1 Introduction

From a global perspective, open-ocean and coastal winds are blowing harder in the Southern Hemisphere (Young and Ribal 2019) and in upwelling systems (Sydeman et al. 2014), whereas terrestrial stilling (i.e., decline in surface winds until the 2010s; Roderick et al. 2007) is expected to continue throughout the present century across mid-to-high latitudes of the Northern Hemisphere (Deng et al. 2021; Zha et al. 2021), despite of the reversal observed in the last decade (Zeng et al. 2019). Among the possible causes discussed for the atmospheric stilling and reversal phenomena (Wu et al. 2018), changes in land-use and surface roughness (Vautard et al. 2010) and the internal decadal ocean–atmosphere oscillations (Zeng et al. 2019; Deng et al. 2021) are the most likely drivers. However, besides the interhemispheric asymmetry of surface wind changes, these are changing differently in sign and magnitude at seasonal-scale (Azorin-Molina et al. 2021), latitudinal-scale (Zhang et al. 2020; Zha et al. 2021), and depending on the locations, i.e., differences between coastal, inland and mountain stations (Minola et al. 2016; Azorin-Molina et al. 2018a). For example, increased trends in summer-coastal winds have been observed around the world (Azorin-Molina et al. 2014a, 2016, 2018a; Kim and Paik 2015; Zhang et al. 2020), and most of these authors point out that a reinforcement in local circulations (i.e., sea-breeze, hereafter SB) could be driving these trends, as SB represent the dominant local wind in most coastal regions of the world (Simpson 1996). Nevertheless, the mechanisms behind seasonal and local-based wind speed trends remain largely unknown as the stilling vs. reversal research has not yet quantified changes according to the types of winds; e.g., westerlies, trade winds, local winds, etc. (Azorin-Molina et al. 2018b). This study covers the research niche of assessing changes and multidecadal variability in SB.

Despite being a widely studied phenomenon through numerical simulations (Zhong and Tackle 1993; Crosman and Horel 2010; Robinson et al. 2013; Birch et al. 2015; Drobinski et al. 2018), cases-study (Atkins and Wakimoto 1997; Berri and Dezzutti 2020; Cana et al. 2020), and short-term climatology (Furberg et al. 2002; Papanastasiou and Melas 2009; Azorin-Molina et al. 2011a; Liang and Wang 2017; Cafaro et al. 2019; El-Geziry et al. 2021), the research on long-term SB changes has been limited because of the short length, low spatial and temporal resolution and unreliability of observations over the land and ocean surfaces. To date, few studies have addressed more than a decade of SB climate (Redaño et al. 1991; Laird et al. 2001; Masselink and Pattiaratchi 2001; Zhu and Atkinson 2004; Alomar and Grimalt 2008; Misra et al. 2011; Steele et al. 2015; Perez and Silva Dias 2017; Khan et al. 2018; Guedje et al. 2019; Coulibaly et al. 2019; Hwang et al. 2020; Grau et al. 2021; Shen et al. 2021a, 2022), and only six recent studies have focused on analyzing their long-term trends and multidecadal variability (Shen et al. 2019, 2021a, b, 2022; Pazandeh-Masouleh et al. 2019; Shen and Zhao 2020). In fact, there is no study examining how changes in SB influence the seasonal trends of winds, but only one has quantified the contribution of SB to wind variability (Shen et al. 2022). The recent state-of-the-art of the SB trend estimates over the globe is not conclusive, with strong suggestions to the key role of local mechanisms. For example, the dynamic factors (i.e., increasing surface roughness and urbanization expansion) have primarily driven the slowdown of summer SB speed (-0.033 m s−1 decade−1) in Shanghai (China) for 1994–2014 (Shen et al. 2019); and also in Colombo (Sri Lanka) for the same period, with a greater wind speed decline after 2010 (Shen et al. 2021b). On the opposite, the increase in air temperature has driven an annual and seasonal strengthening of the maximum SB speed in Adelaide (Australia) for 1955–2007 (Pazandeh-Masouleh et al. 2019), and recent studies evidence ongoing positive wind trends in Adelaide and Perth for 1994–2014 (Shen et al. 2021b). In addition, solar radiation also determined winter variations on SB speed in Los Angeles (USA) for 1994–2014 (Shen et al. 2021a). In the Mediterranean area, where climate conditions favor the development of SB, trends vary locally, e.g., Thyna (Tunez) and Tel Aviv (Israel) present no significant variations for 1994–2014, Rome (Italy) shows positive trends and Barcelona (Spain) negative ones for the same period (Shen et al. 2021b). These discrepancies evidence that SB changes are location-based, due to the different coastline orientation and complex terrain (Qian et al. 2012), land-use changes (Kusaka et al. 2000), large-scale atmospheric circulation (Arritt 1993), air-surface temperature rise (Lebassi-Habtezion et al. 2011) or ocean–atmosphere oscillations (Azorin-Molina and Lopez-Bustins 2008) to mention a few, which can interact each other at once and control the SB features (Miller et al. 2003), thereby challenging the understanding of their changes and variability.

