1 Introduction
The sea-land breeze (SLB) is a mesoscale local circulation that is prevalent in coastal regions, driven by the differences in thermal properties between land and sea. This phenomenon occurs across a wide range of latitudes, from the equator to polar regions (
Miller et al., 2003). SLB can modify local meteorological factors such as temperature, humidity, and wind fields by influencing the atmospheric boundary layer, thereby affecting local weather and climate conditions in coastal areas (
Shen et al., 2021,
2024;
Lei et al., 2024). Furthermore, SLB plays a crucial role in aerosol generation, ozone formation, and the transport and dispersion of other air pollutants, significantly impacting air quality in coastal-urban regions (
Ryu and Baik, 2013;
Nie et al., 2020;
Caicedo et al., 2021;
Han et al., 2022, 2023). Further understanding of the SLB circulation is essential for promoting sustainable development concerning natural resources and environment management in coastal areas. Additionally, recent research has increasingly focused on SLB (
Shen et al., 2024) due to its association with meteorological disasters caused by the interactions with other weather systems (
Miao and Yang, 2020), its relationship with urbanization (
Yang et al., 2023), and its significant impact on wind power utilization (
Guo et al., 2023).
The effects of SLB on air quality in coastal regions have been extensively investigated (
Chang et al., 2022;
Chen et al., 2023;
Hao et al., 2024;
Meng et al., 2024). SLB affects pollutant dispersion and transport by modifying meteorological conditions, altering plume trajectory, and trapping pollutants within the shallow thermal internal boundary layer (
Miller et al., 2003;
Viana et al., 2014), thereby deteriorating air quality in coastal urban areas (
Banta et al., 2005;
Caicedo et al., 2021). The generation of SLB can create a convergence zone between onshore and offshore winds, leading to calm wind conditions that facilitate pollutant accumulation over coastal areas (
Darby, 2005), thereby directly or indirectly affecting the health of residents living in these areas (
Budhiraja et al., 2019;
Wang et al., 2022).
Boyouk et al. (2011) observed that PM2.5 concentrations increased with the onset of SLB in Dunkerque, France, attributed to the development of the thermal internal boundary layer.
Augustin et al. (2020) found that SLB circulations provide favorable meteorological conditions for elevated concentrations of precursors, promoting secondary aerosol formation. Previous studies have indicated that SLB amplifies ozone pollution over both the East and South China Seas; this amplified ozone pollution can be recirculated back to coastal areas via sea breeze (SB), further exacerbating ozone levels in coastal urban areas such as Hong Kong, Tianjin, the Yangtze River Delta, and the Bohai Gulf (
Bei et al., 2018;
Wang et al., 2018;
Hu et al., 2019;
Meng et al., 2019;
Xu et al., 2021). Observational studies have demonstrated that total gaseous mercury concentrations could increase by up to 24% under land breeze (LB) conditions compared to SB scenarios. Furthermore, SLB circulation has been shown to facilitate the accumulation of total gaseous mercury specifically in coastal Qingdao (
Nie et al., 2020).
The impact of SLB on meteorological fields has also been well investigated, particularly regarding its interaction with the urban heat island (UHI). On one hand, SB can transport cooler air from the sea into urban areas, which tends to lower near-surface temperatures and mitigate the UHI intensity (
Zhou et al., 2020;
Hu et al., 2022;
Guo et al., 2023;
Yang et al., 2023). On the other hand, the UHI effect can create increased temperature differentials between land and sea, leading to intensified SLB circulation characterized by higher wind speed, greater depth, and stronger vertical upward motion (
Ma et al., 2013). The cooling effect induced by SLB in coastal urban areas has been widely documented globally, including Bilbao, Spain (
Acero et al., 2013), north central Texas (
Hu and Xue, 2016), New York (
Han et al., 2022), Hania (
Kolokotsa et al., 2009), Tianjin (
Shang et al., 2024), and Shanghai (
Yang et al., 2023).
The fundamental dynamics, properties, and structures of the SLB—including its formation mechanisms, structural characteristics, and life cycle—have been extensively studied and are well understood through observational analyses, theoretical research, and numerical simulations (
Miller et al., 2003;
Chen et al., 2019;
Zhao et al., 2022). Researchers have utilized a variety of methods, such as ships, meteorological balloons, aircraft, weather towers, and radar (
Kraus et al., 1990;
Banta et al., 1993) to observe SLB. These observations reveal that the SLB circulation exhibits asymmetric, with SB being significantly stronger than LB (
Finkele et al., 1995). Furthermore, advancements in technology, including the establishment of automatic weather station networks (
Prtenjak and Grisogono, 2007), satellite data and cloud image recognition techniques (
Gille et al., 2005), wind profile lidar, and microwave radiometer radar (
Jia et al., 2023), have enhanced our ability to analyze SLB structures and their development in great detail and comprehensiveness. The meteorological characteristics of SLB, such as prevailing seasons, occurrence frequency, onset and cessation times, variations in wind direction and speed, intensity, and inland invasion distances, have been systematically investigated (
Miller et al., 2003). Findings indicate a pronounced diurnal variation in SLB activity, which predominantly occurs during summer and autumn when the SB exhibits strong intensity while the LB remains weak. Additionally, the timing of onset and cessation varies across different regions (
Qiu and Fan, 2013).
