile, and 10th–90th percentile ranges) overlaid show the statistical distribution of layer depth between in-cloud mixed and stable layer depths for Barrow LWP 0–50 g m−2 (cyan), SHEBA LWP 0–50 g m−2 (magenta), and Barrow LWP 75–300 g m−2 (black).
Citation: Journal of Applied Meteorology and Climatology 53, 12; 10.1175/JAMC-D-14-0065.1
In Fig. 8b, the depth of the in-cloud stable layer (layer between in-cloud mixed layer top and cloud top) shows no dominant signal relative to cloud LWP. Approximately 45% of the larger LWP subclass at Barrow (Fig. 8, red) has no in-cloud stable layer, relative to 37% for the smaller LWP subclass (Fig. 8, blue); only 16% of SHEBA cloud observations are observed with no in-cloud stable layer. Stable layer thickness medians and interquartile ranges are quite similar for both Barrow LWP subclasses (cyan and black), with modest increases observed for SHEBA. These observations indicate an in-cloud stable layer is a frequent feature, regardless of cloud LWP. This suggests a rather limited correlation between increases in ϵ, via LWP increases, and a fully mixed cloud layer; the situation may be different for a transitioning cloud layer, unlike the statistics from “static” cloud layers examined here. Thus, while increasing LWP may lead to a deeper in-cloud mixed layer, the stable layer above up to cloud top is largely unaffected. This result suggests the in-cloud stable layer is most notably controlled by the thermodynamic characteristics aloft, hinting again toward the important role of horizontal advection.
e. Downwelling longwave radiation at cloud top
In this section, atmospheric longwave downwelling (LWD) radiation from above the cloud is examined to investigate its impact on in-cloud static mixing. Following the Stefan–Boltzmann law, broadband longwave radiation is a function of ϵ [Eq. (3)] and emitting temperature to the fourth power. The emissivity of the cloud-free atmosphere above these low clouds is, to a first order, dominated by the water vapor w content. Radiosounding profiles of ρa and w are integrated vertically to obtain column-integrated water vapor (CIWV) estimates. These estimates are used to examine the amount of CIWV above the cloud compared to the total CIWV value (Fig. 9). The observed distributions of the ratio of CIWV above cloud to the total CIWV generally show an increase with the full column estimates for both LWP subclasses (Fig. 9). However, the distributions in Fig. 9 indicate two distinguishable differences between LWP subclasses. First, the lower LWP subclasses have the distribution maxima at the total CIWV ranging over 0.25–0.5 cm (Fig. 9a), while the larger LWP subclass has a maxima found at slightly larger total CIWV values (0.5–0.75 cm, Fig. 9b). Second, the distributions of ratios of CIWV above cloud top relative to the total CIWV are larger for the 0–50 g m−2 LWP subclass relative to the larger LWP subclass; median CIWV ratios are 81% and 80% for Barrow and SHEBA clouds with LWP < 50 g m−2, as compared with 69% for larger LWP clouds at Barrow.
Relative frequency 2D distributions of the ratio of CIWV above cloud top normalized by the full CIWV content shown as a function of full CIWV (cm) for the (a) Barrow (color shading) and SHEBA (stars) 0–50 g m−2 LWP subclass and (b) Barrow (color shading) 75–300 g m−2 LWP subclass.
Citation: Journal of Applied Meteorology and Climatology 53, 12; 10.1175/JAMC-D-14-0065.1
Clear-sky longwave radiation fluxes are calculated using the Rapid Radiative Transfer Model (RRTM; Mlawer et al. 1997). Radiosounding thermodynamic profiles are input into RRTM. The relationship of CIWV to LWD reaching cloud top is shown in Fig. 10a. The increase in LWD with increasing CIWV is expected following the exponential relationship of ϵ [e.g., Eq. (3)]. The nonlinear behavior generally results in the largest increases/decreases in LWD as CIWV values range from near 0 to roughly 0.5 cm. As shown in Fig. 9a, the most frequently observed atmospheric water vapor estimates fall within this range when the low clouds have LWP < 50 g m−2. Thus, any changes to the CIWV under this relatively dry atmospheric column will have a large impact on the resultant LWD reaching cloud top. Additionally, the data indicate that clouds with LWP 0–50 g m−2 have a tendency to have larger values of LWD reaching cloud top compared to clouds with LWP > 75 g m−2. To understand the impact of column temperature on LWD, the relationship of LWD to the average pressure-weighted column temperature is shown in Fig. 10b. Both subclasses of LWP and both locations generally occurred under similar temperature ranges. The observed relationship in LWD is more linear compared to changes in CIWV, although slight nonlinearity is present. Comparing the two LWP subclasses, the data indicate a general similarity in the relationship between temperature and LWD, unlike CIWV where the lower LWP subclass tends to have the largest flux of LWD reaching cloud top (Fig. 10a). This suggests that CIWV value has a more dominant control on the LWD reaching cloud top than temperature.