Focusing on the Iberian Peninsula, Azorin-Molina et al. (2014a, 2016) reported positive trends of wind speed and gusts in the warm semester (May–October), which might be associated with a strengthening of local wind circulations; e.g., meanwhile Jerez et at. (2012) pointed out a possible reinforcement of local winds due to the soil moisture depletion and the enhancement of the Iberian thermal low. However, it is still unknown how local SB has reacted to the mix of local and regional feedbacks and a warming climate. Therefore, it is of huge scientific interest deepening in the knowledge of SB long-term trends in the framework of the stilling vs. reversal phenomena. The multidecadal research of SB is necessary, as these winds have a large impact on broad socioeconomic and environmental spheres (Simpson 1996). By bringing moisture to coastal and continental areas from ocean basins (Drobinski et al. 2018; Davis et al. 2019), SB affect evapotranspiration and precipitation processes (Azorin-Molina et al. 2009, 2015) with direct implications over agriculture, hydrology and local weather and climate (Mahrer and Rytwo 1991; Simpson 1996; Zhu et al. 2017; Pausas and Millán 2019). Besides, changes on SB speed can also may alter the efficiency of wind power industry during peak demand periods (Steele et al. 2015). Moreover, these onshore winds greatly affect air quality and human health, by transporting air pollutants from coastal areas toward inland cities (Papanastasiou et al. 2010; Bei et al. 2018). Thus, a better understanding of changes in SB in a warming climate could support stakeholders and policy-makers to adapt to future wind projections.

The aim of this study is to analyze for the first time the long-term trends, multidecadal variability and possible drivers of the near-surface wind speed changes (mean and gusts) on days dominated by SB circulations in Eastern Spain for 1961–2019. The specific objectives of this study are (i) to develop a robust and automatic selection approach for creating the first long-term SB database from reliable and homogenized wind series; (ii) to advance in the study of the seasonal changes in the stilling-reversal phenomena from a wind type (i.e., SB circulations) perspective and for both mean and gust speeds; and (iii) to improve the understanding of the influence of large-scale atmospheric circulation on SB variability and trends in a warming climate.

2 Data and methods

2.1 Study region

We define Eastern Spain as the area comprising the Mediterranean coast of Spain between 36°–41 °N and 4.5 °W–4.5 °E (Fig. 1), characterized by a complex topography: i.e., coastal plains, deltas, river mouths and valleys, flanked by inland mountain ranges such as the Betic, Prebetic and Iberian systems. The study area covers 82,165 km2 of territory and 3017 km of coastline (from Malaga to Barcelona, including the Balearic Islands), and it is located in the Western Mediterranean basin. The climate is typically Mediterranean with mild winters and dry hot summers, modulated by the cooling effect of the SB, which represent the most dominant wind circulation at summertime (Olcina-Cantos and Azorin-Molina 2004). Because stable atmospheric conditions (i.e., high insolation and weak surface pressure gradient) are frequent throughout the year, SB can develop even during the cold semester (November–April), occurring in 2/3 of the days in the year (Azorin-Molina and Martin-Vide 2007). In addition, this local wind circulation can penetrate ~ 150 km inland (Kottmeier et al. 2000) through river valleys (Simpson et al. 1977; Arritt 1993), being responsible for most convection (Azorin-Molina et al. 2009) and summer precipitation (Azorin-Molina et al. 2014b) in mountainous areas inland.