China possesses a coastline exceeding 18000 km, placing it among the countries with the longest coastlines in the world. The complex coastal topography leads to diverse spatial distribution patterns of SLB across different regions of China. Given the significant influence of local factors on the development and evolution of SLB, research typically focuses on individual cities or specific regions. Moreover, due to variations in latitude, topography, and coastal characteristics, there is currently no standardized method for identifying SLB (
Papanastasiou and Melas, 2009;
Han et al., 2019).
Adams (1997) classified SB into four types based on wind direction: pure sea breeze (PSB), corkscrew sea breeze (CSB), backdoor sea breeze (BSB), and synoptic sea breeze (SSB). Each type exhibits distinct structural characteristics that have received limited attention in the literature (
Guo et al., 2023).
Shang et al. (2024) emphasized that further investigation into the effects of PSB and SSB on UHI and air quality is particularly warranted.
To date, there have been few studies that provide a standardized analysis and comparison of SLB across extensive coastal areas in China. Analyzing the various types of SLB is essential for the rational development of both the economy and industry, as well as for improving air quality. Consequently, it is crucial to consider different types of SLB to achieve a comprehensive understanding of their significance throughout China. To better comprehend the unified characteristics and distinctions of SLB in China’s coastal areas, this study primarily focuses on five major coastal urban agglomerations: the Bohai Rim region, Jiaozhou Bay, Hangzhou Bay, the west coast of Taiwan Strait and the Pearl River estuary. We aim to analyze the statistical and meteorological characteristics of SLB using two identification methods.
2 Data and methods
2.1 Observation data
The observational data used in this study were obtained from the National Meteorological Observing Stations of the China Meteorological Administration, including four years of hourly 2-m air temperature, 10-m wind speed and direction from January 2015 to December 2018. The data period analyzed in this study extends beyond that of most previous research conducted in coastal areas of China. To ensure data quality and to minimize the impact of topographic changes, we selected six stations for analysis, all situated at elevations below 200 m. These stations include Tanggu, Qingdao, Fengxian, Zhenhai, Xiamen and Zhuhai, representing the Bohai Rim region, Jiaozhou Bay, Hangzhou Bay, the west coast of the Taiwan Strait and the Pearl River estuary respectively. Given the complex topography along the coastline in Hangzhou Bay, we specifically chose one station on each shore: Zhenhai on the southern shore and Fengxian on the northern shore. The locations of these six stations are illustrated in Fig.1. Data quality control followed the methodology outlined by
Guo et al. (2023): if missing data within a single day exceeded seven hours, that day was excluded from diurnal average statistics. Conversely, if the number of valid data collected within a day exceeded 17 h, it was classified as a valid day. Furthermore, sea/land breeze days must occur within these valid days (
Qiu and Fan, 2013). Fortunately, the validity of data at each station exceeded 99.5%, with all recorded days being deemed valid, and the proportion of valid days across every station reached 100%.
The land surface temperature (LST) data were derived from the MODIS MYD11A1 daily land surface temperature product (Version 6) (MODIS/Aqua Land Surface Temperature/Emission Daily L3 Global 1 km SIN Grid) as detailed by
Wan (2014). For each station, the average value within a latitude and longitude range of 0.15° × 0.15°, was selected to correspond to the station’ location. However, due to significant local topographical influences at the Qingdao and Xiamen stations, a more restricted latitude and longitude range of 0.1° × 0.1° was utilized. The sea surface temperature (SST) data used in this study adopted the daily mean real-time global sea surface temperature high-resolution data set from the National Centers for Environmental Prediction (NCEP), which has a spatial resolution of 0.5° × 0.5° . Given the relatively smooth variations in SST, we calculated the average temperature over an area measuring 0.2° × 0.2°, encompassing the region where each respective station is located, to effectively represent its SST data.
2.2 Identification of SLB
In this study, we employed the methodologies proposed by
Adams (1997) and
Shen et al. (2019) to identify SLB. A key characteristic of the occurrence and development of SLB is the shifts in wind direction, which initially appears as instantaneous changes before becoming sustained (
Zhao et al., 2022). Furthermore, synoptic winds also play a crucial role in the identification of SLB.
