Abstract

Small-scale turbulence has an essential role in sea-fog formation and evolution, but is not completely understood. This study analyzes measurements of the small-scale turbulence, together with the boundary layer structure and the synoptic and mesoscale conditions over the life cycle of a cold advection fog event and a warm advection fog event, both off the coast of southern China. The measurement data come from two sites: one on the coast and one at sea. These findings include the following: 1) For cold advection fog, the top can extend above the inversion base, but formation of an overlaying cloud causes the fog to dissipate. 2) For warm advection fog, two layers of low cloud can merge to form deep fog, with the depth exceeding 1000 m, when strong advection of warm moist air produces active thermal-turbulence mixing above the thermal-turbulence interface. 3) Turbulence near the sea surface is mainly thermally driven for cold advection fog, but mechanically driven for warm advection fog. 4) The momentum fluxes of both fog cases are below 0.04 kg m−1 s−2. However, the sensible and latent heat flux differ between the cases: in the cold advection fog case, the sensible and latent heat fluxes are roughly upward, averaging 2.58 and 26.75 W m−2, respectively; however, in the warm advection fog case, the sensible and latent heat flux are mostly downward, averaging −6.98 and −6.22 W m−2, respectively. 5) Low-level vertical advection is important for both fogs, but has a larger influence on fog development in the warm advection fog case.

1. Introduction

Sea-fog forecasting under different synoptic and mesoscale conditions is crucial for commerce and travel, yet still a complicated problem worldwide. Most studies of sea fog focus on particular areas such as the U.S. West Coast and China’s Yellow Sea because sea-fog characteristics differ by area (Koračin et al. 2014). For example, fog characteristics over the South China Sea are quite unique in many aspects including the formation, seasonal aspect, and extremely large fog-layer depths, but this area has seen relatively little study. The present study seeks to improve our knowledge of sea fog in this unique area.

However, improvement of sea-fog forecasting will require better observations of fog. In particular, the role of small-scale turbulence in fog needs to be clarified (Lewis et al. 2004; Koračin et al. 2014). Such small-scale characteristics depend on the fog type. And the type of fog is generally based on the primary formation mechanism (Gultepe et al. 2007). For sea fog, Lewis et al. (2004) describe four main types: advection fog, thermal buoyancy fog, radiative cooling fog, and steam fog. But of these, only advection fog is common off the coast of southern China, so we focus on advection fog here.

Advection fog was first studied by Taylor (1917). He showed that both warm air passing over a cold sea and cold air blowing over a warm sea can form sea fog. The former case is called warm advection fog here, which has the typical feature that the surface air temperature (SAT) exceeds the sea surface temperature (SST). We use this feature as its defining characteristic. Warm advection fog forms by the transport of heat via eddy diffusion (mechanical turbulence) from the upper warm saturated air to the sea surface and thereby cools the air adjacent to the sea. Taylor (1917) pointed out the formation mechanism, which was then later confirmed by Byers (1930) and Lamb (1943). The fog is then maintained by strong advection of warm, moist air over the cold sea surface. In contrast, the defining characteristic of cold advection fog is having the value of SAT be continuously less than SST.

Cold advection fog off the coast of southern China is similar to the sea fog on the U.S. West Coast, the latter of which was studied initially by Petterssen (1938), Emmons and Montgomery (1947), and Leipper (1948). It is also similar to the sea fog on the Scottish coast, known as the Haar, which has been studied by Lamb (1943) and Findlater et al. (1989). The fog forms initially by contact of warm air with cold sea, but then is maintained mainly by radiation from fog top, which depresses the fog temperature below SST (Findlater et al. 1989). Buoyancy is another driving force. When a stratus layer forms anterior to the fog, a combination of buoyancy and radiative cooling lowers the cloud to the sea surface, transforming the cloud to fog (Pilié et al. 1979; Findlater et al. 1989; Leipper 1994; Koračin et al. 2001, 2005). Sea fog with these two types of boundary layer characteristics also appears in the Yellow Sea, the East China Sea (Wang 1985; Huang and Zhou 2006; Zhang et al. 2012), the west coast of the Korean Peninsula (Kim and Yum 2012), and the northwest Pacific (Tachibana et al. 2008; Tanimoto et al. 2009).

Several studies have evaluated the turbulence in sea fog indirectly. Telford and Chai (1993) took a long aircraft flight in sea fog at altitude of 60 m on 25 September 1981, heading due west from San Francisco until reaching clear air. They found that the turbulence in fog (characterized by vertical motion) was extremely low. The vertical velocity ranged from 0.0 to −0.5 m s−1 in the sea fog with SAT > SST and fluctuated from −1.0 to 1.0 m s−1 under low cloud with SAT < SST. Later, Tachibana et al. (2008) used a series of GPS radiosondes and oceanic observations over the Okhotsk Sea in July 1998 to estimate the surface sensible and latent heat fluxes using bulk methods. They showed that most of the sensible heat flux was upward, as well as the latent heat flux, averaging 4.2 and 20 W m−2, respectively, from two cold advection sea-fog cases. Tanimoto et al. (2009) observed the Kuroshio Extension front using a GPS radiosonde and a research vessel over the sea to the east of Japan during the 2005 summer. They found that suppressed turbulent mixing within the marine atmospheric boundary layer favors the formation of warm advection fog.

Direct measurements of turbulence in sea fog are difficult, requiring sea platforms equipped for measurements of flux and other key air–sea elements. Two examples are the air–sea interaction tower (ASIT) in the Coupled Boundary Layers and Air–Sea Transfer Experiment in low winds (CBLAST-LOW; Edson et al. 2007) and the Ieodo Ocean Research Station (IORS) of Korea (Heo et al. 2010). Another problem with observational instruments arises from the liquid water. On a fast-response moisture sensor, such as a CSAT3 sonic anemometer (Campbell Scientific, Inc.), the moisture flux data can become erratic after fog formation (Heo et al. 2010).

Heo et al. (2010) analyzed warm advection fog cases over the East China Sea using the IORS data from 2004 to 2006. To reduce the flux-data noise, they used averaging times as long as 1 h, and even 3 h under a fog condition. They found that the sensible heat flux ranged mainly from −40 to 40 W m−2, with an extreme value of −70 W m−2. The latent heat flux instead fluctuated from −100 to 100 W m−2, with an extreme value of −200 W m−2. Although Huang et al. (2011) reveal the general characteristics of boundary layer structure in warm advection fog, the boundary layer structure characteristics of cold advection fog, and the turbulence characteristics of both types of sea fog, are not well known. Thus, more observations of turbulence in sea fog are needed.

To address this need, we investigate here the turbulence structure in two fog episodes that occurred in a tropical area just off the south China coast. One case is cold advection fog, the other is warm advection fog. As both fog types have been extensively studied by pioneers [cold advection fog has been well-studied since Petterssen (1938), and warm advection fog since Taylor (1917)], our aim is to use a well-equipped observation platform and GPS sondes to identify distinctive characteristics of the fog in this region. We use the data from the instruments to analyze the small- and large-scale characteristics of both fog types. We also present an investigation of boundary layer processes that drive the life cycle of sea fog near the southern China coast.

