Abstract

In this study, collocated satellite and buoy observations as well as satellite observations over an extended region during 2006–10 were used to quantify the humidity effects on marine boundary layer (MBL) aerosols. Although the near-surface aerosol size increases with increasing near-surface relative humidity (RH), the influence of RH decreases with increasing height and is mainly limited to the lower well-mixed layer. In addition, the size changes of MBL aerosols with RH are different for low and high surface wind () conditions as revealed by observations and Mie scattering calculations, which may be related to different dominant processes (i.e., the hygroscopic growth process during low wind and the evaporation process during sea salt production during high wind). These different hygroscopic processes under the different conditions, together with the MBL processes, control the behaviors of the MBL aerosol optical depth () with RH. In particular, under high conditions, the MBL stratifications effects can overwhelm the humidity effects, resulting in a weak relationship of MBL on RH. Under low conditions, the stronger hygroscopic growth can overwhelm the MBL stratification effects and enhance the MBL with increasing RH. These results are important to evaluate and to improve MBL aerosols simulations in climate models.

1. Introduction

Sea salt is one of the largest natural contributors to the global aerosol loading and thus plays a significant role in the global climate (Solomon et al. 2007). Sea salt dominates submicron and supermicron scatterers and total aerosol mass concentration in the marine boundary layer (MBL) (Sievering et al. 2004). However, its radiative forcing is still poorly simulated in models (Textor et al. 2006; Kinne et al. 2006). One important uncertainty is from the lack of a better understanding of the source mechanisms while another major challenge is its hygroscopic growth (Textor et al. 2006; Rind et al. 2009). Sea salt optical properties are dependent on size distribution, shape, chemical composition, and mixing state, all of which are strong functions of relative humidity (RH) (Pilinis et al. 1995; Nessler et al. 2005). Moreover, the water uptake of sea salt could affect their residence times, act as efficient cloud condensation nuclei (CCN), catalyze heterogeneous reactions, etc. (Charlson et al. 1984; Hegg et al. 1993; Covert et al. 1972; Gong et al. 2002; Randles et al. 2004; Wise et al. 2009; Jaeglé et al. 2011). However, the humidity effect on sea salt over the extended ocean is hard to assess and it remains an open topic for aerosol modeling (Textor et al. 2006).

Sea salt particles are formed by evaporation of the sea salt droplets introduced by the wind-driven sea-spray processes (Blanchard and Woodcock 1957; Smith et al. 1993; Clarke et al. 2003; O’Dowd and de Leeuw 2007). Whitecap formation onset occurs at surface wind speeds () of ~4 m s−1 and produces sea salt droplets through bubble bursting (O’Dowd and de Leeuw 2007). When wind speeds are above ~9 m s−1, sea salt droplets can be produced by the direct tearing of wave crests. Near the seawater surface, the RH is approximately 98% (Blanchard and Woodcock 1980; Lewis and Schwartz 2006). Therefore, when a sea salt droplet is introduced to the atmosphere, it has an ambient RH of 98%. Then the sea salt droplets evaporate to form sea salt aerosols. The particles reach equilibrium with the ambient environment instantaneously (Fitzgerald et al. 1998; Lewis and Schwartz 2006; Fan and Toon 2011). The near-surface-produced sea salt aerosols are mixed and transported vertically by complex boundary layer processes, which could counteract the humidity effects (Smirnov and Shifrin 1989; Smirnov et al. 1995; Sayer et al. 2012). The MBL is frequently decoupled (Bretherton and Wyant 1997; Wood and Bretherton 2004; Jones et al. 2011), which affects the vertical distributions of MBL aerosols. Nonetheless, the simple mixing layer model is still widely used in the current MBL studies and simulations (Caldwell et al. 2013).

Sea salt particles contain mostly NaCl and are naturally hygroscopic (Tang and Munkelwitz 1994; Martin 2000). When the RH increases to the deliquescence relative humidity (DRH), sea salt particles abruptly take up water to form saturated droplets. For a pure sea salt such as NaCl, the DRH is ~75%. If the RH decreases, the particle loses water and eventually reforms crystals (efflorescence) at a RH value significantly lower than the DRH. These hygroscopic properties of sea salt are well characterized in the laboratory (Tang and Munkelwitz 1994; Tang 1996 and 1997). However, the humidity effects on sea salt aerosols are still not well understood when interacting with MBL processes. Model-based study indicates that the hygroscopic growth of marine aerosols account for up to 50% of the MBL aerosol optical depth (τ) (Glantz et al. 2009). On the other hand, several observational studies showed that there is no correlation between the surface RH and as a whole, while a negative correlation and a positive correlation of and RH could be observed for RH < 75% and RH > 75%, respectively (Smirnov and Shifrin 1989; Smirnov et al. 1995; Sayer et al. 2012). The challenges in quantifying the humidity effects on MBL aerosols are mainly due to limited reliable observations and complex processes controlling MBL aerosols (Smirnov and Shifrin 1989; Smirnov et al. 1995; Glantz et al. 2009; Sayer et al. 2012; Luo et al. 2014b).

Unlike previous studies that focused on either laboratory studies which cannot account for MBL processes (Wise et al. 2009) or model-based studies with unresolved MBL processes (Glantz et al. 2009), this study aims to provide new insights on the humidity effects on MBL aerosol using extensive observations in the MBL. To achieve the goal, satellite-based buoy observations and operational meteorology datasets are combined. Buoy measurements provide reliable near-surface RH measurements, which overcome the large uncertainties in the current satellite remote sensing products when trying to retrieve RH at low altitudes. On the other hand, Cloud–Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO) provides aerosol optical properties and MBL structure information (Luo et al. 2014a,b). Combining these measurements enables us to analyze the humidity effects and the influence of MBL processes on MBL aerosol properties over remote oceans. The data sources and analysis method are introduced in section 2. The results and discussions are presented in section 3. Key findings are given in section 4.

2. Data and methodology

a. Data

Multiple satellite remote sensing, operational meteorology datasets, and buoy data during the period from June 2006 to December 2010 were used to carry out the study and are briefly summarized here.

  1. CALIPSO level 1B data: Cloud–Aerosol Lidar with Orthogonal Polarization (CALIOP) is a polarization-sensitive lidar capable of measuring backscatter intensity at wavelengths of 532 and 1064 nm (Winker et al. 2009), with a 333-m along-track footprint. CALIOP level 1B data provide calibrated and geolocated 532- and 1064-nm total attenuated backscatter ( and ) and 532-nm perpendicular polarization component (Hunt et al. 2009). The backscatter data were used to retrieve MBL and provide MBL aerosol size information by the total color ratio ().