In this study, we compared two definitions of SLB: the pure SLB, which excludes the synoptic wind field (hereafter referred to as P-standard SLB), and the true SLB, which incorporates the synoptic wind field (hereafter referred to as T-standard SLB). The observed hourly wind vector is denoted as UO, while the synoptic wind field, representing large-scale circulation, is defined as US. In contrast, SLB refers to small-scale local winds, represented by UL. Consequently, UO and US can be expressed as follows:
where
H represents the 24-h period in a day. In this method, the hourly observed wind, which comprises both large-scale and small-scale local winds, is subtracted from the daily large-scale wind to isolate the small-scale SLB circulation from the larger wind field (
Shen et al., 2019). Consequently,
UL represents the P-standard SLB, while
UO denotes the T-standard SLB. In terms of SLB circulation, SB typically persist between 12:00 and 20:00 LST, whereas LB is generally observed from 00:00 to 08:00 LST. For wind direction, it is classified as a SB if it flows inland from the sea; conversely, it is categorized as a LB if it flows offshore toward the sea within a range of 45° to either side of a line perpendicular to the coastline. A day is designated as an SLB day if it meets all of the following criteria: 1) SB occurs for at least five hours and LB for at least three hours; 2) maximum wind speed does not exceed 8 m/s; and 3) there exists a surface temperature difference between land and sea of at least 0°C during daylight hours on that day. If SB or LB persists for more than two-thirds of the duration of that day, it is classified respectively as a SB day or a LB day. Due to variations in coastal geography and proximity to shorelines, the ranges for SB and LB may differ slightly across various coastal regions.
3 Results
3.1 The frequency of SLB
According to the two defined criteria for SLB days, Tab.1 presents an analysis of the annual mean frequency of SLB days at the six stations from 2015 to 2018. The frequencies of P-standard SLB days are as follows: in the western Bohai Bay (Tanggu), Jiaozhou Bay (Qingdao), the north (Fengxian) and south (Zhenhai) coasts of Hangzhou Bay and the Pearl River estuary (Zhuhai), they are recorded at 21.1%, 20.7%, 24.0%, 28.0%, and 21.8%, respectively. Notably, the frequency of P-standard SLB days on the west coast of the Taiwan Strait (Xiamen) is much higher, reaching 55.4%. This disparity can be attributed to the relatively flat topography surrounding most stations, whereas Xiamen station is characterized by complex topographical features, with mountains located to the north-west and extensive oceanic areas to the south-east (as shown in Fig.1). The local SLB may be in phase and superimposed with the terrain-induced valley breeze. During the day, the upslope wind combines with SB, while at night, the downslope wind interacts with LB. This interaction increases the frequency of SLB occurrences in XiaMen.
Zhang et al. (2014) also noted that valley breezes significantly enhance SLB activity in the coastal areas of Hainan. The frequency of T-standard SLB days is significantly lower than that of P-standard SLB days. The highest frequency of T-standard SLB days reaches 15.7% in Xiamen, compared to only 5.8% in Zhuhai, and 6.9%, 9.8%, 9.9%, 11.1% for the four other stations, respectively. This indicates that the presence of synoptic wind fields can obscure more than half of the local SLB signals. With the exception of Zhuhai, results from other stations show that both P-standard and T-standard SLB days are more prevalent in low-latitude regions than in high-latitude regions. This disparity arises because higher latitude regions are more affected by westerly weather systems, which are unfavorable for SLB generation. In contrast, lower latitude regions experience less impact from these westerly weather systems, thereby facilitating SLB development. Furthermore, lower latitude regions receive a greater amount of total solar radiation compared to higher latitude regions, which further promotes conditions favorable for SLB formation.
The frequency of SLB days demonstrates pronounced seasonal variations. Fig.2 illustrates the number of seasonal mean P-standard and T-standard SLB days for each season from 2015 to 2018. Consistent with prior studies (
Qiu and Fan, 2013), the number of SLB days is generally higher during summer and autumn compared to winter and spring. This pattern can be attributed to increased solar radiation in summer and autumn, leading to more rapid land heating relative to sea surfaces and thus a greater temperature difference between them. Such enhanced sea-land temperature gradient facilitates the formation of SLB. Furthermore, during summer and autumn, the synoptic circulation tends to produce a relatively uniform pressure field with weaker gradient winds, which enhances the observability of SLB. Conversely, during winter, intensified cold air activity and stronger synoptic winds may obscure the signals indicative of SLB events.
In terms of P-standard SLB, the number of SLB days in TangGu during summer is the highest among all seasons, totaling 27 days. In contrast, autumn records the highest number of SLB days along the coast of the Yellow Sea and the East China Sea, with 24, 26, and 30 days at Qingdao, Fengxian, and Zhenhai stations, respectively. This phenomenon can be attributed to a higher frequency of typhoons in these regions during summer. The intense circulation associated with typhoons can disrupt local-scale SLB circulation, thereby reducing the overall number of SLB days (
Wang et al., 2016). At Xiamen station, the number of SLB days peaks in summer at 55 days and reaches its lowest point in winter at 42 days. Conversely, Zhuhai station exhibits relatively consistent numbers across all seasons, ranging from 17 to 18 days.