The rest of the paper is organized as follows. Section 2 introduces the observational site and data. Section 3 describes the synoptic background of two sea fog cases. The role of physical processes in sea fog is discussed in section 4. Section 5 analyzes the diabatic heating exchanges of sea fog. Section 6 summarizes the results and discusses the boundary layer, turbulence, and radiation characteristics of the sea fog.

2. Observational site and data

a. Observational site

Since 2006, the Institute of Tropical and Marine Meteorology (ITMM) has made a series of boundary layer structure observations of the sea fog off the coast of southern China. Observational data in this study were collected at the Marine Meteorological Science Experiment Base (MMSEB; Huang and Chan 2011) at Bohe, Maoming (21.45°N, 111.32°E, 7.0 m MSL), operated by the ITMM in spring 2011, site A in Fig. 1, hereafter the coast base, and at the Integrated Observation Platform for Marine Meteorology (IOPMM, 21.44°N, 111.39°E, operated by MMSEB), site B in Fig. 1, hereafter the sea base. The coast base lies on the west coastline of Guangdong province (Fig. 1a). The sea base lies just offshore over water about 15 m deep, about 8.2 km eastward from MMSEB, and about 6.5 km from the nearest coast (Fig. 1b). Established in 2008, IOPMM of the sea base is the first one of its type in China. The platform forms a regular triangle about 7 m on a side and about 11 m above sea level. The triangular platform has one angle pointing north, the two others pointing southwest and southeast. These southern corners each have a 10-m-high observational tower. The main observational tower is about 25 m high, and has 7 measurement level heights. Measured from sea level, these heights are 13.4, 16.4, 20.0, 23.4, 27.3, 31.3, and 35.1 m (Fig. 2). See Table 1 for the instruments used here. As they were solar powered, data collection was occasionally interrupted for lack of power on cloudy days (which are more common in warm advection fogs).

Fig. 1.

Observational site locations. (a) Overview of the Marine Meteorological Science Experiment Base (MMSEB; 21.45°N, 111.32°E). (b) Close-up of the region (from Google Earth) near site A, the MMSEB, and site B, the Integrated Observation Platform for Marine Meteorology (IOPMM; 21.44°N, 111.39°E).

Fig. 1.

Observational site locations. (a) Overview of the Marine Meteorological Science Experiment Base (MMSEB; 21.45°N, 111.32°E). (b) Close-up of the region (from Google Earth) near site A, the MMSEB, and site B, the Integrated Observation Platform for Marine Meteorology (IOPMM; 21.44°N, 111.39°E).

Fig. 2.

The IOPMM operated by the Guangzhou Institute of Tropical Marine and Meteorology (ITMM). The top-right corner of the image is the mounting position of the Gill R3-50 and Li-cor 7500A instruments at the 27.3-m level during the spring of 2011. (Main photo courtesy of Weikang Mao on 31 Aug 2009.)

Fig. 2.

The IOPMM operated by the Guangzhou Institute of Tropical Marine and Meteorology (ITMM). The top-right corner of the image is the mounting position of the Gill R3-50 and Li-cor 7500A instruments at the 27.3-m level during the spring of 2011. (Main photo courtesy of Weikang Mao on 31 Aug 2009.)

Table 1.

Sea base instruments on the IOPMM.

Sea base instruments on the IOPMM.
Sea base instruments on the IOPMM.

b. IOPMM data

We collected data on wind direction, wind speed, atmospheric temperature, and humidity (at the 13.4-m level). In addition, we used a net radiometer at about 12 m above sea level, infrared CO2/H2O analyzers and a sonic anemometer at the 27.3-m level, and a thermistor unit 5 m below the sea surface. We calibrated the gradient of temperature and humidity at five levels and the sea temperature at nine levels using a suite of thermometers and ventilated psychrometers. Our three dataloggers were synchronized prior to each sea-fog period.

As we could not avoid deposition of fog droplets, we used Gill R3-50 ultrasonic anemometers, which are unaffected by droplet deposition under cloudy conditions (Siebert and Teichmann 2000). We use eddypro 4.0 software (http://www.licor.com/env/products/eddy_covariance/software.html) to calculate turbulence flux over time intervals of 30 min. The software uses the Foken et al. (2004) method to run quality control tests of the fluxes. The overall quality flags follow the Spoleto agreement (the Second CarboEurope Workshop on QA/QC of eddy covariance measurements in Spoleto, January 2004) for CarboEurope-IP, as shown in Table 2.

Table 2.

Overall quality flags of the eddypro 4.0 software. RNcov is the parameter to evaluate the steady state of turbulence, and ITCσ is the parameter to evaluate the integral turbulence characteristic. More information is available in Foken et al. (2004).

Overall quality flags of the eddypro 4.0 software. RNcov is the parameter to evaluate the steady state of turbulence, and ITCσ is the parameter to evaluate the integral turbulence characteristic. More information is available in Foken et al. (2004).
Overall quality flags of the eddypro 4.0 software. RNcov is the parameter to evaluate the steady state of turbulence, and ITCσ is the parameter to evaluate the integral turbulence characteristic. More information is available in Foken et al. (2004).

c. Additional data

GPS sondes (Vaisala model RS-92) were released from observation coast base during the two sea-fog events (see Table 3 for details). One event was cold advection fog and the other was for warm advection fog. The sonde measurement accuracy for temperature is 0.2 K, for humidity 2%, for pressure 0.5 hPa, for wind direction 2°, and for wind speed 0.15 m s−1. A sonde was released every 2 h on the even hours, except during the warm advection fog, when they were also released at 1500 LST 20 March, 1330 LST 21 March, and 1900 LST 21 March. The sampling rate was 0.5 Hz and the vertical resolution was 4–10 m. The cold advection fog event had a total of six sondes and the warm advection fog event had eight. To achieve uniform data, the GPS sonde data were interpolated in the vertical with a 10-m interval. Calculation of the equivalent potential temperature () followed Bolton (1980). All the meteorological elements obtained from the GPS sonde data are smoothed using a three-point smoothing method in the vertical direction. We also use the logbook entries of observations by the crew at MMSEB (coast base), which were taken at 0200, 0800, 1100, 1400, 1700, 2000, and 2300 LST. Furthermore, no rain occurred during the two sea-fog cases. We use soundings from MMSEB to analyze their boundary layer structures. Both cases of sea fog lasted longer at the sea base than at the coast base (Table 3).

Table 3.

Analysis periods for two cases of sea fog at the coast base and the sea base (LST).

Analysis periods for two cases of sea fog at the coast base and the sea base (LST).
Analysis periods for two cases of sea fog at the coast base and the sea base (LST).