Fig. 1
figure 1

Terrain map of the Eastern Spain with location of the 16 weather stations (triangles), the numbers and limits of each region are detailed in Tables 1 and 2. The inset upper left picture shows the transect of the WeMOI used to detect SB events

Full size image
Table 1 Description of the meteorological stations with homogeneous wind speed measurements over the Eastern Spain (for locations see Fig. 1) for 1961–2019
Full size table
Table 2 SB days identified for each weather station and sub-regional coast for 1961–2019
Full size table

We split the Eastern Spain into five regions according to the orientation of the coastline, which strongly influences SB dominant direction (Azorin-Molina and Lopez-Bustins 2008; see Fig. 1): (i) the Southern coast (coastline oriented from W to E) on the Betic system, with the highest mountain range peak (Mulhacén, 3478.6 m a.s.l); (ii) the South-eastern coast (SW to NE) on the Prebetic system and irregular coastal relief (mostly headlands and cliffs); (iii) the Eastern coast (S to N) with the Iberian mountain range at west, large coastal plains at the coast, and the most important delta of the region: Ebro delta; (iv) the North-eastern coast (SW to NE) on the Catalan coastal range; and (v) the Balearic Islands (W to E), a group of islands with quite complex orography (altitudes reach 1445 m a.s.l in the Serra de Tramuntana mountain range).

2.2 Observed wind speed data

The Spanish Meteorological Agency (AEMET) provided (i) the daily mean wind speed data (in m s−1) averaged from standard 10 m height measurements at 0700, 1300 and 1800 UTC, and (ii) the daily peak wind gusts (observed peak gust in ~ 3 s, from 0000 to 2400 h UTC, in m s−1), for 16 stations along the Eastern coast of Spain (Fig. 1) for 1961–2019. Spanish historical wind observations have been measured with different types of anemometers as detailed in Azorin-Molina et al. (2014a). We selected those existing meteorological stations where the distance to the coastline does not exceed 18 km (given in Table 1), because the inland penetration of SB could be reduced in wintertime or inhibited by offshore synoptic flows (Salvador and Millán 2003; Azorin-Molina and Chen 2009; Azorin-Molina et al. 2011a). Five of the selected stations are sited in urban environments (see Table 1), while the rest are located in well exposed sites at airports, reducing the frequency of potential inhomogeneities due to surrounding changes in urbanization (Azorin-Molina et al. 2014a). This is the longest wind dataset for studying SB changes in Spain.

2.3 Homogenization of wind speed data

Wind speed series were created by applying a robust quality control and homogenization protocol, following the method used by Azorin-Molina et al. (2019). Reconstructions of wind speed series were made by checking for discontinuities, due to the installation of automatic weather stations (after the 1980s) and relocation (see Table 1). All stations were homogenized to avoid disturbances in climate trends due to non-climatic break-points in wind series. Although Azorin-Molina et al. (2014a) suggest the use of some regional climate models (i.e., MM5) as reference series for homogenization due to their reliability to reproduce mesoscale circulations (such as SB), we rather used wind speed data from the National Center for Environmental Prediction-National Center for Atmospheric Research (NCEP-NCAR) reanalysis (https://psl.noaa.gov/data/gridded/data.ncep.reanalysis.html; last accessed July 14, 2022), as it covers the entire study period. Moreover, high correlated (r = 0.4, p < 0.05; up to 25 km) neighboring stations were also used to detect inhomogeneities, according to Azorin-Molina et al. (2014a). We applied the CLIMATOL package (http://www.climatol.eu/, last accessed July 14, 2022), which implements a variation of the Alexandersson’s Standard Normal Homogeneity Test (SNHT; Alexandersson 1986) in all 16 stations on a daily basis, which represent one of the most used tests to detect inhomogeneities in climate series. In summary, the procedure did not detect any clearly outlier value, but corrected 68 and 76 break-points with SNHT > 25 for SB speeds and gusts respectively.