The number of T-standard SLB days is significantly lower than that of P-standard SLB days, and the seasonal variation in T-standard SLB days exhibits a more pronounced correlation with prevailing wind patterns. This phenomenon can be attributed to the fact that large-scale synoptic wind fields tend to obscure local-scale SLB signals. Furthermore, seasonal differences at higher latitudes are more pronounced compared to those at lower latitudes. This disparity arises from the relatively minor seasonal variations in solar radiation observed at low latitudes, leading to smaller seasonal differences in SLB within this region. Notably, the highest frequency of T-standard SLB days at Zhuhai station occurs during winter. This trend is closely associated with the large-scale synoptic circulation prevalent in this region. Under the influence of a weak ridge resulting from the southward extension of the mainland cold high, winter weather conditions over the Pearl River estuary tend to be clear, accompanied by a weak synoptic wind field. Such meteorological conditions are particularly conducive to the formation of SLB (
Zhang et al., 1999).
3.2 Onset, cessation and duration times of SLB
Due to the differences in thermal properties between land and sea, LB typically initiates around midnight and persists until sunrise, while SB manifests during the daytime. The transition between LB and SB consistently occurs around noon and midnight. In a SLB day, if SB starts at time A and continues for three hours or more (T1), then time A (A + T1) is defined as the onset (cessation) time of SB. If LB starts at time B and continues for two hours or more (T2), concluding before 14:00 LST, then time B (B + T2) is defined as the onset (cessation) time of LB.
3.2.1 Spatial variations of the onset, cessation and duration times of SLB
Fig.3 and Fig.4 show the frequency distributions of the onset and cessation times for P- and T-standard SB and LB, respectively. A notable feature of SLB is its distinct diurnal cycle. As shown in Fig.3 and Fig.4, the onset and cessation times for both SB and LB span a broad range of hours throughout the observed period. Specifically, SB predominantly (≥ 10%) initiates between 11:00−15:00 LST and ceases between 16:00−21:00 LST, with a duration of approximately 5 to 7 h. In contrast, LB primarily (≥ 10%) initiates between 01:00−05:00 LST and ceases between 07:00−10:00 LST. Notably, as latitude decreases, there is a gradual delay in the onset time of SB. The distribution patterns for both onset and cessation times of SB exhibit a pronounced hump-shaped curve; moreover, the peaks associated with these times are more obvious compared to those observed for LB. This observation aligns with the asymmetric characteristics inherent in SLB circulation.
The frequency distributions of the onset and cessation times for sea/land breezes differ between P-standard and T-standard criteria. For P-standard SB in Tanggu, the most probable onset time occurs between 14:00 and 15:00 LST, while the most frequent cessation time is observed between 18:00 and 19:00 LST. In contrast, for T-standard SB, the most likely onset time is at 11:00 LST, with the most common cessation times occurring at 18:00 and 20:00 LST. In Qingdao, the anticipated onset time for T-standard SB is at 15:00 LST (Fig.4(c)), which is one hour later than that of P-standard SB. Additionally, the predominant cessation time for T-standard SB of this site is at 18:00 LST, two hours earlier than that of P-standard SB. On the north coast of Hangzhou Bay (Fengxian), the peak frequency of P-standard SB occurs at 13:00 LST, while that of T-standard SB is observed at 10:00 LST. Conversely, both P-standard and T-standard SB in Fengxian predominantly cease by 18:00 LST. The frequency distributions for the onset and cessation times of P- and T-standard SB in the other three regions exhibit similar patterns. At Zhenhai station, the most probable onset and cessation times for P- and T-standard SB are recorded at 12:00 and 17:00 LST, respectively. In Xiamen, P- and T-standard SB primarily occur between 13:00 and 14:00 LST, and conclude around 20:00 LST. In Zhuhai, P-standard SB typically initiates at 14:00 LST and concludes at 17:00 LST, while T-standard SB starts at 12:00 or 15:00 LST and ends at 18:00 LST.
The distributions of the onset and cessation times for P- and T-standard LB don not exhibit a clear bimodal character. This lack of distinct modes can be attributed to the relatively weak nature of LB, which makes it challenging to sustain; consequently, the onset and cessation times tend to be in close proximity to one another. In Tanggu, the onset time for P-standard LB typically occurs between 00:00−07:00 LST, with cessation time between 05:00−12:00 LST. For T-standard LB, the onset time is observed between 00:00−07:00 LST, while cessation time is noted between 01:00−12:00 LST. The onsets of P- and T-standard LB in Qingdao primarily occur between 00:00 and 07:00 LST, while their cessations predominantly take place between 05:00 and 12:00 LST. At Fengxian station, the onsets of P- and T-standard LB are mainly observed from 00:00 to 06:00 LST. The cessation of P-standard LB typically occurs between 07:00 and 09:00 LST, whereas that of T-standard LB is noted between 02:00 and 09:00 LST. In Zhenhai, the onsets of P- and T-standard LB exhibit patterns similar to those at Fengxian; however, their cessations lag by approximately two hours. In Xiamen, the onsets of P- and T-standard LB are primarily recorded from 00:00 to 07:00 LST, with cessations occurring mainly between 03:00 and 12:00 LST. Finally, in Zhuhai, the onsets of P- and T-standard LB primarily occur between 00:00 and 07:00 LST, with P-standard LB cessations predominantly observed between 05:00 and 12:00 LST, and T-standard LB cessations between 01:00 and 08:00 LST.