For the sea surface temperature data, we use the Real Time Global (RTG High Res 0.083, http://polar.ncep.noaa.gov/sst/) data from the National Centers for Environmental Prediction (NCEP) (Thiébaux et al. 2003). We calculate the backward airflow trajectories by using the HYSPLIT trajectory model and the reanalysis data from the Global Data Assimilation System (GDAS) (Draxler and Hess 1998). We also use the NCEP Global Analysis on 1° × 1° grid spacing datasets from the National Centers for Environmental Prediction/National Weather Service/NOAA/U.S. Department of Commerce (2000). The research data archive is at the National Center for Atmospheric Research, Computational and Information Systems Laboratory (http://rda.ucar.edu/datasets/ds083.2). Finally, we use cloud-top temperature data from MODIS [the Level 1 and Atmosphere Archive and Distribution System (LAADS)] website (Platnick et al. 2003).

3. Synoptic backgrounds of the two cases

Sea fog in southern China always occurs with low clouds (Huang et al. 2011). During winter and early spring here, stratus is very common, having a cloud fraction of nearly 50%. And each 6% increase in stratus fractional-area coverage is associated with an increase in static stability of 1 K (Klein and Hartmann 1993).

Sea fog in southern China has several climatic characteristics: 1) The sea fog occurs from January to May, averaging 3–5 days per month from February to April. 2) The fog appears about 100–200 km from the coast, usually surrounding the Leizhou Peninsula. 3) The fog always forms at the afternoon or night, and maintains through the next morning (Wang 1985). Moreover, a heavy fog case can extend over the entire southern China coast, lingering over 2–3 days. Of the 60 fog cases observed at MMSEB in 2007–12, 69% was warm advection fog, 28% was cold advection fog, and just 3% was other types (Huang 2013). We select two distinct fog cases according to the following criteria: 1) the fog case has a clear difference between SAT and SST, 2) the fog-layer depth exceeds 100 m, 3) the fog lasts over 4 h, 4) the fog is representative of its type, and 5) the dataset for the fog event is sufficient and with good quality.

The two sea-fog events in this study had similarities and differences. In both cases, a cold front moved south from Lingnan Mountain to the coast of southern China, and light fog existed at the coast base before each case (Table 4). Differences between the two fog events come mainly from the different airflows at low levels. In the cold advection fog case, the isobars are nearly parallel to the cold frontal line, and the airflow came mainly from the coastal cold-water area from the east at 20-m level (Fig. 3a). In the warm advection fog case, the isobars are nearly perpendicular to the cold frontal line, and the airflow mainly comes from the south (Fig. 3b).

Table 4.

Weather observations before and after two sea-fog events at the coast base. Stratocumulus translucidus (Sc tra), Fracto-cumulus (Fc), Altocumulus translucidus (Ac tra), Cumulus humilis (Cu hum), Stratocumulus opacus (Sc op), Stratocumulus cumulogenitus (Sc cug), total cloud cover/low cloud amount (10/10; using a scale of 1 to 10; 10 means larger than 9.5, but less than 10; 0 means less than 0.5). All these symbols follow the specifications for surface meteorological observations of China.

Weather observations before and after two sea-fog events at the coast base. Stratocumulus translucidus (Sc tra), Fracto-cumulus (Fc), Altocumulus translucidus (Ac tra), Cumulus humilis (Cu hum), Stratocumulus opacus (Sc op), Stratocumulus cumulogenitus (Sc cug), total cloud cover/low cloud amount (10/10; using a scale of 1 to 10; 10− means larger than 9.5, but less than 10; 0 means less than 0.5). All these symbols follow the specifications for surface meteorological observations of China.
Weather observations before and after two sea-fog events at the coast base. Stratocumulus translucidus (Sc tra), Fracto-cumulus (Fc), Altocumulus translucidus (Ac tra), Cumulus humilis (Cu hum), Stratocumulus opacus (Sc op), Stratocumulus cumulogenitus (Sc cug), total cloud cover/low cloud amount (10/10; using a scale of 1 to 10; 10− means larger than 9.5, but less than 10; 0 means less than 0.5). All these symbols follow the specifications for surface meteorological observations of China.
Fig. 3.

Airflow trajectories and daily average sea surface temperatures (SSTs) (°C). (a) SST on 28 Feb 2011 and cold-front location at 0000 LST 1 Mar. Dotted lines denote the 48-h backward trajectory of airflow at 20-, 100-, and 300-m heights at 0000 LST 1 Mar. (b) As in (a), but for SST at 21 Mar 2011 and the front and trajectories at 2000 LST 21 Mar.

Fig. 3.

Airflow trajectories and daily average sea surface temperatures (SSTs) (°C). (a) SST on 28 Feb 2011 and cold-front location at 0000 LST 1 Mar. Dotted lines denote the 48-h backward trajectory of airflow at 20-, 100-, and 300-m heights at 0000 LST 1 Mar. (b) As in (a), but for SST at 21 Mar 2011 and the front and trajectories at 2000 LST 21 Mar.

Moreover, the clouds differed between the two cases. Preceding the cold advection fog case, there were thin, sparse clouds that tended to dissipate (Fig. 4a, Table 4). But preceding the warm advection case, the sky was uniformly overcast (Fig. 4c, Table 4). During the diurnal period of the warm advection fog, the fog stretches across the coast (Fig. 4d). After the cold advection fog, the cloud amounts are sparse (Fig. 4b, Table 4). However, the sky was still overcast after the warm advection fog case at MMSEB (Fig. 4e, Table 4). Both fogs dissipated similarly as the cold front passed over the coast, pushing the fog offshore (Figs. 4b,e).

Fig. 4.

Cloud-top temperature (K) of the sea-fog events. In both cases, the yellow color represents the fog-top temperature, namely 290–293 K in the cold advection fog case and 291–294 K in the warm advection fog case. (a) 1400 LST 28 Feb, before the cold advection fog case. (b) 1130 LST 1 Mar, after the cold advection fog case. (c) 1335 LST 20 Mar, before the warm advection fog case. (d) 1415 LST 21 Mar, during the warm advection fog case. (e) 1320 LST 22 Mar, after the warm advection fog case. (From MODIS: http://ladsweb.nascom.nasa.gov/data/search.html.)

Fig. 4.

Cloud-top temperature (K) of the sea-fog events. In both cases, the yellow color represents the fog-top temperature, namely 290–293 K in the cold advection fog case and 291–294 K in the warm advection fog case. (a) 1400 LST 28 Feb, before the cold advection fog case. (b) 1130 LST 1 Mar, after the cold advection fog case. (c) 1335 LST 20 Mar, before the warm advection fog case. (d) 1415 LST 21 Mar, during the warm advection fog case. (e) 1320 LST 22 Mar, after the warm advection fog case. (From MODIS: http://ladsweb.nascom.nasa.gov/data/search.html.)