  2. CloudSat geometric profile product (2B-GEOPROF): CloudSat carries a 94-GHz cloud profiling radar (CPR). The CloudSat antenna pattern provides an instantaneous footprint of approximately 1.3 km (at mean sea level). It has 125 vertical bins with a bin size of about 240 m. The cloud mask in the 2B-GEOPROF product is used to identify hydrometeors occurrence (Mace 2007).

  3. ECMWF-AUX data: ECMWF-AUX contains temperature and pressure profiles from the European Centre for Medium-Range Weather Forecasts (ECMWF) operational analysis interpolated in time and space to the CloudSat track (Partain 2004).

  4. Advanced Microwave Scanning Radiometer for Earth Observing System (EOS) (AMSR-E) level 3 daily Ocean Products version 5: The AMSR-E aboard Aqua (part of the A-Train) is a 12-channel, 6-frequency, passive microwave radiometer system (Kawanishi et al. 2003). It measures horizontally and vertically polarized brightness temperatures at 6.9, 10.7, 18.7, 23.8, 36.5, and 89.0 GHz. Spatial resolutions of the individual measurements vary from 5.4 km at 89 GHz to 56 km at 6.9 GHz. The daily AMSR-E Ocean Products are produced by Remote Sensing Systems (RSS; http://www.remss.com/). The products contain several important geophysical parameters retrieved from observations collected by the AMSR-E instrument, such as sea surface temperature (SST) and . The orbital data are mapped to 0.25° grid box and is divided into two maps based on ascending and descending passes. Preliminary validations of AMSR-E (Wentz et al. 2003) showed that the root-mean-square (RMS) difference in wind speed retrievals is 0.92 m s−1 with a bias of 0.57 m s−1. In this study, validation with buoy data showed that the RMS difference in AMSR-E wind speed retrievals is 1.63 m s−1 with a bias of −0.15 m s−1.

  5. Buoy data: Buoy data were obtained from the National Data Buoy Center (NDBC; http://www.ndbc.noaa.gov/). Fig. 1 shows the map of 63 NDBC moored-buoy locations used in this study. All historical buoy standard meteorological data was reported on the hour and represent 8-min averages. Several quality controls were performed to ensure that the measurements meet the WMO (2006) requirements (NDBC 2009), with wind speed accuracy of 0.55 m s−1, air temperature accuracy of 0.09°C, and dewpoint accuracy of 0.31°C. The RH is calculated by air temperature and dewpoint (NDBC 2009). The absolute error in RH resulted by the uncertainties in air temperature and dewpoint is less than 2.5%.

  6. AIRS level 2 products version 5 (AIRS Science Team and Texeira 2013): The Atmospheric Infrared Sounder (AIRS) is a facility instrument aboard the Aqua satellite, as a part of the A-Train satellites (Aumann et al. 2003). In combination with the Advanced Microwave Sounding Unit (AMSU) and the Humidity Sounder for Brazil (HSB), AIRS constitutes an innovative atmospheric sounding group of visible, infrared, and microwave sensors. The moisture profile is retrieved at 14 atmospheric layers between 1100 and 50 hPa and at 50-km horizontal resolution.

Fig. 1.

Buoy locations used in this study indicated by black dotted asterisks.

Fig. 1.

Buoy locations used in this study indicated by black dotted asterisks.

b. Methodology

Two datasets, temporally and spatially collocated satellite and buoy observations and satellite-only observations in the buoy data regions, were constructed for analysis.

1) Collocated satellite and buoy observations (DataSB)

A 3-h running average is applied to the buoy observations to account for the different spatial coverage of buoy and satellite measurements. Then, ECMWF datasets and cloud-free CALIOP backscattering data were collocated and averaged into ±0.5° grid box centered at each buoy location. The cloud information was determined by combining CloudSat cloud mask and attenuated lidar scattering ratio from CALIOP (Wang et al. 2008; Adhikari et al. 2010). The cloud-free condition here refers to no-cloud or optically thin cloud with cloud top higher than 8 km. A total 4588 cases were collocated.

The aerosol-layer identification and sea salt at 532-nm () retrieval method are the same as those in Luo et al. (2014b). The aerosol layer was identified by an improved threshold method to overcome the lower signal-to-noise ratio issue in the spaceborne lidar (Luo et al. 2014a,b). Then, several criteria were applied to identify the clean marine aerosol cases (Luo et al. 2014b). First, it should be a single layer with layer top within 0.3–3 km and a layer bottom below 0.1 km in order to exclude elevated dust or smoke layer. Second, a threshold of volume depolarization ratio (VDR) larger than 0.06 was further used to identify and exclude dust aerosols (Liu et al. 2008). Third, a threshold of integrated attenuated backscatter at 532 nm smaller than 0.01 sr−1 was applied to the remaining profiles to avoid possible clouds, continental aerosol, or smoke cases (Kiliyanpilakkil and Meskhidze 2011). After sea salt classification, the was retrieved by a forward iteration method. The lidar ratio in the retrieval is chosen as 25 referring to the comparison with MODIS observations (Luo et al. 2014b).

The derived aerosol layer top height is a good proxy for the boundary layer height (), as evaluated by the ground-based lidar and radio sound observations (Luo et al. 2014a). Luo et al. (2014a) showed that the bias and RMS difference of lidar-derived ΔH is −0.12 ± 0.24 km relative to the sounding observations, and further evaluation of CALIOP-derived ΔH with marine stratiform cloud-top height over the global ocean showed the bias and RMS difference of −0.08 ± 0.37 km. The and RH () were obtained from the collocated buoy observations. The is calculated by the collocated backscattering coefficients at 532 and 1064 nm.

2) Satellite-only observations (DataS)

To increase the data samples, a subset between 30°N and 30°S of the dataset achieved in Luo et al. (2014b) together with new estimation of near-surface RH from satellite observations was developed (total 291 222 cases). The dataset in Luo et al. (2014b) is a new 25-km global sea salt database with cloud-free CALIOP backscattering data, AIRS level 2 products, and ECMWF-AUX data collocated and averaged into AMSR-E 25-km footprint. The cloud information, aerosol-layer identification, aerosol-type classification, and retrievals were the same as detailed in the last section. Additionally, for DataS, the data within 200 km to the continents was removed to further lower the possible influence of misclassified other type aerosols from continents on the results in this paper.