For a more comprehensive comparison of the differences in onset and cessation times of SLB across various regions, Tab.2 and Tab.3 present the average onset, cessation, and duration times for P- and T-standard SB and LB in different geographical areas. The results indicate that both P- and T-standard SB and LB exhibit later onset and cessation times in low-latitude regions compared to high-latitude ones. Additionally, the corresponding duration times are shorter in low-latitude regions than those observed in high-latitude regions. Fig.5 and Fig.6 illustrate scatter plots depicting the relationship between latitude and onset times of P- and T-standard SB and LB for further analysis. Due to the complex topography and curvature of the coastline in Hangzhou Bay, the onset times of SB and LB at Fengxian and Zhenhai stations do not fully conform to the expected pattern associated with changes in latitude; consequently, these two stations have been excluded from this analysis. For each 10° increase in latitude, it is observed that P-standard SB starts approximately 0.3 h earlier, while LB starts about 0.6 h earlier (Fig.5). The duration of P-standard SB across six stations from north to south is as follows: 6.2, 6.7, 5.7, 5.3, 6.0, and 5.5 h, respectively (Tab.2). With the exception of Fengxian station, the duration of LB gradually decreases from 5.8 h to 5.0 h as latitude decreases.
The differences in mean onset, cessation, and duration times of T-standard SB and LB across the six stations are more pronounced compared to those observed for P-standard SB and LB. Notably, the onset time of T-standard SB at higher latitudes occurs obviously earlier than at lower latitudes. Specifically, for each 10° increase in latitude, the onset time of T-standard SB advances by approximately 0.7 h. The synoptic wind field plays a crucial role in influencing both the onset and cessation times of SB. The duration of T-standard SB at the six stations is as follows: 8.1, 7.9, 7.6, 6.6, 6.4, and 7.4 h; while the corresponding duration for T-standard LB is 4.0, 6.1, 4.5, 3.4, 3.5, and 3.5 h, respectively (Tab.3). The duration of T-standard SB is consistently longer than that of P-standard SB across all stations, whereas the duration of LB is comparatively shorter. The narrow sea expanse and convoluted coastline of Hangzhou Bay limits the ability to sustain significant temperature differences between land and sea over extended periods. Moreover, the duration of T-standard SB in Hangzhou Bay is shorter compared to other regions due to topographic variations. The cessation time of T-standard LB in Qingdao is notably later than that in other regions, and its duration is much longer. This phenomenon is attributed to the unique geographical characteristics of the Shandong Peninsula. During the nighttime LB period, the extensive land area of the Shandong Peninsula enhances the sea-land temperature differential, which persists for a more extended duration.
The large-scale synoptic wind field and the latitudinal variation in solar radiation are both critical factors influencing the timing of SLB onset. Moreover, the Coriolis force significantly impacts SLB circulation. As latitude increases, the influence of the Coriolis force intensifies, leading to a more pronounced rotational effect and greater deflection of horizontal winds. This enhanced deflection results in an earlier onset of LB (
Yan and Anthes, 1987). Consequently, at TangGu station, which is situated at the highest latitude, the T-standard LB commences earliest.
3.2.2 Seasonal variations of the onset, cessation, and duration times of SLB
The onset, cessation, and duration of SLB exhibits distinct seasonal variations. Tab.4 and Tab.5 present the seasonal mean duration times for P- and T-standard SB and LB, respectively. The seasonal differences in duration times for P-standard SB and LB are both approximately 1−2 h across all six stations. Specifically, the duration of SB ranges from 5.0 to 6.8 h across different seasons, while that of LB varies from 4.5 to 6.7 h. In contrast, the seasonal differences in duration times for T-standard SB and LB are more pronounced. For T-standard measurements, the duration of SB ranges from 5.6 to 9.7 h, and that of LB varies from 3.0 to 7.2 h across different seasons. For P-standard SB, most stations experience the longest duration times in spring, including Tanggu, Qingdao, Fengxian, and Zhuhai. In contrast, for P-standard LB, the longest duration times occur during summer at locations such as Fengxian, Zhenhai, and Xiamen. For both T-standard SB and LB, most stations experience their maximum duration times in autumn. For instance, in Qingdao, the longest duration times recorded for T-standard SB and LB are 8.5 and 7.2 h in autumn, respectively. The seasonal variation in T-standard SLB duration is primarily affected by large-scale synoptic circulation patterns. During summer and autumn, south-east and south-west winds dominate along the east coast of China. The prevailing synoptic wind direction aligns relatively consistently with SB during daylight hours, thereby enhancing both the strength and longevity of SB. Moreover, the seasonal disparity in thermal energy between land and sea, driven by variations in solar radiation, plays a significant role in influencing the seasonal variations in SLB duration.