The difference in fog is due to the dissimilar characteristics of the advection of temperature and vapor. Hence, we need to analyze the transport of heat and moisture to understand the synoptic background of the two cases. The horizontal advection of temperature can be written as

 
formula

where is the horizontal wind vector, is the horizontal temperature gradient, and implies the difference across 1°. We used Eq. (1) for the advection of temperature calculated with NCEP Global Analyses (1° × 1° grid datasets at 1000 and 925 hPa), and we also calculate the advection of vapor with the same datasets. Based on the advection of temperature and wind vectors at the lower level, the cold advection fog case had weaker advection than the warm advection fog case at both 1000 and 925 hPa (Fig. 5). The region of 21°–22°N, 111°–112°E shows temperature changes of −0.040 × 10−4 and 0.048 × 10−4 K s−1 at 1000 hPa in the cold advection fog and warm advection fog cases, respectively. Similarly, the cold advection fog case also had less water vapor transport than the warm advection fog case at both 1000 and 925 hPa; specifically, for 1000 hPa, being −0.070 × 10−4 g kg−1 s−1 in the cold case and 0.260 × 10−4 g kg−1 s−1 in the warm case.

Fig. 5.

Horizontal advection of temperature (10−4 K s−1) and wind vectors at 1000 and 925 hPa before the sea fog. The red line is the zero line for the sign of advection. (top) Cold advection fog case at 0800 LST 28 Feb: (a) 1000 and (b) 925 hPa. (bottom) Warm advection fog case on 2000 LST 20 Mar: (c) 1000 and (d) 925 hPa.

Fig. 5.

Horizontal advection of temperature (10−4 K s−1) and wind vectors at 1000 and 925 hPa before the sea fog. The red line is the zero line for the sign of advection. (top) Cold advection fog case at 0800 LST 28 Feb: (a) 1000 and (b) 925 hPa. (bottom) Warm advection fog case on 2000 LST 20 Mar: (c) 1000 and (d) 925 hPa.

Overall, the large-scale synoptic background produces different flow patterns during the two types of sea fog. This indicates that the warm advection fog case has favorable conditions for both heating and vapor transport, which leads to a deeper fog layer than the cold advection fog case.

4. The role of physical processes in sea fog

a. Cold advection fog life cycle

First consider the formation mechanism of cold advection sea fog. As with advection-radiation fog over land, the conditions favorable for the formation of cold advection fog over the sea are (i) a high relative humidity, (ii) a clear sky, and (iii) a lack of turbulence (or stable stratification and a lack of wind). This particular cold advection fog case forms at 1650 LST 28 February and dissipates at 0840 LST 1 March.

We now analyze the meteorological and turbulent characteristics of the cold advection case at the sea base (Fig. 6, left side). Here, the SAT is below SST from 1440 LST 28 February to 0650 LST 1 March, starting before the fog appears and ending before the fog dissipates (Fig. 6a). Wind direction reveals a sea–land-breeze phenomenon in the cold advection fog case, which was absent in the warm advection fog case (Fig. 6b). The downward longwave radiation (DLR) and the upward longwave radiation (ULR) have a close relationship to the difference of the SAT and SST. In particular, when the SAT is lower than the SST in the cold advection fog case, the DLR is smaller than the ULR (Fig. 6c). The cold advection fog case endures lower shortwave radiation than in the warm advection fog case (Fig. 6d). In the cold advection fog case, the vertical velocity varies from negative to positive, but its magnitude remains below 0.1 m s−1 (Fig. 6e). By using the results from a one-dimensional mixed layer model, Findlater et al. (1989) point out that the sea fog in Scotland always undergoes a similar diurnal-modulated process. Similarly, we can see the sea–land breeze also connects with the formation of a cold advection fog in this area. Initially, when the land breeze ceases at 0940 LST 28 February, the wind direction turns from the north to the southeast and a sea breeze begins (Fig. 6b). The sea breeze accompanies the downward motion of the cold moist air, which results in the SAT going below SST, promoting the fog formation (Figs. 6a,e). Later, the sea breeze diminishes, ceasing around midnight (Fig. 6b) as the wind direction slowly changes from southeast to north. The vertical velocity also changes, going from negative to positive, and because the north wind brings warm, dry air, the fog gradually dissipates (Figs. 6a,e). This connection between the sea–land breeze and the formation–dissipation of cold advection fog occurs in other cases (Huang 2013).

Fig. 6.

Meteorological parameters and turbulence characteristics of the (left) cold advection fog and (right) warm advection fog. The missing data in the warm advection case from 2240 LST 20 Mar to 1020 LST 21 Mar are due to a power outage. (a) Relative humidity (RH) and SAT − SST is the difference of air and sea temperature. (b) Wind direction (Wd) and wind speed (Ws). (c) Downward longwave radiation (DLR) and upward longwave radiation (ULR). (d) Downward shortwave radiation (DSR) and upward shortwave radiation (USR). (e) Vertical velocity (w). (f) Tau is the momentum flux. (g) Sensible heat flux (H). (h) LE is the latent heat flux (missing data are due to being out of range or doubtful). Averaging interval for the meteorological parameters is 10 min and for the fluxes it is 30 min.

Fig. 6.

Meteorological parameters and turbulence characteristics of the (left) cold advection fog and (right) warm advection fog. The missing data in the warm advection case from 2240 LST 20 Mar to 1020 LST 21 Mar are due to a power outage. (a) Relative humidity (RH) and SAT − SST is the difference of air and sea temperature. (b) Wind direction (Wd) and wind speed (Ws). (c) Downward longwave radiation (DLR) and upward longwave radiation (ULR). (d) Downward shortwave radiation (DSR) and upward shortwave radiation (USR). (e) Vertical velocity (w). (f) Tau is the momentum flux. (g) Sensible heat flux (H). (h) LE is the latent heat flux (missing data are due to being out of range or doubtful). Averaging interval for the meteorological parameters is 10 min and for the fluxes it is 30 min.

Consider the fluxes. The momentum flux is very small, never exceeding 0.04 kg m−1 s−2 (Fig. 6f), and the sensible heat flux is mainly positive, ranging from 0 to 20 W·m−2 for SAT − SST < 0 (Fig. 6g). The shifting in the vertical profiles of the mixing ratio indicates complexity in the latent heat flux (Fig. 7). This flux fluctuates and has a lot of missing data (~39% missing, similar to other cold advection cases), with no evident characteristics (Fig. 6h).

Fig. 7.

(a)–(f) Boundary layer structure for the cold advection fog on 28 Feb–1 Mar. Horizontal lines: 1 is the height of the thermal turbulence interface (only exists in warm advection fog, determined as ), 2 is the top of the fog (determined as RH = 98%, the same as 3 and 4), 3 is the base of the cloud, and 4 is the top of the cloud. Symbols: temperature (T; K), equivalent potential temperature (θe; K), mixing ratio (q; kg kg−1), wind direction (Wd; °), wind speed (Ws; m s−1), relative humidity (RH; %), and height (AGL; m).

Fig. 7.

(a)–(f) Boundary layer structure for the cold advection fog on 28 Feb–1 Mar. Horizontal lines: 1 is the height of the thermal turbulence interface (only exists in warm advection fog, determined as ), 2 is the top of the fog (determined as RH = 98%, the same as 3 and 4), 3 is the base of the cloud, and 4 is the top of the cloud. Symbols: temperature (T; K), equivalent potential temperature (θe; K), mixing ratio (q; kg kg−1), wind direction (Wd; °), wind speed (Ws; m s−1), relative humidity (RH; %), and height (AGL; m).