Although collocated buoy observations and RH estimations from AIRS L2 observations show small mean bias and RMS (−0.6 ± 5.8%), the poor correlation coefficient (0.27), indicating large random error, made it difficult to study the humidity effect on by using the satellite-based humidity data directly (Luo et al. 2014b). The RH estimated from ECMWF products has better correlation coefficient of 0.52 and RMS of , but with large mean bias −10.7%. Therefore, an alternative way to estimate RH was explored by combining AIRS near-surface water vapor mixing ratio (), ECMWF near-surface air temperature (), and from AMSR-E. To achieve this, , , and from AMSR-E were collocated into DataSB, and regression of on these parameters was performed to give the relationship as following

 
formula

Here, is the estimated RH in percentage. The unit of is grams per kilogram. The unit of is kelvins. The unit of is meters per second. Relating to buoy observations, the regressive relationship provides better mean bias and RMS of with the correlation coefficients of 0.52. Furthermore, as shown in Fig. 2, the mean accords well with the mean along the 1:1 line under different bins, and thus the statistical analysis based on in this study is convincible. Relating to , RH from AIRS L2 or from ECMWF has poor slopes of ~0.2 or ~0.4 under different bins, respectively.

Fig. 2.

Comparison between and for different bins. The Us bins are 1 ≤ Us < 3 m s−1, 3 ≤ Us < 6 m s−1, and 6 ≤ Us < 9 m s−1. The vertical lines show the standard deviations.

Fig. 2.

Comparison between and for different bins. The Us bins are 1 ≤ Us < 3 m s−1, 3 ≤ Us < 6 m s−1, and 6 ≤ Us < 9 m s−1. The vertical lines show the standard deviations.

In DataS, was obtained from the collocated AMSR-E observations and was estimated by applying Eq. (1) to DataS. The is calculated by the collocated backscattering coefficients at 532 and 1064 nm.

c. Mie scattering calculations

Mie scattering calculations were performed to further investigate the optical characteristics in the hygroscopic growth process and evaporation process under different and RH conditions, so as to understand how CALIOP measured and can depend on sea salt size and RH. The Mie scattering code is taken from Bohren and Huffman (1983). The refractive indexes of sea salt at 532 and 1064 nm are taken from Shettle and Robert (1979) and are interpolated to ambient RHs. The effective radius () is calculated according to Hansen and Travis (1974). The particle color ratio () definition in Mie scattering is based on Eq. (3) in Bi et al. (2009). The definition with the columnar aerosol size distribution is based on Eq. (15) in King (1982).

The size distributions used in the simulations at different were taken from Table 4 in Sayer et al. (2012), which are best-fit relationships to the averaged columnar size distributions over several ranges from the AERONET observations, as shown in Fig. 3a. Because the size distributions were derived from the averaged distributions, the RH of 80% for these size distributions was assumed according to the mean RH in the buoy observations. The assumed RH of 80% is larger than DRH, which ensures that the sea salt at lower RH than 80% has the same starting size distributions in the simulated hygroscopic growth process and evaporation process according to the particle weight changes in these processes (Tang and Munkelwitz 1994).

Fig. 3.

(a) The observed size distributions at different groupings; (b) the hygroscopic growth factors; (c) the relationship between and from Mie simulations by using the size distributions in (a); (d) the relationship between and from Mie simulations by using the size distributions in (a); (e) simulated –RH relationships under different groupings; (f) simulated –RH relationships under different groupings. In (c)–(f), the lines with asterisks denote the simulation results by GFS, the lines with circles denote the simulation results by GFSR, and different colors indicate different value ranges. The Us ranges are 0 ≤ Us < 4 m s−1, 4 ≤ Us < 6 m s−1, 6 ≤ Us < 8 m s−1, and 8 ≤ Us < 10 m s−1.

Fig. 3.

(a) The observed size distributions at different groupings; (b) the hygroscopic growth factors; (c) the relationship between and from Mie simulations by using the size distributions in (a); (d) the relationship between and from Mie simulations by using the size distributions in (a); (e) simulated –RH relationships under different groupings; (f) simulated –RH relationships under different groupings. In (c)–(f), the lines with asterisks denote the simulation results by GFS, the lines with circles denote the simulation results by GFSR, and different colors indicate different value ranges. The Us ranges are 0 ≤ Us < 4 m s−1, 4 ≤ Us < 6 m s−1, 6 ≤ Us < 8 m s−1, and 8 ≤ Us < 10 m s−1.

The size distributions were then adjusted from 80% to 18 RHs by using the method of Zhang e t al. (2005) according to changes of the growth factor. Referring to the particle weight changes in the hygroscopic growth process and evaporation process (Tang and Munkelwitz 1994), two sets of growth factors at ambient RHs—Shettle and Robert (1979, GFSR) and Spada et al. (2013, GFS)—were used to perform the size adjustments. As shown in Fig. 3b, a jump could be found in the GFSR near the DRH of ~75%, which describes the particle weight changes in the hygroscopic growth, while the GFS has a smooth trend down to the efflorescence point of ~45%, which describes the particle weight changes in the evaporation growth. After the RH becomes larger than 80%, GFS is close to GFSR. The simulation results with RH by GFS are similar to that by the widely used growth factor from Gerber (1985, GFG). Therefore, the results by GFG will not be shown in this paper.

Two sets of Mie scattering simulations under different and RH conditions were performed with size distributions adjusted by GFSR and GFS, respectively. The simulation results were presented in Figs. 3c–f.

As shown in Fig. 3c, the increases with under the same . Though the is the ratio of attenuated backscatter with contributions from both molecules and aerosols, the variation in can be resulted only by the aerosol size variation when the molecule backscattering is the same. Therefore, the changing of with RH under the same wind speed and the same molecular scatterings indicate the changing of the mean aerosol size with RH, which will be employed in the rest of the analysis to avoid errors in calculating associated with retrieval uncertainties. Figure 3d showed that under the same the increases with increase of too. Figures 3e and 3f present the change of and under hygroscopic growth and evaporation process, respectively. Both and showed jumps in hygroscopic growth process near the deliquescence point of 75%, which correspond well with the size changes according to the growth factors in Fig. 3b. The increases of from RH of 75%–80% is ~0.13 and ~0.042 for , at slopes of 0.026 and 0.0084 %−1, respectively. The simulations produced the gradual change of and with RH in the evaporation process.