3.3 Wind speed characteristics of SLB
To investigate the wind speed characteristics associated with SLB, this section analyzes the intensity of both SB and LB. Wind speed is categorized into two components relative to the coastline: the perpendicular (U-component) and the parallel (V-component). A positive U-component value, indicating flow from land to sea, is classified as offshore wind, while a negative U-component value, representing flow from sea to land, is classified as onshore wind. The V-component represents alongshore wind, with northward direction designated as positive. The maximum onshore wind speed is defined as SB intensity, whereas the maximum offshore wind speed is defined as LB intensity (
Arritt, 1993).
3.3.1 Diurnal variation of the wind speed of SLB
Due to the diurnal transformation of SB and LB, the SLB circulation exhibits pronounced diurnal variations. Hodographs representing surface wind vectors can serve as an effective means for gauging the amplitude of diurnal processes associated with SLB (
Davis et al., 2019). Conceptually, the SLB circulation involves the rotation of local winds driven by frictional forces and the Coriolis effect, resulting in a hodograph that traces an elliptical path in a cyclonic manner within the Northern Hemisphere (
Haurwitz, 1947). The length of the major axis of this elliptical hodograph serves as an indicator of the intensity of wind speed associated with SLB.
Hodographs derived from the hourly averages of the P-standard SLB wind vectors are illustrated in Fig.7. With the exception of the Fengxian station, the trajectories of wind vectors at other stations exhibit a consistent elliptical pattern. At higher latitudes, there is a more pronounced rotational effect due to the Coriolis force, which results in a more eccentric elliptical path (
Gille et al., 2005). For instance, the elliptical eccentricity associated with Qingdao is significantly greater than that of Zhenhai. Due to the smaller thermal difference between land and sea during nighttime in the lower atmosphere compared to daytime, pronounced asymmetries exist in both the intensity and duration of the SB and LB circulations. The interaction between large-scale synoptic circulation and local SLB further exacerbates this asymmetry. In P-standard SLB, the maximum onshore wind typically occurs between 15:00 and 16:00 LST, with SB intensity ranging from approximately 1.5 to 2.5 m/s. Conversely, the maximum offshore wind occurs between 03:00 and 05:00 LST, with LB intensity averaging approximately 1.0 m/s. The wind direction at the Tanggu, Zhenhai, and Zhuhai stations exhibits a clockwise rotation over time, clearly indicating the influence of the Coriolis force. In contrast, the wind direction at Qingdao and Xiamen changes more smoothly, demonstrating anti-clockwise hodograph rotations. This phenomenon can be attributed to several factors, including the influence of large-scale synoptic circulation, coastal geometry, interactions with the Coriolis force, and nonlinear heating effects (
Davis et al., 2019). The variation in alongshore wind also influences the intensity of SLB (
Miller et al., 2003). The V-component wind speed in Qingdao is notably stronger than that at other stations, indicating a pronounced presence of alongshore winds in this region. This phenomenon triggers seawater upwelling, which subsequently increases the temperature difference between land and sea, thereby enhancing SLB circulation.
The hodographs of the T-standard SLB wind vectors are shown in Fig.8. The fundamental characteristics of onshore and offshore winds for the T-standard SLB exhibit similarities to those observed in P-standard SLB. Specifically, the maximum onshore wind occurs between 15:00 and 16:00 LST, with a corresponding SB intensity of approximately 2.5 to 3.0 m/s. Conversely, the maximum offshore wind occurs between 03:00 and 05:00 LST, with an LB intensity ranging from approximately 0.5 to 1.0 m/s. Compared to the P-standard SB and LB, the T-standard SB exhibits a stronger intensity, whereas the LB intensity is comparatively weaker. This observation suggests that large-scale synoptic circulation amplifies the differences between T- and P-standard SB and LB. Furthermore, at the TangGu and ZhuHai stations, the length of the elliptical hodograph’s short axis for T-standard SLB exceeds that of P-standard SLB, indicating a more substantial alongshore wind in these regions. This phenomenon arises from the interplay between synoptic circulation and local SLB dynamics. The intricate variations in wind direction observed at the Fengxian station illustrate the impact of a weak synoptic weather system on the local SLB, leading to oscillations of alongshore winds between positive and negative values. During the day, LB rotates clockwise toward SB. Conversely, at night, influenced by northerly winds, SB turns counterclockwise toward LB.