Radiative transfer estimates of atmospheric LWD radiation (W m−2) reaching cloud top as a function of (a) observed CIWV (cm) and (b) pressure-weighted average column temperature (K). Blue and red circles represent 0–50 and 75–300 g m−2 LWP subclasses at Barrow, with SHEBA 0–50 g m−2 LWP as stars.
Citation: Journal of Applied Meteorology and Climatology 53, 12; 10.1175/JAMC-D-14-0065.1
The relationship between LWD reaching cloud top and CIWV is further investigated statistically in Fig. 11 as box and whisker distributions of LWD for different ranges of CIWV. Median and interquartile spreads of LWD tend to be larger at cloud tops when LWP is less than 50 g m−2 (Fig. 11, blue). For these CIWV ranges, the differences amount to approximately 5–10 W m−2 more longwave flux reaching the cloud top relative to cases when the cloud LWP was larger than 75 g m−2. Estimates of LWD reaching cloud top for SHEBA generally tend to be even larger than at Barrow for the same LWP range; atmospheric temperatures were generally warmer at SHEBA within this water vapor range and may be the primary cause of the enhanced LWD (Fig. 10b). These results indicate an increase in atmospheric opacity (Figs. 9, 10) has a direct correlation on the atmospheric longwave flux reaching the top of the cloud layer. Increased LWD flux may potentially lead to a modification of radiative divergence at cloud top.
Radiative transfer estimates of atmospheric LWD radiation (W m−2) reaching cloud top as a function of CIWV (cm). Barrow LWP subclasses are shown as box-and-whisker distributions of LWD [median (horizontal line), mean (cross), interquartile, and 10th–90th percentiles] for CIWV bins spanning 0.25 cm for LWP 0–50 g m−2 (blue) and 75–300 g m−2 (red); numbers above each box-and-whisker distribution indicate the number of observations in each CIWV bin. Data for SHEBA LWP 0–50 g m−2 are shown as scatter points (stars).
Citation: Journal of Applied Meteorology and Climatology 53, 12; 10.1175/JAMC-D-14-0065.1
Relationship between radiative cooling rate changes (ΔH) (K day−1) as a function of layer absorption depth (Δz) (m) estimated from Eq. (4). The ΔHs are shown for two anomalous values of atmospheric LWD reaching cloud top: +5 W m−2 (solid black) and +10 W m−2 (dashed black). Negative ΔH represents a reduction in the cooling rate over a particular Δz.
Citation: Journal of Applied Meteorology and Climatology 53, 12; 10.1175/JAMC-D-14-0065.1
A potential feedback link between the in-cloud mixing depth, cloud condensate production, and atmospheric opacity above cloud top appears to be supported by the following observations: given an initial cloud with a respective LWP, 1) in-cloud mixing may be reduced because of decreases in buoyant production when LWP is initially low; 2) a fraction of limited cloud condensate is distributed within the in-cloud stable layer, acting to reduce the effective radiative divergence of the cloud; 3) enhanced downwelling longwave from the atmosphere above the cloud further reduces the cloud cooling rate; and 4) the production of cloud condensate is thus further limited. Similar thermodynamic–radiative feedback loops have been modeled previously (e.g., Garrett et al. 2009; Petters et al. 2012), under the condition that these clouds need to have sufficiently low LWP (<50 g m−2) in order for emissivity changes to affect buoyancy production. If the initial cloud layer is a blackbody (LWP > 50 g m−2; e.g., Stephens 1978b), cloud-driven mixing is observed over a sufficiently deeper layer and able to sustain vertical motions necessary for larger condensate production. Even with a fraction of the condensate distributed in the stable, upper portion of the cloud, longwave divergence continues to drive condensation and sustain the larger LWPs. Also as shown, the emission of LWD from above the cloud is often reduced for these larger LWP clouds, further promoting cloud-top cooling. This system may transition toward an optically thinner cloud state with reduced mixing potential if either the moisture source is depleted or if ice-phase processes consume a portion of the cloud liquid.