2.4 Identification and validation of sea breeze days

SB selection algorithms consist in detecting past events from well-defined criteria, such as certain thresholds for wind or thermal contrast between land-sea air (ΔT), among many others. However, there is no universal method to accurately detect SB days. Instead, SB identification depends mostly on the criteria used in each study site, physical complexity of the area and data availability (see Table 1 in Azorin-Molina et al. 2011b). In general, there is a global lack of historical observed data for many essential variables to automatically detect SB events from long-term climate series. Most historical observed data presents (i) large gaps; (ii) low quality; and (iii) limited temporal resolution to create with reliability a historical and high-quality SB database. Even when procedures such as quality control, missing data infilling, reconstruction and homogenization can be applied to entire datasets (i.e., essential climate variables for SB detection) on a sub-daily basis (Guijarro 2018; Curci et al. 2021), this implies huge computational cost and effort, being still challenging for the climate data homogenization community. To address the need of improving the processing time and accuracy in SB selection methods for long-term studies, we propose a robust automated method to identify potential SB days using alternative variables (e.g., Laird et al. 2001; Azorin-Molina and Lopez-Bustins 2008), which can be used worldwide at any coastal region with few adjustments.

The automated selection algorithm presented here is based on six alternative filters, classified in (a) synoptic and (b) local-scale conditions (see Fig. 2). Thresholds were chosen based on previous SB literature over the world but specifically over Eastern Spain (e.g., Ramis and Alonso 1988; Azorin-Molina and Lopez-Bustins 2008; Azorin-Molina et al. 2011a; Azorin-Molina et al. 2011b; Cafaro et al. 2019; Arrillaga et al. 2020; to name but a few). The first main group (filters 1 to 3) employs regional teleconnection patterns, synoptic weather types and geostrophic winds as alternative criteria to reject days with large-scale synoptic disturbances for SB to develop. The second group (filters 4 to 6) uses objective criteria in local weather to confirm SB passage (by onshore winds and weak wind gusts), and to ensure pure SB events (days without precipitation). After running the data through these filters, we identified about 43.4% potential SB days (i.e., 9349 days) at the regional level across Eastern Spain (Fig. 3) and between 33.3% (7179) and 39.2% (8439) of SB days at the coastal level (Table 2) for 1961–2019. We distinguished between SB days at station level (i.e., exact number of events identified for each weather station), coastal level (i.e., encompass all the SB episodes presented in one same coast-region) and regional level (all SB days identified for all stations).

Fig. 2
figure 2

Workflow diagram of the SB automated selection algorithm

Full size image
Fig. 3
figure 3

Total number of SB days identified across all weather stations (regional level) in every month and year for 1961–2019

Full size image

A validation is presented below, whereas thresholds and descriptions of each criterion are described as follows:


Filter 1: To assure a surface pressure gradient weak enough to develop local winds, we used the neutral phase (interval range between −1 and 1) of the Western Mediterranean Oscillation Index (WeMOI, details in Sect. 2.6). Azorin-Molina and Lopez-Bustins (2008) proposed it as the first criterion to detect potential SB passages associated with weak surface pressure conditions over the Western Mediterranean basin.


Filter 2: Eight weather types (A, E, NE, AE, ANE, C, N, and SE) were filtered from the automated classification by the Jenkinson and Collison’s method (1977; JC), as favorable synoptic conditions for SB (Azorin-Molina et al. 2011a). The method uses daily air pressure from NCEP-NCAR reanalysis to compute the 26 weather types defined by Lamb (1950) for the British Islands. A detailed description of the method can be found in Azorin-Molina et al. (2011a).


Filter 3: Weak geostrophic winds (≤ 12 m s−1) exclude days dominated by strong synoptic flows, which disturb SB development (Ramis and Alonso 1988). We calculated the daily 850 hPa geostrophic wind (in m s−1) at 1200 UTC in terms of geopotential, by using the finite difference method which assumes a horizontal constant pressure gradient. Geopotential data at 1200 UTC were taken from ERA5 reanalysis dataset (available online at https://cds.climate.copernicus.eu/; last accessed July 14, 2022).