3.3.2 Seasonal variation of the wind diurnal cycle of SLB
The diurnal variation of the SLB also exhibits seasonal variations, which can be illustrated by fluctuations in onshore and offshore winds. Fig.9 and Fig.10 present the diurnal variations of the U-component wind for both P-standard and T-standard SLB across different seasons. Overall, the period from 12:00 to 20:00 LST corresponds to onshore winds (i.e., SB), while the period from 00:00 to 08:00 LST is characterized by offshore winds (i.e., LB) for both P- and T-standard SLBs. For most stations, onshore wind intensity peaks during summer and reaches its lowest levels in winter. Additionally, peak times for onshore winds occur earlier in summer, with a shorter transition duration from onshore to offshore winds compared to other seasons.
The seasonal variation of the P-standard SB at Tanggu and Qingdao stations is more pronounced compared to that observed at other stations. In spring, the SB intensity at TangGu can reach 2.2 m/s, whereas during other seasons it remains approximately between 1.5 and 1.6 m/s. At Qingdao station, the SB intensity peaks at 2.8 m/s during both spring and summer, while in winter it averages approximately 2.0 m/s. Conversely, the SB intensity at the other four stations exhibits only slight variations across different seasons, ranging from 1.5 to 2.0 m/s at Fengxian station, 1.8 to 2.2 m/s at Zhenhai station, 2.0 to 2.5 m/s at Xiamen station, and 1.4 to 1.8 m/s at Zhuhai station, respectively. Notably, the P-standard LB intensity is consistently weaker than that of the SB. The seasonal variation of the P-standard LB intensity at Tanggu and Qingdao stations is more pronounced than that observed at other stations. At Tanggu, the LB intensity can reach 1.8 m/s in spring, while it averages approximately 1.2 m/s in winter. In contrast, at Qingdao station, the LB intensity peaks at 1.9 m/s in winter and is approximately 1.1 m/s in autumn. The seasonal variations of LB intensity at other stations are relatively small, ranging from 1.0 to 1.3 m/s at Fengxian station, 1.0 to 1.2 m/s at Zhenhai station, 1.1 to 1.5 m/s at Xiamen station, and 1.1 to 1.6 m/s at Zhuhai station, respectively. Overall, the seasonal variation of the P-standard LB intensity is less pronounced compared to that of the SB intensity.
In terms of the T-standard SLB, the SB intensity at Tanggu, Qingdao, and Fengxian stations exhibits its strongest (weakest) values in spring (winter), with respective records of 3.6 m/s (2.5 m/s), 3.1 m/s (2.6 m/s) and 3.2 m/s (2.5 m/s), respectively. At Zhenhai station, the SB intensity is highest in spring (2.9 m/s) and lowest in autumn (2.5 m/s). For Xianmen station, the SB intensity peaks during summer (3.6 m/s) and reaches its minimum in winter (2.4 m/s). Zhuhai station demonstrates the strongest SB intensity in autumn (3.9 m/s) and the weakest during winter (2.5 m/s). Concerning LB intensity, TangGu records its highest value in spring (1.3 m/s) and its lowest in summer (1.1 m/s); whereas Qingdao shows peak LB intensity during winter (2.0 m/s) and a minimum during spring (1.8 m/s). The seasonal variation in LB intensity is minimal at Fengxian and Zhenhai stations, where the LB intensity across all four seasons ranges from 0.3 to 0.6 m/s and 0.7 to 0.9 m/s, respectively. In contrast, Xiamen station experiences the highest LB intensity in autumn (1.9 m/s) and the lowest in summer (0.9 m/s). Similarly, Zhuhai exhibits its strongest LB intensity during autumn (1.0 m/s) and its weakest in spring (0.5 m/s). Furthermore, the T-standard SB intensity surpasses that of the P-standard, while the T-standard LB intensity is weaker than that of the P-standard. The influence of synoptic circulation renders the seasonal differences in T-standard SB and LB intensities more pronounced compared to those of the P-standard.
In spring, the prevailing wind direction in the coastal regions of China is southerly. The superposition of the sea-land temperature gradient generated by large-scale circulation, along with the local sea-land thermal differences, enhances the strength of SB. In contrast, during winter, the thermal differences driven by the north-west monsoon are partially offset by local sea-land temperature gradients. Consequently, within the SLB circulation framework, the LB intensity is stronger in spring, while the SB intensity is weaker in winter.
3.4 Development trend of the SLB
We analyzed the development trends of the P- and T-standard SLB by examining SLB days from 2015 to 2018. Linear regression was conducted separately for daytime and nighttime wind speeds associated with SLB, with the slope of the linear fit defined as the development trend of SLB (unit: m/s·hr-1). Subsequently, daily mean SST data from NCEP and daytime and nighttime land surface temperature data from the MYD11A1 product of MODIS were utilized to calculate the temperature differences between land and sea during both daytime and nighttime. Finally, we performed a linear fitting analysis between the development trend of SLB and the sea-land temperature difference to assess how this temperature disparity influences the evolution of SLB.