As found by Petterssen (1938), we find that the critical fog-maintaining factor is the longwave radiation cooling at the fog top. But unlike previously studied cases elsewhere, we find more complicated features in the boundary layer structure. For example, here the fog top can be higher than the base of the inversion (Figs. 7a–e). This phenomenon might be due to entrainment at the fog top. The entrainment here is opposite to a case reported off the coast of California (Rogers and Telford 1986). In the latter case, a lower layer was above the fog top, and the entrainment instability resulted in an air parcel from that layer descending through the cloud layer. In the present case, we instead have a higher layer above the fog layer, and thus entrainment cannot bring down an air parcel with higher value. Because the present case is stable, the fog top can be higher than the base of the inversion (Figs. 7a–e).

Also important is the interaction between overlaying clouds and the fog. The cold advection fog forms under a relative strong inversion (about 5 K) at 2100 LST 28 February (Fig. 7a). But vapor accumulates gradually at the base of the upper inversion, until forming a low cloud at 0500 LST 1 March (Figs. 7b–e). During this period, the base of the upper inversion rises from about 1050 to 1450 m. The temperature of the base of the upper inversion decreases from about 289 to 285 K due to longwave cooling. Meanwhile, the intensity of the lower inversion, where the sea fog forms, weakens from 5 to 2 K. This decrease may be due to the upper cloud, which reduces the net outgoing longwave radiation from the fog top, thus decreasing the cooling effect from longwave radiation at the fog top. Hence, the lower inversion weakens because of the vertical mixing. Then, after the strong inversion vanishes, the fog gradually dissipates, though the upper cloud persists at 0700 LST 1 March (Fig. 7f).

Although the cold-front passage over the coast aids in fog dissipation, a typical sea-fog phase persists for several days before the cold front reaches it, suggesting that other processes may in some cases cause dissipation. For instance, as depicted in Fig. 7, the cold advection fog dissipated when a cloud formed above it. Generally, some other factors, such as a decrease in advection of warm, moist air, a heating of the sea surface and the fog layer by shortwave radiation, and an increase in vertical mixing from an increase in wind speed can weaken the cold advection fog or just lift the fog into the stratus cloud. These stratus clouds, which usually exist during the daytime, will turn into fog again when favorable conditions return.

b. Warm advection fog life cycle

For warm advection fog, the conditions favorable for formation over the sea are 1) a large difference between the SAT and the SST (SAT − SST), 2) moderate winds, 3) an initially high relative humidity, and 4) an initially stable stratification (Petterssen 1939). Prior to the formation of the fog, strong, warm, and moist advection heats the cooler sea surface, thus forming a stably stratified layer with a moderate wind speed.

The selected warm advection fog case forms at 2240 LST 20 March and dissipates at 0050 LST 22 March. For the meteorological and turbulent characteristics of the warm advection fog (Fig. 6, right side), SAT exceeds SST before the fog forms and becomes more evident after it forms (Fig. 6a). The wind speed is generally higher in the warm advection fog case than in the cold advection fog case (Fig. 6b). When the SAT exceeds the SST, the DLR exceeds the ULR (Fig. 6c). The shortwave radiation is low because of the thick low cloud and fog layers (Fig. 6d). Compared to the cold advection fog case, the vertical velocity is larger and always negative. Thus, this fog case shows the continuous transport of warm saturated air from the upper levels to the cold sea surface (Fig. 6e). The momentum flux is also small (Fig. 6f). The sensible heat flux is primarily negative, mainly ranging from 0 to −20 W m−2 because SAT − SST > 0 (Fig. 6g). The latent heat flux is also primarily negative, ranging from 0 to −30 W m−2 (Fig. 6h).

For warm advection fog, the critical maintaining factor is the strong advection of warm moist air (Taylor 1917). Another evident difference with the cold advection fog case is the boundary layer structure. Unlike previously warm advection fog cases that have been observed elsewhere, a shallow stratus layer exists before the fog forms at 1500 LST 20 March (Fig. 8a). At 0000 LST 21 March, the warm advection fog forms (Fig. 8b). At this time, advection of warm, moist air maintains the surface-based inversion, making the SAT exceed the SST. At 0200 LST 21 March (Fig. 8c), the thermal turbulence effect produces a turbulent mixing layer above the thermal turbulence interface (TTI, defined as the point where ). The top of the sea fog rises to a high altitude of about 1050 m under the dominant thermal turbulence effect. At 0400 LST 21 March (Fig. 8d), the advection weakens, thereafter bringing in an insufficient amount of water vapor to maintain the whole fog layer. As a result, dry air enters the sea-fog layer via entrainment, causing the fog to split into two layers: the upper is stratus and the lower is the remaining fog. Then from 0600 to 0800 LST 21 March, because of thermal turbulence above the TTI, the two layers merge into one fog layer again with a lower fog top (due to the weaker advection) (Figs. 8e,f). At 1330 LST that day, the fog layer rises and turns into a stratus layer at MMSEB due to the increase in near-surface temperature (Fig. 8g). At 1900 LST 21 March, advection of warm, moist air again increases, causing the sea fog to reform at MMSEB (Fig. 8h). Such evolution of warm advection fog was also reported in Huang et al. (2011).

Fig. 8.

(a)–(h) As in Fig. 7, but for the warm advection fog on 20–21 Mar.

Fig. 8.

(a)–(h) As in Fig. 7, but for the warm advection fog on 20–21 Mar.

The dissipation of the warm advection case is also due to a cold front passage over the coast. However, other processes may dissipate the warm advection fog. The primary one is a decrease in the advection of warm, moist air. The other processes that weaken the warm advection fog are similar to those for cold advection fog.

5. Analysis of the diabatic heating exchanges

To identify the dominant formation contributor for each fog event, we analyze the diabatic heating rate by phase change and radiation at the 1000-hPa level over both sites. Working from the first law of thermodynamics, Petterssen (1939) pointed out that the cooling effect of fog can be analyzed using

 
formula

where is temperature, is the nonadiabatic heating, is the atmospheric pressure, is the specific heat at constant pressure, and is the gas constant. Consider the right side of Eq. (2). The first term depicts the nonadiabatic heating or cooling and the second term is the adiabatic influence resulting from the pressure variation. According to the analysis of Petterssen, the cooling effect of the second term need not be considered unless there is adiabatic upward motion. As this is not the case here (over the sea), we consider only the nonadiabatic cooling effect.