Filter 4: Winds from the sea were selected based on the coast orientation (see Sect. 2.1 and Fig. 1) and the wind direction at 1200 UTC. Further, to include sideshore SB, we extended orientations at Table 1 in ±15°. For the Balearic islands, it is worth mentioning that we allow all wind directions (i.e., 0º–360º) due to the complexity of defining a coastline orientation; islands can be affected by SB flows from all directions (Ramis and Alonso 1988; Liang and Wang 2017; Cana et al. 2020). Because (i) the poor quality of wind direction data recorded in Spain between 1961 and the mid-1980s (i.e., 8-point or 16-point compass; Gallego et al. 2007; Domínguez-Castro et al. 2014); and (ii) the complexity of defining and applying a homogenization protocol to this variable; we computed wind direction from 10-m height zonal (u) and meridional (v) wind components at 1200 UTC provided by ERA5, as some recent studies suggest a good representation of near-surface winds from this reanalysis (Ramon et al. 2019; Minola et al. 2020).


Filter 5: Daily wind gusts ≤ 13.9 m s−1 were required to exclude moderate-strong onshore flows, as SB gusts in this region do not exceed 14 m s−1 (Azorin-Molina et al. 2011b). For this filter, daily observations were provided by AEMET in section 2.2.


Filter 6: Daily total precipitation < 0.1 mm d−1 was needed to reject days with atmospheric instability (Arrillaga et al. 2020; Grau et al. 2021). Daily precipitation series were provided by AEMET for each station described in Table 1. Further, we used daily total precipitation from ERA5 to full length (since 1961) of the shortest observed-precipitation series (see Table 1) plus few missing data therein.

We compared the accuracy of the proposed method (PM) in this study with the unique reference manual method (RM) developed by Azorin-Molina et al. (2011b) in Alicante (Spain) for 2000–2005. Comparisons between databases were made by computing probabilities of detection of potential SB and non-SB episodes (NSB) first described in Laird et al. (2001), and used by Azorin-Molina et al. (2011b). Probabilities of potential detection of SB days were calculated as follows: (i) identical SB days; (ii) identical NSB days; (iii) different SB days; and (iv) different NSB days; for more details see Table S1. Our PM identified 804 SB events for Alicante Ciudad Jardín weather station (Table 3) for 2000–2005, which is more than half of the number of SB days (1,414) detected by the RM. From 37% of episodes identified for 2000–2005, the validation results for equal SB (48%) and NSB (47%) detections suggest that our method is capable to create a robust long “potential” SB database by taking only pure SB days, i.e., discarding most situations in which SB is combined with synoptic winds. Overall, the methods differed from each other in 11% SB days while 89% of NSB days were due to differences in the type of method (i.e., manual or automatic detection), test criteria, and data sources (Table 3).

Table 3 Comparison of sea-breeze (SB) and non-sea-breeze (NSB) days detected by the reference method (RM) and ours (PM) in the Alicante Ciudad Jardín station for 2000–2005
Full size table

2.5 Trend analysis and statistics

Long-term trends for SB speeds, gusts and occurrence were estimated for 1961–2019. As a first step, we created daily time series of wind speed and wind gust for SB days in each weather station. From daily time series, we derive monthly averaged and monthly occurrence series. Weighted averages were performed by grouping all the SB days found in one same region to create regional and coastal series (see Table 2). After this step, we calculated the anomalies with respect to 1981–2010 to assure that no weather station dominated the averaged series (Zhang et al. 2020). Then, we estimated trends with a weighted linear regression in meters per second per decade (m s−1 decade−1). Trend significance was assessed by applying the non-parametric modified Mann–Kendall test (Hamed and Rao 1998), which corrects the variance for highly autocorrelated series. Statistical significance was also assessed from a “process and importance” perspective, following McVicar et al. (2010) and Minola et al. (2021). This perspective uses three p-level thresholds, defined as (i) highly significant (p < 0.05); (ii) significant (p < 0.10) and (iii) not significant (p > 0.10). To evaluate the trend persistence, we performed running trends (Brunetti et al., 2006), which detect weakened or stronger trends sub-periods (Morán-Tejeda et al. 2016) with a minimum window length of 30 years, which runs from 1961 to 2019. We also computed a 15-year Gaussian low pass filter to illustrate the multidecadal variability of SB, and the coefficient of determination value (R2) to estimate the variance between SB speeds and gusts. All calculations are performed over three different time scales: annually, monthly (most can be found in supplemental material) and boreal seasons: December, January and February (DJF) for winter; March, April and May (MAM) for spring; June, July and August (JJA) for summer; and September, October and November (SON) for autumn. Lastly, to explore the influence of large-scale atmospheric circulation on SB trends and variability, we correlated the weighted SB speeds, gusts and days with (i) the atmospheric circulations indices (see Sect. 2.6) at annual and seasonal scale; and (ii) with the frequency of the 26 JC weather types (Jenkinson and Collison 1977) at monthly scale. These types can be grouped as follows (Azorin-Molina et al. 2011a): (i) directional flow types (N, NE, E, SE, S, SW, W, and NW); (ii) anticyclonic (A) and cyclonic (C) types; and (iii) hybrid types (AN, ANE, AE, ASE, AS, ASW, AW, ANW, CN, CNE, CE, CSE, CS, CSW, CW, and CNW).