4. Summary
Cloud and thermodynamic properties from two Arctic locations have been examined from surface-based remote sensors and regular radiosounding profiles. Motivated by recent studies observing stably stratified, low liquid water path (LWP) clouds over the Arctic (Sotiropoulou et al. 2014; Bennartz et al. 2013), a characterization on the occurrence, cloud physical properties, and thermodynamic structure has been undertaken here. Separation has been made between optically thin and thick clouds by segregating the results into subclasses of LWP: a low subclass (LWP 0–50 g m−2) and a high subclass (LWP 75–300 g m−2). Observations from 2002 to 2010 at the ARM North Slope of Alaska super site (Barrow) were analyzed together with 1997/98 SHEBA observations over the central Arctic sea ice.
Both locations are subject to relatively frequent single-layer clouds below 1.5 km, present in 27% to 30% of all profiles. In terms of LWP subclasses, these single-layer clouds were most commonly optically thin, with LWP 0–50 g m−2. As shown in the distributions of LWP, clouds with LWP < 50 g m−2 were below the MWR LWP retrieval uncertainty for nearly half of these observations. It is therefore difficult to postulate whether these clouds actually contain limited traces of liquid condensate or whether they are predominantly ice clouds. Cloud-base temperature distributions corresponding with these cases were often relatively cold (−20°C). This may reflect an increasing fraction of clouds dominated by ice crystals. However, since these observations fall within the retrieval uncertainty, droplets may also be present, with the overall cloud LWP limited by the colder temperatures and reduced water vapor following the Clausius–Clapeyron relation. The amount of single-layer clouds at SHEBA with LWP > 75 g m−2 was so few that it prevented an accurate description of their statistical characteristics. As a result, only the 0–50 g m−2 LWP subclass of clouds at SHEBA was analyzed.
Nearly all observations indicate the presence of a cloud-driven thermodynamic mixed layer. The primary production mechanism of this layer is related to cloud-top longwave divergence (cooling) and turbulent kinetic energy production through buoyant mixing (e.g., Paluch and Lenschow 1991). Generally, clouds with lower LWP had a shallower mixed layer than higher LWP clouds. A thermodynamic stable layer above the in-cloud mixed layer was supported by large increases in equivalent potential temperature; stable layers were observed to occur regardless of cloud LWP subclass, highlighting the importance of horizontal advection. Specific humidity inversions were observed together with temperature inversions; however, they occurred more frequently for clouds with LWPs ranging between 0 and 50 g m−2. Thus, a portion of these clouds was often in contact with a potential source of moisture originating near the cloud top (Solomon et al. 2011; Sedlar et al. 2012).
In-cloud mixing depth is driven by the effectiveness of LWP-increasing cloud emissivity, leading to increases in radiative divergence across rather shallow vertical layers. Both LWP subclasses reveal that the bulk cloud mixed layer was near the moist adiabatic lapse rate value. However, when the mixed layer was thin (approximately <150 m), bulk mixed layer temperature gradients were often subadiabatic and occasionally slightly stable for the lower LWP subclass. While it has been shown that in-cloud mixed layer depth increases with increasing LWP, increased LWP does not necessarily diminish the stable layer presence or depth between in-cloud mixed layer top and cloud top. Instead the stable and often moist layer in the vicinity of the cloud top is largely unaffected by in-cloud mixing and most notably controlled by the thermodynamic characteristics aloft as a result of horizontal advection.
The role of atmospheric emissivity in the layer above cloud top has also been examined. It is observed that the ratio of CIWV above the cloud top relative to the full atmospheric column value is larger above clouds with LWP < 50 g m−2 than above clouds with LWP > 75 g m−2. Radiative transfer calculations indicate increased emissivity above cloud top results in enhanced downwelling longwave radiation reaching cloud top. Statistically, clouds with LWP below 50 g m−2 were exposed to increases in downwelling longwave radiation by 5–10 W m−2 relative to optically thicker clouds. This additional energy flux into the cloud has the ability to reduce the radiative divergence and cause a reduction in radiative cooling rates up to 25 K day−1. The implications of this change may be critical in modifying cloud overturning circulation production, leading to a modification in the production of cloud condensate.