The nonadiabatic heating rate equals the heating by phase change () and radiation ():

 
formula

The first term equals

 
formula

where is the latent heat of vaporization and is the saturation mixing ratio. Writing out the full derivative, we have

 
formula

where is the mixing ratio. The radiative heating is derived by considering the potential temperature. The potential temperature form of the thermodynamic equation is

 
formula

By using Eqs. (3), (4), and (6), we obtain

 
formula

which is rewritten as

 
formula

Then we obtain

 
formula

So we can calculate the heating by radiation using

 
formula

We calculate the heating by phase change and radiation using Eqs. (5) and (10). The local changes [ or ] are calculated using GPS sonde data at the 1000-hPa level, the horizontal advection term { or } from the NCEP global analyses datasets (1° × 1° grid-averaged values at 1000 hPa), and the vertical advection term { or } from the observed vertical velocity and GPS sonde data. In this way, we obtain each component of the diabatic heating rate at 1000 hPa.

We calculated this heating rate at three times (2000 LST 28 February, 0200 LST 1 March, and 0800 LST 1 March) during the cold advection fog and four times (0200 LST 21 March, 0800 LST 21 March, 1400 LST 21 March, and 2000 LST 21 March) during the warm advection fog (Table 5). The vertical advection term is usually larger than the other two terms in cold advection fog. Meanwhile, the local change term also affects the evolution of the cold advection fog. This characteristic also demonstrates that the diurnally modulated process affects the cold advection fog [as it similarly does to the Haar; Findlater et al. (1989)]. Generally speaking, the total diabatic heating cools the moist air during the formation and development stage, but warms the air in the dissipation stage of the cold advection fog.

Table 5.

Diabatic heating rate at the 1000-hPa level in the sea fog (W kg−1).

Diabatic heating rate at the 1000-hPa level in the sea fog (W kg−1).
Diabatic heating rate at the 1000-hPa level in the sea fog (W kg−1).

In contrast, for warm advection fog, the vertical advection term is much greater than the other two terms. The vertical advection occurs with a quick increase of near the sea surface at 1400 and 2000 LST 21 March, and although the latent heating warms the level by 8.76 to 13.36 W kg−1, the radiative cooling is stronger, with values of −14.66 to −21.02 W kg−1. The total diabatic heating of the fog can reach values of −5.90 to −7.66 W kg−1, which are close to the values of cloud diabatic forcing in southern Asia for January 1986–88 found by Siegmund (1993). Thus, for warm advection fog, the atmosphere supplied a large amount of heat to the sea.

In particular, the observed vertical velocity is about an order of magnitude larger than that in the NCEP datasets. So, if one uses the vertical velocity in the NCEP datasets, the vertical advection term should be as small as the horizontal advection term. However, we consider the vertical advection calculated from the observations as ground truth. Moreover, the observed vertical velocity has a magnitude as low as 10−2 m s−1 before the warm advection fog forms, increasing to 10−1 m s−1 after the fog forms. Thus, the vertical advection term should be reliable in this case (similar magnitude to other sea-fog cases). In addition, two major uncertainties may exist: one, from applying a single value (from a point observation) to the entire region, and two, from neglecting possible adiabatic heating/cooling in the cold advection fog case.

6. Discussion and conclusions

We used GPS sondes and flux data from an offshore integrated observation platform to analyze the boundary layer structure near the sea surface for a case of cold advection fog and a case of warm advection fog off the coast of southern China.

Table 6 summarizes the meteorological and turbulent characteristics of the two cases. A critical factor is the difference of SAT and SST. For cold advection fog, SAT is less than SST by −0.7 K on average, but for warm advection fog, SAT exceeds SST by 1.9 K on average. Consequently, for cold advection fog, DLR is less than ULR by −8.9 W m−2 on average, and the sensible heat flux is mainly positive (average value of 2.58 W m−2); but for warm advection fog, DLR exceeds ULR by 6.8 W m−2 on average, and the sensible heat flux is mainly negative (average value of −6.98 W m−2). The latent heat flux is mostly negative in warm advection fog with an average value of −6.22 W m−2. In contrast, for cold advection fog, the latent heat flux averages 26.75 W m−2, but the value is uncertain because of high data variability and too much missing data.

Table 6.

Characteristics of two cases of sea fog.

Characteristics of two cases of sea fog.
Characteristics of two cases of sea fog.

Other properties also distinguish these two types of sea fog. For the airspeed, the cold advection fog has the lower horizontal wind speed (average value of 2.4 m s−1) and the more variable vertical velocity (average value of −0.007 m s−1). For the warm advection fog, the horizontal wind speed is moderate (average value of 5.4 m s−1) and the vertical velocity is always negative (average value of −0.108 m s−1). For the shortwave radiation, the warm advection fog case endures a higher value than the cold advection fog case because it has a thicker fog layer. The only similar characteristic of the two cases is their low momentum fluxes.

The diagram in Fig. 9 shows the evolution of both cases. Before the cold advection fog forms, two types of airflow exist over the sea surface: the upper airflow is weak advection of warm, moist air, and this airflow leads to a continuous stably stratified layer over the sea. The lower airflow is cold, moist air, mainly coming from the east-northeast with low speeds (Fig. 9a). This lower airflow comes from coastal cold waters, causing a decrease in SAT − SST, and because of the decreasing specific humidity in the vertical at this time, a cold advection fog forms in the high relative humidity very near the sea surface. Then, as soon as the fog forms, because of the clear sky, the longwave radiation from fog top becomes the dominating development factor, causing a further decrease in SAT − SST (Fig. 9b). The strong cooling effect of the longwave radiation at the fog top decreases the fog-top temperature. Accompanied by the buoyancy from the warmer air close to the sea surface, the vertical mixing becomes active. Vertical mixing transports heat from the sea surface into the air; hence, the sensible heat flux should be upward for cold advection fog over the sea. The latent heat flux should depend on the vertical specific humidity profile. A specific humidity that decreases with height will result in an upward latent heat flux, whereas an increase with height produces a downward flux (Fig. 9c).

Fig. 9.

Formation mechanism of (a)–(c) cold advection fog and (d)–(f) warm advection fog. Temperature profile (T), warm moist advection (WMA), cold moist advection (CMA), upward longwave radiation (ULR), stratocumulus (Sc), stratus (St), sensible heat flux (H), and latent heat flux (LE). Light shading depicts region in which thermal turbulence dominates, whereas the dark shading shows where mechanical turbulence dominates. Wavy line depicts the sea surface, solid lines mark fog or cloud boundaries; the broken line is the thermal turbulence interface.

Fig. 9.

Formation mechanism of (a)–(c) cold advection fog and (d)–(f) warm advection fog. Temperature profile (T), warm moist advection (WMA), cold moist advection (CMA), upward longwave radiation (ULR), stratocumulus (Sc), stratus (St), sensible heat flux (H), and latent heat flux (LE). Light shading depicts region in which thermal turbulence dominates, whereas the dark shading shows where mechanical turbulence dominates. Wavy line depicts the sea surface, solid lines mark fog or cloud boundaries; the broken line is the thermal turbulence interface.