3. Results and discussion

Luo et al. (2014b) showed that the MBL is mainly controlled by and . To further discriminate the humidity effects, data were sorted into different bins of , , and RH for DataS. For DataSB, because of the insufficient sampling issues, four categories were used in the analysis. All collocated cases were grouped to two categories, which are the high conditions with between 6 and 9 m s−1 (total 2506) and the low conditions with between 1 and 4 m s−1 (total 1209). Because the production of sea salt is controlled by and whitecap formation onset occurs at of ~4 m s−1, these two categories represent high and low sea salt source conditions. Then the two categories were further grouped into the high RH condition (RH > 85%) and the low RH condition (RH < 80%). Furthermore, to minimize the influence of molecular signal on , only in DataS with the near-surface molecular scattering coefficient ranging from to sr−1 km−1 in the 532-nm channel was used here (total 30 005 cases). The range of the molecule scattering coefficient is determined by its distribution as the most population region. The variations in the molecular scattering (standard deviation/mean value) are less than 1.35% below 3 km. The results of in DataS within other ranges of molecular scattering coefficients are similar and thus were not shown here. The range selection of molecular scattering coefficient was not performed on DataSB because of the small data sample.

a. The vertical distribution of aerosol properties

Sea salt aerosol can take up water and grow to larger size at higher RH. The near-surface RH can more directly influence the in lower 15% of the MBL, as indicated by the relationship between the mean with the normalized height by (figure not shown here). Therefore, the relationship of RH with mean within the lower 15% of the MBL (), ranging from ~3 range bins when km to ~13 range bins when km, was presented in Fig. 4. Because of the limited samples in DataSB, Fig. 4a presents the –RH relationships under two wide ranges in DataSB (black lines), overlaid with the corresponding relationships in DataS (red lines). Both datasets confirmed that for different categories the s at higher RHs are larger than those at lower RHs, with significant differences at the confidence level of 0.05 between high and low RH categories as given by the one-way analysis of variance (ANOVA). However, the changes in with increasing RH have different characteristics for different wind conditions. Under high conditions, the increases steadily with RH with slope of ~0.004%−1 for DataSB and ~0.009%−1 for DataS. Under low conditions, there exists a transition region between RHs of ~83% and ~86% with a more rapid increasing of with RH, at a slope of ~0.023 %−1 for DataSB and ~0.020 %−1 for DataS. When RH is smaller than ~80%, the has weak dependence on RH. When RH exceeds ~88%, increases steadily with a slope of 0.013 %−1 for DataSB—smaller than that in the transition region. Referring to the simulated –RH relationships in Fig. 3d, the –RH relationship under high conditions is similar to that in the simulated evaporation processes, while the –RH relationship under low conditions is very similar to that in the simulated hygroscopic growth process, except that the transition region is between RH of 75% and 80% in the simulations. The transition point of RH (~80%) in observations is a little higher than 75%, which is actually quite impressive, can be expected from a procedure with inherently significant uncertainties in RH.

Fig. 4.

(a) Comparison of relationships between the low (lines with dots) and high (lines with open circles) conditions in DataSB (black lines) and DataS (red lines). (b) relationships under different bins in DataS. Here, is the mean in the lower 15% part of the MBL. For results from DataS in (a),(b), only data with near-surface molecular scattering coefficients ranging from to sr−1 km−1 in the 532-nm channel were used.

Fig. 4.

(a) Comparison of relationships between the low (lines with dots) and high (lines with open circles) conditions in DataSB (black lines) and DataS (red lines). (b) relationships under different bins in DataS. Here, is the mean in the lower 15% part of the MBL. For results from DataS in (a),(b), only data with near-surface molecular scattering coefficients ranging from to sr−1 km−1 in the 532-nm channel were used.

To further understand the effects of , the –RH relationships in DataS with 1 m s−1 bins were presented in Fig. 4b. Although the DataS has limited samples with RH > 90%, similar conclusions to those from Fig. 4a could still be drawn. When > 6 m s−1, shows a steady increase with RH with a slope of ~0.013 %−1. When < 6 m s−1, increases slightly with RH for RHs < 80% with a slope < 0.005 %−1, and a more rapid increase in the transition region between RH of ~80%–90% with slopes of ~0.017 %−1–0.023 %−1. The one-way ANOVA showed that the difference between high and low RH categories at each bin is significant at the confidence level of 0.05. The difference between RH < 75% and 77.5% < RH < 82.5% is not significant at the confidence level of 0.05 when < 5 m s−1; but is when ≥ 5 m s−1. The simulations in Fig. 3e show smaller increases of with than in observations (Fig. 4b), which may be due to the fact that the size distributions from Sayer et al. (2012) are the columnar ones and are associated with smaller variations under different ranges (Fig. 3a) than those expected near surface.

The results may be interpreted to indicate that the hygroscopic properties of the MBL aerosols are changing with increasing , which could be the result of different underlying processes. Under high wind conditions, evaporation processes should be dominant because the wind-driven sea-spray processes introduce lots of sea salt droplets with RH of 98% into the atmosphere. Under the low wind conditions, without or with very low sea salt source, the hygroscopic growth process could be dominant in some situations when there are dry sea salt particles, which may closely relate to the MBL processes.

The MBL tends to be frequently decoupled especially when the MBL deepens (Bretherton and Wyant 1997; Wood and Bretherton 2004; Jones et al. 2011). The influences of the near-surface RH are mainly limited in the lower well-mixed layer. Figures 5 and 6 show the MBL structures in terms of and under different conditions. DataSB and DataS were grouped into different RH and categories and then were averaged into different bins. As shown in Figs. 5 and 6, both datasets showed that the MBL has a decoupled structure in terms of aerosol loadings and aerosol size. The lower well-mixed layer has large aerosol size and higher aerosol loading than in the upper-decoupled layer; and with increasing of , the decouple structure can be more obviously observed. A more interesting feature of the decoupling structure is that low RH conditions are associated with deeper mixed layers, while high RH conditions are associated with shallower mixed layers when comparing the MBL structures under different RH but the same and conditions. This indicates stronger stratifications near the mixing-layer top under the high RH conditions, which can limit the vertical transport of the sea salt aerosols. Our results are similar to the natural anticorrelation between RH and as suggested by Sayer et al. (2012), except that the natural negative correlation is between RH and mixing-layer height.

Fig. 5.