These results suggest a thermodynamic–radiative feedback loop may be partially responsible in determining the total condensed cloud water path: 1) in-cloud mixing may be reduced because of decreases in buoyant production when LWP is initially low or when cloud liquid is potentially absent altogether; 2) a fraction of limited cloud condensate is distributed within the in-cloud stable layer, acting to reduce the effective radiative divergence (the fraction of the cloud layer that is characterized by a stable layer is clearly larger for cloud LWP < 50 g m−2 than for LWP > 75 g m−2); 3) enhanced downwelling longwave from the atmosphere above cloud further reduces the cloud cooling rate; and 4) the production of cloud condensate is further limited, leading to a potentially weaker cloud-driven circulation and less condensate production. The results presented support this feedback loop, with implications leading to a preference on the amount of cloud LWP (e.g., Petters et al. 2012). Other processes, such as shortwave absorption by the cloud, the influence of entrainment on the local LWC profile, or mechanical wind mixing, may also be important mechanisms in determining in-cloud mixing, condensate production, and the depth and stability of the in-cloud stable layer; testing these processes, especially shortwave absorption heating in cloud, requires detailed knowledge on cloud microphysical properties. While estimates of droplet effective radius may be possible from these surface-based remote sensors (e.g., Shupe et al. 2008a and references within), the uncertainty in the retrievals would pose as a severe caveat. Thus, these results of the physical cloud thermodynamic features and the presentation of a thermodynamic–radiative feedback loop may serve as observational constraints on future cloud modeling studies. Such studies may prove effective at separating the contributions of physical processes leading to cloud condensate production, in-cloud mixing, and low-level cloud life cycle.
Acknowledgments
Thanks are given to the ARM Program for providing excellent, freely available datasets of clouds, surface meteorological data, and thermodynamic profiles. Thanks are also given to Matthew Shupe for providing the SHEBA cloud dataset. Three anonymous reviewers provided constructive comments and suggestions that greatly improved the scientific scope of this paper.
REFERENCES
Bennartz, R., and Coauthors, 2013: July 2012 Greenland melt extend enhanced by low-level liquid clouds. Nature, 496, 83–86, doi:10.1038/nature12002.
Clothiaux, E. E., T. P. Ackerman, G. G. Mace, K. P. Moran, R. T. Marchand, M. A. Miller, and B. E. Martner, 2000: Objective determination of cloud heights and radar reflectivities using a combination of active remote sensors at the ARM CART sites. J. Appl. Meteor., 39, 645–665, doi:10.1175/1520-0450(2000)039<0645:ODOCHA>2.0.CO;2.
Curry, J. A., 1986: Interactions among turbulence, radiation, and microphysics in Arctic stratus clouds. J. Atmos. Sci., 43, 90–106, doi:10.1175/1520-0469(1986)043<0090:IATRAM>2.0.CO;2.
Curry, J. A., W. B. Rossow, D. Randall, and J. L. Schramm, 1996: Overview of Arctic cloud and radiation characteristics. J. Climate, 9, 1731–1764, doi:10.1175/1520-0442(1996)009<1731:OOACAR>2.0.CO;2.
Garrett, T. J., L. F. Radke, and P. V. Hobbs, 2002: Aerosol effects on cloud emissivity and surface longwave heating in the Arctic. J. Atmos. Sci., 59, 769–778, doi:10.1175/1520-0469(2002)059<0769:AEOCEA>2.0.CO;2.
Garrett, T. J., M. M. Maestas, S. K. Krueger, and C. T. Schmidt, 2009: Acceleration by aerosol of a radiative-thermodynamic cloud feedback influencing Arctic surface warming. Geophys. Res. Lett., 36, L19804, doi:10.1029/2009GL040195.
Guiraud, F. O., J. Howard, and D. C. Hogg, 1979: A dual-channel microwave radiometer for measurement of precipitable water vapor and liquid. IEEE Trans. Geosci. Electron., 17, 129–136, doi:10.1109/TGE.1979.294639.
Harrington, J. Y., T. Reisin, W. R. Cotton, and S. M. Kreidenweis, 1999: Cloud resolving simulations of Arctic stratus: Part II: Transition-season clouds. Atmos. Res., 51, 45–75, doi:10.1016/S0169-8095(98)00098-2.
Herman, G., and R. Goody, 1976: Formation and persistence of summertime Arctic stratus clouds. J. Atmos. Sci., 33, 1537–1553, doi:10.1175/1520-0469(1976)033<1537:FAPOSA>2.0.CO;2.
Intrieri, J. M., C. W. Fairall, M. D. Shupe, P. O. G. Persson, E. L. Andreas, P. S. Guest, and R. E. Moritz, 2002: An annual cycle of Arctic surface cloud forcing at SHEBA. J. Geophys. Res., 107, 8039, doi:10.1029/2000JC000439.