2.6 Atmospheric circulation indices

We selected four atmospheric circulation indices to assess their relationship (via Pearson’s correlation coefficient (r) with the observed SB variability over the 58-year study: (i) The daily WeMOI (Martin-Vide and Lopez-Bustins 2006) provided by the Group of Climatology at the University of Barcelona (http://www.ub.edu/gc/wemo/, last accessed July 14, 2022). This index is based on the regional sea-level pressure difference between the barometric dipole San Fernando (Cádiz) and Padua (Italy), a representation is shown in Fig. 1. We used it to detect potential SB passages for the Eastern Spain (Azorin-Molina and Lopez-Bustins 2008); (ii) The MOI (Palutikof 2003) provided by the Climate Research Unit (CRU), available online at https://crudata.uea.ac.uk/cru/data/moi/ (last accessed July 14, 2022); (iii) The NAOI (Jones et al. 1997) retrieved from CRU (https://crudata.uea.ac.uk/cru/data/nao/, last accessed July 14, 2022); and (iv) the SNAOI (Folland et al. 2009) calculated by averaging the NAOI values for July–August. These indices are well known to drive most of the climatic variability in the Eastern Spain (Corell et al. 2020; Martinez-Artigas et al. 2021), and some authors have analyzed their influence on wind speed variability for this region (Azorin-Molina et al. 2014a, 2016).

3 Results

3.1 SB climatology

Figure 3 displays the regional climatology of the SB occurrence across Eastern Spain for 1961–2019. In general, a marked seasonality of the SB is observed, with a greater occurrence in summer (up to 27 days per month), and a transition period in May and September. Further, the low (but not absent) occurrence of the SB in winter seems to increase slightly in recent years, although it does not exceed 21 days per month in any year. The 58-years climatology of the SB speeds and gusts is also presented in Fig. 4. The annual climatology is dominated by the spring–summer season, therefore presenting a similar pattern. North-eastern and southern regions exhibited the largest values at all scales, while the smallest ones occurred in the eastern region. At the intra-annual scale, summer exhibited the highest SB speeds and gusts values (5.3 and 10.8 m s−1, respectively), while strong differences were found among the regions in winter. The spatial climatology (1961–2019) of the mean, standard deviation, minimum and maximum SB speeds and gusts (Fig. S2 and S3) for 16 stations across the Eastern Spain are also presented in the supplemental material. It is remarkable the greater values in airport stations (i.e., rural areas) than in the cities in e.g., Alicante, Valencia and Mallorca locations, due to the effect of urbanization for all statistical parameters.