Comparisons of the MBL structure in terms of between the different possible low and high US and RH conditions for results from (top) DataSB and (bottom) DataS: (a),(e) both low; (b),(f) low and high RH; (c),(g) high and low RH; and (d),(h) both high. Because the DataS has more data samples, the low and high Us conditions in DataS are defined into a more narrow range as of 3 and 4 m s−1 and of 7 and 8 m s−1, respectively. The results of DataS with the same ranges as DataSB are similar (not shown here).

Fig. 5.

Comparisons of the MBL structure in terms of between the different possible low and high US and RH conditions for results from (top) DataSB and (bottom) DataS: (a),(e) both low; (b),(f) low and high RH; (c),(g) high and low RH; and (d),(h) both high. Because the DataS has more data samples, the low and high Us conditions in DataS are defined into a more narrow range as of 3 and 4 m s−1 and of 7 and 8 m s−1, respectively. The results of DataS with the same ranges as DataSB are similar (not shown here).

Fig. 6.

As in Fig. 5, but for the MBL structure in terms of , except that only data with the near-surface molecular scattering coefficient ranging from to sr−1 km−1 in a 532-nm channel were used.

Fig. 6.

As in Fig. 5, but for the MBL structure in terms of , except that only data with the near-surface molecular scattering coefficient ranging from to sr−1 km−1 in a 532-nm channel were used.

Therefore, low RH conditions are associated with low MBL stratifications, which means stronger convection. The convection results in strong mixing of rising lower wet air and falling upper dry air. This process could result in large fluctuations in the near-surface RH and lower the near-surface RH by turbulence mixing. In a similar way, the sea salt could also be carried with the airflow to a dry environment and lose most of its water, and sea salt aerosols within the upper dry air could be transported into the lower wet air and experience hygroscopic growth there. This process is weaker under high RH conditions because of weaker convection and mixing in the presence of the stronger MBL stratifications or becomes less important when there are rich sea salt particles produced under the high conditions. This hypothesis is also supported by near-surface RH variations in the buoy observations. The standard deviations of the near-surface RH (STDRH) within 3 h centered at the collocated time were calculated under low wind conditions. The STDRH is 12.1% when RH between 65% and 75%, 7.1% when RH between 75% and 85%, and only 4.8% when RH between 85% and 95%.

b. Dependence of τ on RH

Figure 7 shows the –RH relationships from two collocated datasets. Based on the DataSB, Fig. 7a shows that under the low conditions, with high RH are systematically larger than those with lower RH under different bins, with significant difference at the confidence level of 0.05 except when > 1.5 km by the one-way ANOVA. However, under high conditions (Fig. 7b), the enhancement of due to the surface RH is minor, and no uniform relationship with RH under different bins could be observed. The one-way ANOVA shows that there are no significant differences between high and low RH conditions under high conditions at a confidence level of 0.05. With the DataS, Fig. 7c clearly shows the humidity effects on under different with 1 m s−1 bins. When is lower than 80%, the has weak a relationship with RH and even decreases with RH, such as for bins of 4 or 10 m s−1. When is higher than 80%, the shows strong increasing with RH with slope in of ~0.005 %−1 when ≤ 4 m s−1 that decreases to ~0 %−1 when approaches 7 m s−1. Beyond of 7 m s−1, the shows very weak relationship with RH. The one-way ANOVA was performed between the high and low RH conditions under different bins of and . The result shows that there are significant differences at the confidence level of 0.05 between the high and low RH conditions when < 4 m s−1 and < 1.5 km (with mean difference of ~0.01) and when 4 < < 6 m s−1 and < 2 km (with a mean difference of ~0.0041) and no statistical significant differences when > 6 m s−1. These are consistent with results based on the DataS.

Fig. 7.

Comparisons of relationships between the low (blue line) and high (red line) RH conditions under the (a) low and (b) high conditions based on DataSB; (c) comparison of relationships with different bins for cases with between 0.8 and 1 km (total 40 255 cases) in DataS.

Fig. 7.

Comparisons of relationships between the low (blue line) and high (red line) RH conditions under the (a) low and (b) high conditions based on DataSB; (c) comparison of relationships with different bins for cases with between 0.8 and 1 km (total 40 255 cases) in DataS.

Both hygroscopic properties of sea salt particles and MBL processes could affect the behavior of the MBL . Under the high conditions, the size change of the MBL aerosols is relatively slower with RH variations (Fig. 3e). Meanwhile, the MBL stratification becomes stronger with increasing of RH and limits the vertical transport. Additionally, with increasing conditions, increasing of the surface wind shear will regroup the boundary layer turbulence structure and tends to decrease the well-mixed-layer height (Conzemius and Fedorovich 2006; Huang et al. 2009). The Mie simulations (Fig. 3f), which adjust the columnar size distribution with the given near-surface RH and thus ignore the other MBL processes, show that increases steadily with RH in the evaporation process under the high conditions. While in the observations of all MBL processes effects, has only weak or even negative relationship with RH when RH < 90%, as shown in Fig. 7c. Therefore, the complex MBL processes could limit the vertical distributions of MBL aerosols and outweigh the humidity effect, which results in weak humidity effects on the under high conditions. However, under low conditions, the aerosol size grows more rapidly with RH in transition region of RH and thus so does the (Figs. 3e and 3f), which could overwhelm the effects by MBL processes and enhance the with increasing of RH. Furthermore, as suggested by the simulation and observations, a similar phenomenon can also happen when RH close or larger than 90% with large growth factors under high or low conditions.

Several ship- or ground-based observations reported a positive correlation of RH and when RH > 75% and a negative correlation when RH < 75% (i.e., referring to Table 6 in Sayer et al. 2012). These studies suggested that the natural anticorrelation between and RH could outweigh the humidity effects and sometimes even results in the anticorrelation of RH with (Sayer et al. 2012). Because of the fact that the influences from other major factors such as and were not screened out in those past studies, our results are not comparable with theirs. Our results could not confirm the reported negative correlation under dry conditions. A positive correlation could only be found under the low wind conditions, and the transient RH of ~80% is larger than the ~75% in the former studies. Moreover, the decoupling structure of MBL that limits the vertical transport of sea salt aerosols and the surface humidity effects mainly within mixing-layer, and a natural anticorrelation between RH and the mixing-layer height, not as suggested by former studies, were found to be important to the MBL in our datasets.