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.
Marchand, R., T. Ackerman, E. R. Westwater, S. A. Clough, K. Cady-Pereira, and J. C. Liljegren, 2003: An assessment of microwave absorption models and retrievals of cloud liquid water using clear-sky data. J. Geophys. Res., 108, 4773, doi:10.1029/2003JD003843.
McFarquhar, G. M., and Coauthors, 2011: Indirect and semi-direct aerosol campaign: The impact of Arctic aerosols on clouds. Bull. Amer. Meteor. Soc., 92, 183–201, doi:10.1175/2010BAMS2935.1.
Mlawer, E. J., S. J. Taubman, P. D. Brown, M. J. Iacono, and S. A. Clough, 1997: Radiative transfer for inhomogeneous atmospheres: RRTM, a validated correlated-k model for the longwave. J. Geophys. Res., 102, 16 663–16 682, doi:10.1029/97JD00237.
Moran, K. P., B. E. Martner, M. J. Post, R. A. Kropfli, D. C. Welsh, and K. B. Widener, 1998: An unattended cloud-profiling radar for use in climate research. Bull. Amer. Meteor. Soc., 79, 443–455, doi:10.1175/1520-0477(1998)079<0443:AUCPRF>2.0.CO;2.
Morrison, H., G. de Boer, G. Feingold, J. Harrington, M. D. Shupe, and K. Sulia, 2012: Resilience of persistent Arctic mixed-phase clouds. Nat. Geosci., 5, 11–17, doi:10.1038/ngeo1332.
Paluch, I. R., and D. H. Lenschow, 1991: Stratiform cloud formation in the marine boundary layer. J. Atmos. Sci., 48, 2141–2158, doi:10.1175/1520-0469(1991)048<2141:SCFITM>2.0.CO;2.
Persson, P. O. G., 2012: Onset and end of the summer melt season over sea ice: Thermal structure and surface energy perspective from SHEBA. Climate Dyn., 39, 1349–1371, doi:10.1007/s00382-011-1196-9.
Petters, J. L., J. Y. Harrington, and E. E. Clothiaux, 2012: Radiative–dynamical feedbacks in low liquid water path stratiform clouds. J. Atmos. Sci., 69, 1498–1512, doi:10.1175/JAS-D-11-0169.1.
Pinto, J. O., 1998: Autumnal mixed-phase cloudy boundary layers in the Arctic. J. Atmos. Sci., 55, 2016–2038, doi:10.1175/1520-0469(1998)055<2016:AMPCBL>2.0.CO;2.
Rambukkange, M. P., J. Verlinde, E. W. Eloranta, C. J. Flynn, and E. E. Clothiaux, 2011: Using Doppler spectra to separate hydrometeor populations and analyze ice precipitation in multilayered mixed-phase clouds. IEEE Geosci. Remote Sens. Lett., 8, 108–112, doi:10.1109/LGRS.2010.2052781.
Sedlar, J., and M. Tjernström, 2009: Stratiform cloud-inversion characterization during the Arctic melt season. Bound.-Layer Meteor., 132, 455–474, doi:10.1007/s10546-009-9407-1.
Sedlar, J., and M. D. Shupe, 2014: Characteristic nature of vertical motions observed in Arctic mixed-phase stratocumulus. Atmos. Chem. Phys., 14, 3461–3478, doi:10.5194/acp-14-3461-2014.
Sedlar, J., M. Tjernström, T. Mauritsen, M. D. Shupe, I. M. Brooks, P. O. G. Persson, C. E. Birch, and C. Leck, 2011: A transitioning Arctic surface energy budget: The impacts of solar zenith angle, surface albedo and cloud radiative forcing. Climate Dyn., 37, 1643–1660, doi:10.1007/s00382-010-0937-5.
Sedlar, J., M. D. Shupe, and M. Tjernström, 2012: On the relationship between thermodynamic structure and cloud top, and its climate significance in the Arctic. J. Climate, 25, 2374–2393, doi:10.1175/JCLI-D-11-00186.1.
Shupe, M. D., 2011: Clouds at Arctic observatories. Part II: Thermodynamic phase characteristics. J. Appl. Meteor. Climatol., 50, 645–661, doi:10.1175/2010JAMC2468.1.
Shupe, M. D., and J. M. Intrieri, 2004: Cloud radiative forcing at the Arctic surface: The influence of cloud properties, surface albedo, and solar zenith angle. J. Climate, 17, 616–628, doi:10.1175/1520-0442(2004)017<0616:CRFOTA>2.0.CO;2.