Before warm advection fog forms on the coast of southern China, stratocumulus and/or stratus form as a result of the strong, warm, and moist advection (Fig. 9d). As this advection strengthens, the cloud layer becomes thicker due to two types of turbulence, one on either side of the TTI. Thermal turbulence dominates above the TTI because in this layer or , whereas mechanical turbulence is prevalent under the TTI for the layer when (Huang et al. 2011). This upper cloud layer is similar to that of the cold advection fog. Vertical thermal mixing causes the cloud top to grow, accompanied by a strong cooling from the upward longwave radiation of the cloud top. The lower layer has a lowering stratus cloud base dominated by mechanical turbulence mixing (Fig. 9e). Although the vertical thermal mixing may hinder sea-fog formation, the cooling effect of the airflow across the isotherms of the SST creates a large surface temperature difference (SAT − SST). This large difference results in a net cooling effect that overcomes the heating from the upper level. This cooling drives the cloud base down, until it reaches the sea surface. Because the upper stratocumulus and the lower fog have merged together into a thick fog, the fog top has a high altitude, sometimes exceeding 1000 m (up to 1660 m over the past 5 years). In addition, because of the increasing vertical temperature and specific humidity profiles, the sensible and latent heat fluxes are downward in the warm advection fog (Fig. 9f).

The main results are as follows:

  1. Two types of boundary layer structure exist in the sea fog in this area: that for cold advection fog (SAT − SST < 0) and that for warm advection fog (SAT − SST > 0). Both structures depend on the large-scale airflows of the synoptic conditions. The cold advection fog formed near the surface under a clear sky, driven mainly by the advection, but then the main physical forcing changed to longwave radiation at the fog top. The warm advection fog formed by the lowering of an existing cloud base and its growth was mainly driven by advection of warm moist air.

  2. The boundary layer structures for these sea-fog events differed from those of previous studies. In the cold advection fog, the fog top exceeded the base of the inversion, and the formation of a cloud above reduced the effect of the outgoing longwave radiation at the fog top, leading to dissipation of the fog. In the warm advection fog, two layers of low cloud merged to form deep fog, and the fog-top altitude could exceed 1000 m due to strong advection of warm moist air and the thermal turbulence mixing that occurs above the TTI.

  3. The cold advection fog case always had a lower wind speed and endured lower shortwave radiation than the warm advection fog. The vertical velocity, small in both types of sea fog, slowly changed from negative to positive in the cold advection fog, but remained negative in the warm advection fog. The momentum fluxes of both cases of sea fog were low, never exceeding 0.04 kg m−1 s−2.

  4. Turbulence near the sea surface differed between the two cases of sea fog. For the cold advection fog, SAT < SST, DLR < ULR, the latent heat flux, averaging 26.75 W m−2, did not have a clear trend. The sensible heat flux was mainly upward, with an average value of 2.58 W m−2; here the turbulence was primarily thermal. For the warm advection fog, SAT > SST, DLR > ULR, the latent heat flux was downward, with average value of −6.22 W m−2, and the sensible heat flux was mostly downward, with average value of −6.98 W m−2; nearest the sea surface, the turbulence was primarily mechanical.

  5. In both fog cases, vertical advection at the low level had a major influence on fog development. For cold advection fog, vertical advection had a greater influence than the local effects. For warm advection fog, vertical advection transferred a significant amount of heat from the atmosphere to the sea.

For warm advection fog, the direction of vertical advection, the sensible and latent heat flux, all demonstrated that eddy diffusion (mechanical turbulence) transports warm moist air to the sea surface, thus heating the sea. In contrast, for cold advection fog, the average values of the sensible and latent heat flux showed the sea surface heating the air, whereas the effects of the vertical advection and the latent heat flux were variable.

The values of the latent heat flux of fog (mostly between 0 and −30 W m−2) in the CBLAST-LOW experiment (Edson et al. 2007) are similar to our warm advection fog results. However, the values of the latent flux of warm advection fog of the East China Sea (Heo et al. 2010) are larger than ours. The difference might come from two reasons: 1) the turbulence characteristics of sea fog differ between the East China Sea and the South China Sea due to their different synoptic backgrounds; for example, different wind directions and differences in SAT − SST. 2) The two studies used different observational instrumentation and methods of computing the fluxes. Furthermore, an enclosed, high-performance CO2/H2O analyzer might help us solve the problem of the latent heat flux.

Low clouds interacted with both types of fog. For the cold advection fog, the presence of an overhead cloud caused the fog to dissipate. For the warm advection fog, an overlaying stratocumulus layer was crucial for the fog to form. The warm advection fog also split into an upper cloud layer and a lower fog layer, and merged to form a fog layer again afterward. These complex variations are due to the constantly changing advection of warm moist air that always prevails in the boundary layer of a tropical area during the fog season. Further insight into the physical processes of these two types of fog will be possible using modeling efforts that are currently planned.

Acknowledgments

We thank three anonymous reviewers whose constructive suggestions and comments greatly improved this presentation. We also benefited from informal reviews by Darko Koračin and John Lewis. Special thanks to the crew of the Marine Meteorological Science Experiment Base at Bohe, for their help in conducting the field program and providing the data. This study was supported jointly by the National Natural Science Foundation of China (Grants 41275025, 41175013, and 40906023), the Guangdong Science and Technology Plan Project (2012A061400012), the Meteorological Sciences Research Project (Grant 2013B06), and the early warning and forecasting technology for marine meteorology of the Weather Bureau of Guangdong Province.