4. Conclusions

The hygroscopic growth properties of MBL aerosol are different under different surface wind conditions. Generally, near-surface aerosol size increase with an increase of the near-surface RH. However, the influence of RH decreases with increasing height and is mainly limited within the mixing layer. Under the high wind conditions, the MBL aerosols grow gradually with increasing RH at a slope of ~0.0087 %−1 in terms of based on DataS. However, under the low wind conditions, there exists a transition point close to RH of 80%. Below the transition point, the MBL aerosols grow slowly with increasing of RH at a slope similar to that under the high wind condition; while above the point, the growth of aerosol size becomes rapid with a slope ~0.0203 %−1 in terms of based on DataS. The Mie scattering simulations results suggest that the different size growth behaviors under different wind condition may be related to 1) the evaporation process during sea salt production processes and 2) the hygroscopic growth process.

Further investigation showed that the influence of RH on MBL is controlled by both hygroscopic properties of sea salt particles and the MBL processes. The increase of RH could enhance the MBL under low conditions. With the increasing , the humidity effects on become smaller. Mie simulations without MBL processes showed that the increases in in the transition region (between RH of 75% and 80% in the simulation) can be as strong as ~0.042 in the hygroscopic growth process and larger than ~0.0065 in the evaporation process. With the influence of MBL processes, the observed increases in in the transition region between RH of 82% and 86% is ~0.02 when < 4 m s−1 and gradually decreases with increasing ; and generally no statistically significant increase of with RH can be observed when RH < 80%.

Therefore, the minor humidity effect on under high conditions can be connected to the formation of MBL stratifications. The MBL is frequently decoupled and the impacts of near-surface RH are limited within the shallow mixing layer. In this case, the natural anticorrelation between mixing-layer height with RH could outweigh the humidity effects. Sea salt aerosol has weaker hygroscopic growth under the high conditions or at low RH conditions than that at the transition region under the low conditions. Therefore, the humidity effects on are overwhelmed by the MBL stratification effects under the high conditions; while under the low conditions, the stronger hygroscopic growth could overwhelm the MBL stratification effects and enhance the MBL with increasing of RH.

However, owing to the lack of reliable observations of RH over high-latitude oceans, the results presented in this study are mainly representative of the tropics. Over the oceans, the high occurrence range for usually lies between 5 and 10 m s−1. Therefore, it is difficult to discriminate the humidity effects on integrated MBL through remote sensing in most of the situations over the oceans. On the contrary, the humidity effects on near-surface aerosol size could be easily identified by remote sensing of up to 10 m s−1. This paper provides a good guideline for the further study in other regions to investigate humidity effects on MBL aerosols and infers other possible regional differences. The results should prove useful for model improvements and evaluations.

Acknowledgments

This research was partially funded by NASA Grant NNX13AQ41G. Dr. Tao Luo’s effort is also funded by the National Natural Science Foundation of China (41105018). The authors acknowledge the editors’ and referees’ efforts in improving the manuscript.