Shupe, M. D., T. Uttal, S. Y. Matrosov, and A. S. Frisch, 2001: Cloud water contents and hydrometeor sizes during the FIRE Arctic Clouds Experiment. J. Geophys. Res., 106, 15 015–15 028, doi:10.1029/2000JD900476.
Shupe, M. D., and Coauthors, 2008a: A focus on mixed-phase clouds: The status of ground-based observational methods. Bull. Amer. Meteor. Soc., 89, 1549–1562, doi:10.1175/2008BAMS2378.1.
Shupe, M. D., P. Kollias, P. O. G. Persson, and G. M. McFarquhar, 2008b: Vertical motions in Arctic mixed-phase stratiform clouds. J. Atmos. Sci., 65, 1304–1322, doi:10.1175/2007JAS2479.1.
Shupe, M. D., V. P. Walden, E. Eloranta, T. Uttal, J. R. Campbell, S. M. Starkweather, and M. Shiobara, 2011: Clouds at Arctic observatories. Part I: Occurrence and macrophysical properties. J. Appl. Meteor. Climatol., 50, 626–644, doi:10.1175/2010JAMC2467.1.
Shupe, M. D., P. O. G. Persson, I. M. Brooks, M. Tjernström, J. Sedlar, T. Mauritsen, S. Sjogren, and C. Leck, 2013: Cloud and boundary layer interactions over the Arctic sea ice in late summer. Atmos. Chem. Phys., 13, 9379–9400, doi:10.5194/acp-13-9379-2013.
Solomon, A., M. D. Shupe, P. O. G. Persson, and H. Morrison, 2011: Moisture and dynamical interactions maintaining decoupled Arctic mixed-phase stratocumulus in the presence of a humidity inversion. Atmos. Chem. Phys., 11, 10 127–10 148, doi:10.5194/acp-11-10127-2011.
Sotiropoulou, G., J. Sedlar, M. Tjernström, M. D. Shupe, I. M. Brooks, and P. O. G. Persson, 2014: The thermodynamic structure of summer Arctic stratocumulus and the dynamic coupling to the surface. Atmos. Chem. Phys. Discuss., 14, 3815–3874, doi:10.5194/acpd-14-3815-2014.
Stephens, G. L., 1978a: Radiation profiles in extended water clouds. I: Theory. J. Atmos. Sci., 35, 2111–2122, doi:10.1175/1520-0469(1978)035<2111:RPIEWC>2.0.CO;2.
Stephens, G. L., 1978b: Radiation profiles in extended water clouds. II: Parameterization schemes. J. Atmos. Sci., 35, 2123–2132, doi:10.1175/1520-0469(1978)035<2123:RPIEWC>2.0.CO;2.
Trenberth, K. E., J. T. Fasullo, and J. Kiehl, 2009: Earth’s global energy budget. Bull. Amer. Meteor. Soc., 90, 311–323, doi:10.1175/2008BAMS2634.1.
Uttal, T., and Coauthors, 2002: Surface heat budget of the Arctic Ocean. Bull. Amer. Meteor. Soc., 83, 255–275, doi:10.1175/1520-0477(2002)083<0255:SHBOTA>2.3.CO;2.
Verlinde, J., M. P. Rambukkange, E. E. Clothiaux, G. M. McFarquhar, and E. W. Eloranta, 2013: Arctic multilayered, mixed-phase cloud processes revealed in millimeter-wave cloud radar Doppler spectra. J. Geophys. Res. Atmos., 118, 13 199–13 213, doi:10.1002/2013JD020183.
Walsh, J. E., and W. L. Chapman, 1998: Arctic cloud–radiation–temperature associations in observational data and atmospheric reanalyses. J. Climate, 11, 3030–3045, doi:10.1175/1520-0442(1998)011<3030:ACRTAI>2.0.CO;2.
Wang, X., and J. R. Key, 2005: Arctic surface, cloud, and radiation properties based on the AVHRR Polar Pathfinder dataset. Part I: Spatial and temporal characteristics. J. Climate, 18, 2558–2574, doi:10.1175/JCLI3438.1.
Westwater, E. R., Y. Han, M. D. Shupe, and S. Y. Matrosov, 2001: Analysis of integrated cloud liquid and precipitable water vapor retrievals from microwave radiometers during SHEBA. J. Geophys. Res., 106, 32 019–32 030, doi:10.1029/2000JD000055.