REFERENCES

REFERENCES
Bolton
,
D.
,
1980
:
The computation of equivalent potential temperature
.
Mon. Wea. Rev.
,
108
,
1046
1053
, doi:.
Byers
,
H.
,
1930
:
Summer sea fogs of the central California coast
.
Publ. Geogr.
,
3
,
291
338
.
Draxler
,
R. R.
, and
G. D.
Hess
,
1998
:
An overview of the HYSPLIT_4 modeling system of trajectories, dispersion, and deposition
.
Aust. Meteor. Mag.
,
47
,
295
308
.
Edson
,
J.
, and Coauthors
,
2007
:
The Coupled Boundary Layers and Air–Sea Transfer Experiment in low winds
.
Bull. Amer. Meteor. Soc.
,
88
,
341
356
, doi:.
Emmons
,
G.
, and
R.
Montgomery
,
1947
:
Note on the physics of fog formation
.
J. Meteor.
,
4
, 206, doi:.
Findlater
,
J.
,
W. T.
Roach
, and
B. C.
Mchugh
,
1989
:
The Haar of northeast Scotland
.
Quart. J. Roy. Meteor. Soc.
,
115
,
581
608
, doi:.
Foken
,
T.
,
M.
Göckede
,
M.
Mauder
,
L.
Mahrt
,
B.
Amiro
, and
W.
Munger
,
2004
: Post-field data quality control. Handbook of Micrometeorology: A Guide for Surface Flux Measurement and Analysis, X. Lee et al., Eds., Kluwer Academic Publishers, 181–208.
Gultepe
,
I.
, and Coauthors
,
2007
:
Fog research: A review of past achievements and future perspectives
.
Pure Appl. Geophys.
,
164
,
1121
1159
, doi:.
Heo
,
K. Y.
,
K. J.
Ha
,
L.
Mahrt
, and
J.-S.
Shim
,
2010
:
Comparison of advection and steam fogs: From direct observation over the sea
.
Atmos. Res.
,
98
,
426
437
, doi:.
Huang
,
H. J.
,
2013
: Observational analysis of sea fog and its boundary layer structure on the coast of Southern China. Ph.D. dissertation, Nanjing University, 129 pp.
Huang
,
H. J.
,
H. N.
Liu
,
W. M.
Jiang
,
J.
Huang
, and
W. K.
Mao
,
2011
:
Characteristics of the boundary layer structure of sea fog on the coast of southern China
.
Adv. Atmos. Sci.
,
28
,
1377
1389
, doi:.
Huang
,
J.
, and
F. X.
Zhou
,
2006
:
The cooling and moistening effect on the formation of sea fog in the Huanghai Sea
.
Acta Oceanol. Sin.
,
25
,
49
62
.
Huang
,
J.
, and
P.
Chan
,
2011
:
Progress of marine meteorological observation experiment at Maoming of South China
.
J. Trop. Meteor.
,
17
,
418
429
.
Kim
,
C. K.
, and
S. S.
Yum
,
2012
:
Marine boundary layer structure for the sea fog formation off the west coast of the Korean Peninsula
.
Pure Appl. Geophys.
,
169
,
1121
1135
, doi:.
Klein
,
S. A.
, and
D. L.
Hartmann
,
1993
:
The seasonal cycle of low stratiform clouds
.
J. Climate
,
6
,
1587
1606
, doi:.
Koračin
,
D.
,
J. M.
Lewis
,
W. T.
Thompson
,
C. E.
Dorman
, and
J. A.
Businger
,
2001
:
Transition of stratus into fog along the California coast: Observations and modeling
.
J. Atmos. Sci.
,
58
,
1714
1731
, doi:.
Koračin
,
D.
,
J. A.
Businger
,
C. E.
Dorman
, and
J. M.
Lewis
,
2005
:
Formation, evolution, and dissipation of coastal sea fog
.
Bound.-Layer Meteor.
,
117
,
447
478
, doi:.
Koračin
,
D.
,
C. E.
Dorman
,
J. M.
Lewis
,
J. G.
Hudson
,
E. M.
Wilcox
, and
A.
Torregrosa
,
2014
:
Marine fog: A review
.
Atmos. Res.
,
143
,
142
175
, doi:.
Lamb
,
H.
,
1943
: Haars or North Sea fogs on the coasts of Great Britain. Meteorology Office Publication, M.O. 50424, 34 pp.
Leipper
,
D. F.
,
1948
:
Fog development at San Diego, California
.
J. Mar. Res.
,
7
,
337
346
.
Leipper
,
D. F.
,
1994
:
Fog on the United States West Coast: A review
.
Bull. Amer. Meteor. Soc.
,
75
,
229
240
, doi:.
Lewis
,
J. M.
,
D.
Koračin
, and
K. T.
Redmond
,
2004
:
Sea fog research in the United Kingdom and United States—A historical essay including outlook
.
Bull. Amer. Meteor. Soc.
,
85
,
395
408
, doi:.
National Centers for Environmental Prediction/National Weather Service/NOAA/U.S. Department of Commerce
,
2000
: NCEP FNL Operational Model Global Tropospheric Analyses, continuing from July 1999 (updated daily). Research Data Archive at the National Center for Atmospheric Research, Computational and Information Systems Laboratory. Accessed 12 March 2012. [Available online at http://dx.doi.org/10.5065/D6M043C6.]
Petterssen
,
S.
,
1938
:
On the causes and the forecasting of the California fog
.
Bull. Amer. Meteor. Soc.
,
19
,
49
55
.
Petterssen
,
S.
,
1939
:
Some aspects of formation and dissipation of fog
.
Geofys. Publ.
,
12
,
5
22
.
Pilié
,
R. J.
,
E. J.
Mack
,
C. W.
Rogers
,
U.
Katz
, and
W. C.
Kocmond
,
1979
:
The formation of marine fog and the development of fog-stratus systems along the California coast
.
J. Appl. Meteor.
,
18
,
1275
1286
, doi:.
Platnick
,
S.
,
M. D.
King
,
S. A.
Ackerman
,
W. P.
Menzel
,
B. A.
Baum
,
J. C.
Riedi
, and
R. A.
Frey
,
2003
:
The MODIS cloud products: Algorithms and examples from Terra
.
IEEE Trans. Geosci. Remote Sens.
,
41
,
459
473
, doi:.
Rogers
,
D. P.
, and
J. W.
Telford
,
1986
:
Metastable stratus tops
.
Quart. J. Roy. Meteor. Soc.
,
112
,
481
500
, doi:.
Siebert
,
H.
, and
U.
Teichmann
,
2000
:
Behaviour of an ultrasonic anemometer under cloudy conditions
.
Bound.-Layer Meteor.
,
94
,
165
169
, doi:.
Siegmund
,
P.
,
1993
:
Cloud diabatic forcing of the atmosphere, estimated from simultaneous ECMWF diabatic heating and ISCCP cloud amount observations
.
J. Climate
,
6
,
2419
2433
, doi:.
Tachibana
,
Y.
,
K.
Iwamoto
,
H.
Ogawa
,
M.
Shiohara
,
K.
Takeuchi
, and
M.
Wakatsuchi
,
2008
:
Observational study on atmospheric and oceanic boundary-layer structures accompanying the Okhotsk anticyclone under fog and non-fog conditions
.
J. Meteor. Soc. Japan
,
86
,
753
771
, doi:.
Tanimoto
,
Y.
,
S.-P.
Xie
,
K.
Kai
,
H.
Okajima
,
H.
Tokinaga
,
T.
Murayama
,
M.
Nonaka
, and
H.
Nakamura
,
2009
:
Observations of marine atmospheric boundary layer transitions across the summer Kuroshio Extension
.
J. Climate
,
22
,
1360
1374
, doi:.
Taylor
,
G. I.
,
1917
:
The formation of fog and mist
.
Quart. J. Roy. Meteor. Soc.
,
43
,
241
268
, doi:.
Telford
,
J. W.
, and
S. K.
Chai
,
1993
:
Marine fog and its dissipation over warm water
.
J. Atmos. Sci.
,
50
,
3336
3349
, doi:.
Thiébaux
,
J.
,
E.
Rogers
,
W. Q.
Wang
, and
B.
Katz
,
2003
:
A new high-resolution blended real-time global sea surface temperature analysis
.
Bull. Amer. Meteor. Soc.
,
84
,
645
656
, doi:.
Wang
,
B. H.
,
1985
:
Sea Fog.
China Ocean Press, 330 pp.
Zhang
,
S. P.
,
M.
Li
,
X. G.
Meng
,
G.
Fu
,
Z. P.
Ren
, and
S. H.
Gao
,
2012
:
A comparison study between spring and summer fogs in the Yellow Sea—Observations and mechanisms
.
Pure Appl. Geophys.
,
169
,
1001
1017
, doi:.