REFERENCES

REFERENCES
Adhikari
,
L.
,
Z.
Wang
, and
D.
Liu
,
2010
:
Microphysical properties of Antarctic polar stratospheric clouds and their dependence on tropospheric cloud systems
.
J. Geophys. Res.
,
115
,
D00H18
, doi:.
AIRS Science Team
, and
J.
Texeira
,
2013
: Aqua AIRS level 2 standard physical retrieval (AIRS+AMSU), version 006. NASA Goddard Earth Science Data and Information Services Center (GES DISC), doi:.
Aumann
,
H. H.
, and Coauthors
,
2003
:
AIRS/AMSU/HSB on the Aqua mission: Design, science objectives, data products, and processing systems
. IEEE Trans. Geosci. Remote Sens.,41, 253–264, doi:.
Bi
,
L.
,
P.
Yang
,
G. W.
Kattawar
,
B. A.
Baum
,
Y. X.
Hu
,
D. M.
Winker
,
R. S.
Brock
, and
J. Q.
Lu
,
2009
:
Simulation of the color ratio associated with the backscattering of radiation by ice particles at the wavelengths of 0.532 and 1.064 μm
.
J. Geophys. Res.
,
114
,
D00H08
, doi:.
Blanchard
,
D. C.
, and
A. H.
Woodcock
,
1957
:
Bubble formation and modification in the sea and its meteorological significance
.
Tellus
,
9
,
145
158
, doi:.
Blanchard
,
D. C.
, and
A. H.
Woodcock
,
1980
:
The production, concentration, and vertical distribution of the sea-salt aerosol
.
Ann. N. Y. Acad. Sci.
,
338
,
330
347
, doi:.
Bohren
,
C. F.
, and
D. R.
Huffman
,
1983
: Absorption and Scattering of Light by Small Particles. John Wiley & Sons, 544 pp.
Bretherton
,
C. S.
, and
M. C.
Wyant
,
1997
:
Moisture transport, lower-tropospheric stability, and decoupling of cloud-topped boundary layers
.
J. Atmos. Sci.
,
54
,
148
167
, doi:.
Caldwell
,
P.
,
Y.
Zhang
, and
S.
Klein
,
2013
:
CMIP3 subtropical stratocumulus cloud feedback interpreted through a mixed-layer model
.
J. Climate
, 26,
1607
1625
, doi:.
Charlson
,
R. J.
,
D. S.
Covert
, and
T. V.
Larson
,
1984
: Observations of the effect of humidity on light scattering by aerosols. Hygroscopic Aerosols, T. H. Ruhnke and A. Deepak, Eds., A. Deepak Publishers, 35–44.
Clarke
,
A. D.
,
V.
Kapustin
,
S.
Howell
,
K.
Moore
,
B.
Lienert
,
S.
Masonis
,
T.
Anderson
, and
D.
Covert
,
2003
:
Sea salt size distributions from breaking waves: Implications for marine aerosol production and optical extinction measurements during SEAS
.
J. Atmos. Oceanic Technol.
,
20
,
1362
1374
, doi:.
Conzemius
,
R. J.
, and
E.
Fedorovich
,
2006
:
Dynamics of sheared convective boundary layer entrainment. Part I: Methodological background and large-eddy simulations
.
J. Atmos. Sci.
,
63
,
1151
1178
, doi:.
Covert
,
D. S.
,
R. J.
Charlson
, and
N. C.
Ahlquist
,
1972
:
A study of the relationship of chemical composition and humidity to light scattering by aerosols
.
J. Appl. Meteor.
,
11
,
968
976
, doi:.
Fan
,
T.
, and
O. B.
Toon
,
2011
:
Modeling sea-salt aerosol in a coupled climate and sectional microphysical model: Mass, optical depth and number concentration
.
Atmos. Chem. Phys.
,
11
,
4587
4610
, doi:.
Fitzgerald
,
J. W.
,
W. A.
Hoppel
, and
F.
Gelbard
,
1998
:
A one-dimensional sectional model to simulate multicomponent aerosol dynamics in the marine boundary layer: 1. Model description
.
J. Geophys. Res.
,
103
,
16 085
16 102
, doi:.
Gerber
,
H. E.
,
1985
: Relative-humidity parameterization of the Navy Aerosol Model (NAM). Naval Research Laboratory Rep. 8956, 13 pp. [Available online at http://www.dtic.mil/cgi-bin/GetTRDoc?Location=U2&doc=GetTRDoc.pdf&AD=ADA163209.]
Glantz
,
P.
,
E.
Nilsson
, and
W.
von Hoyningen-Huene
,
2009
:
Estimating a relationship between aerosol optical thickness and surface wind speed over the ocean
.
Atmos. Res.
,
92
,
58
68
, doi:.
Gong
,
S. L.
,
L. A.
Barrie
, and
M.
Lazare
,
2002
:
Canadian Aerosol Module (CAM): A size-segregated simulation of atmospheric aerosol processes for climate and air quality models 2. Global sea-salt aerosol and its budgets
.
J. Geophys. Res.
,
107
,
4779
, doi:.
Hansen
,
J. E.
, and
L. D.
Travis
,
1974
:
Light scattering in planetary atmospheres
.
Space Sci. Rev.
,
16
,
527
610
, doi:.
Hegg
,
D. A.
,
T. L.
Larson
, and
P. F.
Yuen
,
1993
:
A theoretical study of the effect of relative humidity on light scattering by tropospheric aerosols
.
J. Geophys. Res.
,
98
,
18 435
18 439
, doi:.
Huang
,
Q.
,
J. H.
Marsham
,
D. J.
Parker
,
W.
Tian
, and
T.
Weckwert
,
2009
:
A comparison of roll and nonroll convection and the subsequent deepening moist convection: An LEM case study based on SCMS data
.
Mon. Wea. Rev.
,
137
,
350
365
, doi:.
Hunt
,
W. H.
,
D. M.
Winker
,
M. A.
Vaughan
,
K. A.
Powell
,
P. L.
Lucker
, and
C.
Weimer
,
2009
:
CALIPSO lidar description and performance assessment
. J. Atmos. Oceanic Technol.,
26
,
1214
1228
, doi:.
Jaeglé
,
L.
,
P. K.
Quinn
,
T. S.
Bates
,
B.
Alexander
, and
J.-T.
Lin
,
2011
:
Global distribution of sea salt aerosols: New constraints from in situ and remote sensing observations
.
Atmos. Chem. Phys.
,
11
,
3137
3157
, doi:.
Jones
,
C. R.
,
C. S.
Bretherton
, and
D.
Leon
,
2011
:
Coupled vs. decoupled boundary layers in VOCALS-Rex
.
Atmos. Chem. Phys.
,
11
,
7143
7153
, doi:.
Kawanishi
,
T.
, and Coauthors
,
2003
:
The Advanced Microwave Scanning Radiometer for the Earth Observing System (AMSR-E), NASDA’s contribution to the EOS for global energy and water cycle studies
.
IEEE Trans. Geosci. Remote Sens.
,
41
,
184
194
, doi:.
Kiliyanpilakkil
,
V. P.
, and
N.
Meskhidze
,
2011
:
Deriving the effect of wind speed on clean maritime aerosol optical properties using the A-Train satellites
.
Atmos. Chem. Phys. Discuss.
,
11
,
4599
4630
, doi:.
King
,
M. D.
,
1982
:
Sensitivity of constrained linear inversions to the selection of the Lagrange multiplier
.
J. Atmos. Sci.
,
39
,
1356
1369
, doi:.
Kinne
,
S.
, and Coauthors
,
2006
:
An AeroCom initial assessment—Optical properties in aerosol component modules of global models
.
Atmos. Chem. Phys.
,
6
,
1815
1834
, doi:.
Lewis
,
E. R.
, and
S. E.
Schwartz
,
2006
:
Comment on “Size distribution of sea-salt emissions as a function of relative humidity.”
Atmos. Environ.
,
40
,
588
590
, doi:.
Liu
,
D.
,
Z.
Wang
,
Z.
Liu
,
D.
Winker
, and
C.
Trepte
,
2008
:
A height resolved global view of dust aerosols from the first year CALIPSO lidar measurements
.
J. Geophys. Res.
,
113
,
D16214
, doi:.
Luo
,
T.
,
R. M.
Yuan
, and
Z.
Wang
,
2014a
:
Lidar-based remote sensing of atmospheric boundary layer height over land and ocean
.
Atmos. Meas. Tech.
,
7
,
173
182
, doi:.
Luo
,
T.
,
R. M.
Yuan
, and
Z.
Wang
,
2014b
:
On factors controlling marine boundary layer aerosol optical depth
.
J. Geophys. Res. Atmos.
,
119
, 3321–3334, doi:.
Mace
,
G.
,
2007
: Level 2 GEOPROF product process description and interface control document algorithm version 5.3. CloudSat Project, 44 pp. [Available online at http://www.cloudsat.cira.colostate.edu/ICD/2B-GEOPROF/2B-GEOPROF_PDICD_5.3.doc.]
Martin
,
S. T.
,
2000
:
Phase transitions of aqueous atmospheric particles
.
Chem. Rev.
,
100
,
3403
3454
, doi:.
NDBC
,
2009
: Handbook of automated data quality control checks and procedures. National Data Buoy Center Tech. Doc. 09-02, 78 pp. [Available online at http://www.ndbc.noaa.gov/NDBCHandbookofAutomatedDataQualityControl2009.pdf.]
Nessler
,
R.
,
E.
Weingartner
, and
U.
Baltensperger
,
2005
:
Effect of humidity on aerosol light absorption and its implications for extinction and the single scattering albedo illustrated for a site in the lower free troposphere
.
J. Aerosol Sci.
,
36
,
958
972
, doi:.
O’Dowd
,
C. D.
, and
G.
de Leeuw
,
2007
:
Marine aerosol production: A review of the current knowledge
.
Philos. Trans. Roy. Soc. London
,
A365
,
1753
1774
, doi:.
Partain
,
P.
,
2004
: Cloudsat ECMWF-AUX auxiliary data process description and interface control document. CloudSat Project, 8 pp. [Available online at http://www.cloudsat.cira.colostate.edu/ICD/ECMWF-AUX/ECMWF-AUX_PDICD_3.0.pdf.]
Pilinis
,
C.
,
S. N.
Pandis
, and
J. H.
Seinfeld
,
1995
:
Sensitivity of direct climate forcing by atmospheric aerosols to aerosol size and composition
.
J. Geophys. Res.
,
100
,
18 739
18 754
, doi:.
Randles
,
C. A.
,
L. M.
Russell
, and
V.
Ramaswamy
,
2004
:
Hygroscopic and optical properties of organic sea salt aerosol and consequences for climate forcing
.
Geophys. Res. Lett.
,
31
,
L16108
, doi:.
Rind
,
D.
, and Coauthors
,
2009
: The way forward. Atmospheric aerosol properties and climate impacts, U.S. Climate Change Science Program Synthesis and Assessment Product 2.3, National Aeronautics and Space Administration, 85–90.
Sayer
,
A. M.
,
A.
Smirnov
,
N. C.
Hsu
, and
B. N.
Holben
,
2012
:
A pure marine aerosol model, for use in remote sensing applications
.
J. Geophys. Res.
,
117
,
D05213
, doi:.
Shettle
,
E. P.
, and
W. F.
Robert
,
1979
: Models for the aerosols of the lower atmosphere and the effects of humidity variations on their optical properties. Air Force Geophysics Laboratory Tech. Rep. AFGL-TR-79-0214, 23 pp. [Available online at http://www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA085951.]
Sievering
,
H.
,
J.
Cainey
,
M.
Harvey
,
J.
McGregor
,
S.
Nichol
, and
A. P.
Quinn
,
2004
:
Aerosol non-sea-salt sulfate in the remote marine boundary layer under clear-sky and normal cloudiness conditions: Ocean-derived biogenic alkalinity enhances sea-salt sulfate production by ozone oxidation
.
J. Geophys. Res.
,
109
,
D19317
, doi:.
Smirnov
,
A. V.
, and
K. S.
Shifrin
,
1989
:
Relationship of optical thickness to humidity of air above the ocean
. Izv. Atmos. Oceanic Phys.,25, 374–379.
Smirnov
,
A. V.
,
Y.
Villevalde
,
N. T.
O’Neill
,
A.
Royer
, and
A.
Tarussov
,
1995
:
Aerosol optical depth over the oceans: Analysis in terms of synoptic air mass types
.
J. Geophys. Res.
,
100
,
16 639
16 650
, doi:.
Smith
,
M. H.
,
P. M.
Park
, and
I. E.
Consterdine
,
1993
:
Marine aerosol concentrations and estimated fluxes over the sea
.
Quart. J. Roy. Meteor. Soc.
,
119
,
809
824
, doi:.
Solomon
,
S.
,
D.
Qin
,
M.
Manning
,
Z.
Chen
,
M.
Marquis
,
K.
Averyt
,
M.
Tignor
, and
H. L.
Miller
Jr.
, Eds.,
2007
: Climate Change 2007: The Physical Science Basis. Cambridge University Press, 996 pp.
Spada
,
M.
,
O.
Jorba
,
C.
Pérez
,
Z.
García-Pando
,
Z.
Janjic
, and
J. M.
Baldasano
,
2013
:
Modeling and evaluation of the global sea-salt aerosol distribution: Sensitivity to size-resolved and sea-surface temperature dependent emission schemes
.
Atmos. Chem. Phys.
,
13
,
11 735
11 755
, doi:.
Tang
,
I. N.
,
1996
:
Chemical and size effects of hygroscopic aerosols on light scattering coefficients
.
J. Geophys. Res.
,
101
,
19 245
19 250
, doi:.
Tang
,
I. N.
,
1997
:
Thermodynamic and optical properties of mixed-salt aerosols of atmospheric importance
.
J. Geophys. Res.
,
102
,
1883
1893
, doi:.
Tang
,
I. N.
, and
H. R.
Munkelwitz
,
1994
:
Aerosol phase transformation and growth in the atmosphere
.
J. Appl. Meteor.
,
33
,
791
796
, doi:.
Textor
,
C.
, and Coauthors
,
2006
:
Analysis and quantification of the diversities of aerosol life cycles within AEROCOM
.
Atmos. Chem. Phys.
,
6
,
1777
1813
, doi:.
Wang
,
Z.
,
G.
Stephens
, and
T.
Deshler
,
2008
:
Association of Antarctic polar stratospheric cloud formation on tropospheric cloud systems
.
Geophys. Res. Lett.
,
35
,
L13806
, doi:.
Wentz
,
F. J.
,
C.
Gentemann
, and
P.
Ashcroft
,
2003
: On-orbit calibration of AMSR-E and the retrieval of ocean products. 12th Conf. on Satellite Meteorology and Oceanography, Long Beach, CA, Amer. Meteor. Soc., P5.9. [Available online at https://ams.confex.com/ams/annual2003/techprogram/paper_56760.htm.]
Winker
,
D. M.
,
M. A.
Vaughan
,
A.
Omar
,
Y.
Hu
,
K. A.
Powell
,
Z.
Liu
,
W. H.
Hunt
, and
S. A.
Young
,
2009
:
Overview of the CALIPSO mission and CALIOP data processing algorithms
.
J. Atmos. Oceanic Technol.
,
26
,
2310
2323
, doi:.
Wise
,
M. E.
,
E. J.
Freney
,
C. A.
Tyree
,
J. O.
Allen
,
S. T.
Martin
,
L. M.
Russell
, and
P. R.
Buseck
,
2009
:
Hygroscopic behavior and liquid-layer composition of aerosol particles generated from natural and artificial seawater
.
J. Geophys. Res.
,
114
,
D03201
, doi:.
WMO
,
2006
: Guide to meteorological instruments and observations. 7th ed. WMO 8, 681 pp.
Wood
,
R.
, and
C. S.
Bretherton
,
2004
:
Boundary layer depth, entrainment, and decoupling in the cloud-capped subtropical and tropical marine boundary layer
.
J. Climate
,
17
,
3576
3588
, doi:.
Zhang
,
K. M.
,
E. M.
Knipping
,
A. S.
Wexler
,
P. V.
Bhave
, and
G. S.
Tonnesen
,
2005
:
Size distribution of sea-salt emissions as a function of relative humidity
.
Atmos. Environ.
,
39
,
3373
3379
, doi:.