Fig. 4
figure 4

Annual and seasonal box and whisker plot of SB speeds (a) and gusts (b) for each coastal region for 1961–2019. The mean is represented with a black dot, while the median with a black line. The boxes describe the 25th and 75th percentile range, while the whiskers exhibit the 10th and 90th percentiles. Outlier values are presented with colored dots

Full size image

3.2 Trends in the SB speeds, gusts and occurrence

We found widespread opposite trends between SB speeds and gusts, shown in Table 4a and b, as well as in Fig. 5. For 1961–2019, the regional SB speeds show significant decreases at all timescales, with an annual reduction of − 0.08 m s−1 decade−1 (p < 0.05). On the contrary, SB gusts presented non-significant increasing annual trends (+ 0.01 m s−1 decade−1, p > 0.10). Seasonal trends of gusts are positive in winter (non-significant) and autumn (significant), and opposite negative slopes appear in spring and summer (p > 0.10). The weakening of the SB speeds and the strengthening of the SB gusts occurred during most of the 1961–2019 period (Fig. 5). It is noticeable the recent significant slowdown for the SB speeds in autumn–winter after a long period of non-significant stability or weak increases (Fig. 5b and c). Likewise, in recent years there has been a recent weakening of the winter SB gusts (Fig. 5b), while on the contrary, spring–summer ones began to reinforce.

Table 4 Annual and seasonal trends (m s−1 decade−1) in the mean SB speeds (a), gusts (b), and occurrence (c) in all Eastern Spain and each coastal region for 1961–2019, abbreviated as southern (S), south-eastern (SE), eastern (E), north-eastern (NE), and Balearic Islands (BI)
Full size table
Fig. 5
figure 5

Regional running trends of total SB days (in days decade−1), SB speeds and gusts (in ms−1 decade−1) obtained from weighted average over all 16 stations across Eastern Spain for 1961–2019. White dots represent trends significance at p < 0.10 (small ones) and p < 0.05 (bigger ones). Coastal level running trends are shown in the supplemental material (Figs. S4–S6)

Full size image

In addition, we found contrasting spatial patterns on trends which influence the regional tendency. A marked slowdown in SB speeds was found all year round for coastal regions (p < 0.05; Table 4a), where the weakening significantly dominated 93.7% of the 16-station annual trends (p < 0.05; Fig. 6). Nevertheless, eastern and north-eastern regions (i.e., around 25% of the stations) displayed significant increases in autumn–winter (p < 0.10), although these began to weaken significantly in the recent years (p < 0.10; Fig. S4). SB gusts trends were overall heterogeneous and non-significant (p > 0.10, Fig. 6), but small increases dominated eastern and north-eastern regions in winter and autumn (p < 0.05; Fig. 6) while contrary, the decline was evident in spring for much of the stations, and persistent in the Balearic Islands for almost all seasons (p < 0.05; Fig. 6, Table 4b). Particularly, some cities (e.g., Alicante, Barcelona, Valencia and Palma Mallorca) evidenced opposing trends between airports and urban stations at different time-scales for both parameters.

Fig. 6
figure 6

Spatial distribution of the sign, magnitude of trend (in m s−1 decade−1), and statistical significance (red and blue triangles are significant at p < 0.05; cyan and magenta are significant at p < 0.10; and non-filled triangles are not significant at p > 0.10) of SB speeds, gusts and occurrence trends for 16 stations in Eastern Spain for 1961–2019

Full size image

The regional SB occurrence showed non-significant positive trends annually (+ 0.83 days decade−1, p > 0.10; Table 4c), highlighting the significant upward winter trend (+ 1.70 days decade−1, p < 0.05) and the downward one in summer (− 0.58 days decade−1, p < 0.05). The increasing SB occurrence is even more evident at coastal level for the southern and eastern region with annual magnitudes of change of + 3.15 and + 2.18 days decade−1 respectively (p < 0.05, Table 4c), with a remarkable significant positive trend in winter for all regions (p < 0.05, Table 4c). Lastly, it is worth noting the upward trends at airport stations compared to the decline in urban stations within the same city (e.g., Alicante or Valencia), being Valencia the most noticeable case, showing opposite trends on all time scales (except in winter). This is a clear feature of the effect of the local surface roughness.