The temperature difference between land and sea is a primary factor influencing the formation of SLB. In the P-standard SLB, most data points exhibit strong positive correlations. A larger sea-land temperature differential leads to a more pronounced development trend of SLB. During the day, LB transitions to SB, resulting in a negative development trend for SLB, as the sea surface temperature remains lower than that of the land surface. Conversely, at night, SB transitions back to LB, leading to a positive development trend for SLB when the sea surface temperature exceeds that of the land surface (Fig.11). The fitting slopes at the six stations are 0.021, 0.017, 0.026, 0.022, 0.017, and 0.017, respectively, which are close to 0.02, indicating that there is no significant difference in slope changes across different latitudes. For every 5°C increase in the sea-land temperature difference, the development trend of SLB increases by approximately 0.1 m/s/hr. The distribution of SB points in Tanggu exhibits the greatest dispersion among all stations. In contrast, the sea-land temperature difference corresponding to SB points in Qingdao is larger and demonstrates greater overall dispersion compared to LB points. At other stations, both SB and LB point distributions are relatively concentrated, with minimal differences observed between them.
Fig.12 shows that the distribution of T-standard SB at TangGu and Qingdao stations is relatively dispersed, while the LB distribution is more concentrated. The fitting slopes between the sea-land temperature differences and SLB intensity across the six stations are 0.029, 0.014, 0.022, 0.021, 0.019, and 0.014, respectively. The slope at Tanggu (Xiamen) is obviously (slightly) greater than that of the P-standard SLB. In contrast, the fitting slopes for the other four stations are smaller than that of the P-standard SLB, indicating that the development trend of T-standard SLB is weaker compared to that of the P-standard SLB under the same sea-land temperature differences. This discrepancy can be attributed to the fact that T-standard SLB incorporates synoptic circulation, meaning the sea-land temperature difference is influenced not only by the local wind field but also by large-scale synoptic wind patterns. With the exception of Qingdao station, it is observed that the fitting slope decreases as latitude decreases at the other stations.
4 Conclusions
In this study, we utilize observational data and reanalysis data from 2015 to 2018 to investigate the SLB characteristics in six coastal urban regions of China. The characteristics of both P-standard SLB (pure sea-land breeze without synoptic circulation) and T-standard SLB (true sea-land breeze including synoptic circulation) are statistically analyzed, and the differences among the six regions are compared.
The occurrence of strong sea-breeze days is more frequent in summer, while strong land-breeze days predominantly occur in winter. The frequency of SLB increases as latitude decreases. Notably, synoptic circulation can obscure over half of the local SLB signals. Across all stations, the frequency of P-standard SLB days exceeds 20%, whereas the frequency of T-standard SLB days is less than 10% at most stations. Seasonal variations in P-standard SLB are primarily driven by local thermal differences, while T-standard SLB is mainly affected by prevailing wind patterns.
As latitude increases, the frequency of SLB also increases, and the onset time of SB gradually occurs earlier. For every 10° increase in latitude, the onset time for P-standard (T-standard) SB is approximately 0.3 (0.7) hours earlier. The duration of T-standard SB is longer than that of P-standard SB, whereas LB has a shorter duration. There are no obvious seasonal variations in the duration of P-standard SB and LB; however, the duration of T-standard SB is longer during summer, while the duration of LB is longer in winter and autumn.
The SLB circulation represents an asymmetric system. Both the P-standard and T-standard SB exhibit greater intensity and duration compared to LB, with notable seasonal variations observed in the intensities of both SB and LB. Specifically, SB intensity tends to be stronger during spring and summer, while it is weaker in autumn and winter. Conversely, LB intensity is generally more pronounced in winter compared to other seasons.
The sea-land temperature difference is the primary factor driving the SLB circulation. For every 5°C increase in this temperature difference, the development trend of SLB increases by approximately 0.1 m/s·hr-1. Notably, the fitting development trend of LB at night exceeds that of SB during the day. Furthermore, under the same sea-land temperature differences, the development trend of T-standard SLB is weaker than that of P-standard SLB. The fitting slope for T-standard SLB is generally smaller than that for P-standard SLB, because the sea-land temperature difference is influenced not only by local wind fields but also by large-scale synoptic wind patterns.
Previous studies on SLB in China have primarily focused on limited coastal regions, and there has been a lack of standardized criteria for defining SLB. This study employs two unified standards to analyze the characteristics of SLB across six coastal urban areas in China, thereby facilitating a systematic examination and deeper understanding of SLB characteristics in these regions. However, due to limitations in observational data, this study acknowledges several shortcomings that warrant further investigation in future research.
PDF
(5708KB)