3.3 Multidecadal variability of the SB speeds, gusts and days

The regional annual and seasonal multidecadal variability of SB speeds, gusts and occurrence for 1961–2019 are shown in Fig. 7. Overall, the 15-years Gaussian low-pass filter revealed a decoupled variability between SB speeds and gusts in terms of r and R2. This clearly occurs at the annual scale, as well as in summer. For these time-scales, anomalies for SB speeds decreased and for SB gusts increased with simultaneous changes for both parameters around the 1990s. The strongest intra-annual variability of anomalies occurred in winter-autumn (Fig. 7b, c), especially for SB gusts, with the weakest one in summer (Fig. 7d). Winter also presents a strong positive correlation between SB speeds and gusts variability (r = 0.56 and R2 = 0.31; Fig. 7b), the same for spring and autumn. For the SB occurrence, it is remarkable the increasing variability observed in winter (+ 1.70 days decade−1, p < 0.05) and the decreasing trend in summer (− 0.58 days decade−1, p < 0.05). It is evident from Fig. 8 that coastal series also showed a decoupled pattern, especially in the north-eastern region.

Fig. 7
figure 7

Regional variability (weighted) for the SB speeds (blue), gusts (red) and occurrence (green) across Eastern Spain for 1961–2019. The 15-years Gaussian low-pass filter (dashed line) is also shown, while the shaded area represents the standard deviation. Trends for the 58-year period are indicated with their statistical significance, i.e., p < 0.10 (one asterisk) and p < 0.05 (two asterisks), while Pearson correlation coefficient (r) and coefficient of determination (R2) between the SB speeds and gusts are also shown

Full size image
Fig. 8
figure 8

Annual variability (weighted) for the SB speeds (blue), gusts (red) and occurrence (green) for each coastal region over the Eastern Spain for 1961–2019. The 15-years Gaussian low-pass filter (dashed line) is also shown. Trends for the 58-year period are indicated with their statistical significance, i.e., p < 0.10 (one asterisk) and p < 0.05 (two asterisks). Seasonal variability is displayed in the supplemental material (Fig. S10)

Full size image

3.4 Influence of large-scale atmospheric circulation

Figures 9, 10, 11 show spatial r correlations between SB time series and large-scale circulation indices. These statistical relationships are better captured when stations are grouped into coastal regions (Tables 5, 6, 7). Overall, we found that the WeMOI exerted its major influence on SB speeds at the annual scale and in spring for all regions (p < 0.05; Table 5, Fig. 10). The positive correlation is stronger in the south-eastern in all time-scales (p < 0.10), except for winter, when it reverses its sign. Particularly, this positive relationship between SB speeds and the WeMOI may be related to its tendency towards negative phases at all time scales (p < 0.05, except in winter; Table 5). Other atmospheric indices, such as the MOI and the SNAOI partly influenced SB speeds in summer for specific regions or stations. Moreover, the results for the SB gusts were not conclusive, with very few significant correlations, most of them heterogeneous and weak in specific regions at different time scales (Table 6). For the SB occurrence, there is a significant positive relationship with the NAOI at all timescales (p < 0.05; Table 7; Fig. 9), followed by the MOI that also exerted a strong positive and significant influence on the SB days both annually and seasonally, except in summer (p < 0.05; Table 7; Fig. 11), while the WeMOI only showed a negative and significant correlation in winter, but positive in autumn for all coastal regions.

Fig. 9
figure 9

Spatial distribution of the sign and significance of Pearson’s correlation relationship (r) between the SB speeds and the SB gusts anomalies (m s−1), the SB days and the a NAOI and b SNAOI for 16 stations in Eastern Spain for 1961–2019. Markers (plus and minus) indicate the sign of the relationship while colors represent the statistical significance at p < 0.05 (blue and red), p < 0.10 (green and magenta), and p > 0.10 (gray). Pearson’s correlation coefficients for the regional and coastal series are presented in Tables 5, 6, 7

Full size image
Fig. 10
figure 10

As Fig. 9 but for the WeMOI

Full size image
Fig. 11
figure 11

As Fig. 9 but for the MOI

Full size image
Table 5 Trends (decade−1) of annual and seasonal WeMOI, MOI, NAOI, and SNAOI for 1961–2019 and their Pearson’s correlation coefficients (r) with the SB speeds in all Eastern Spain and each coastal region, abbreviated as southern (S), south-eastern (SE), eastern (E), north-eastern (NE), and Balearic Islands (BI)
Full size table
Table 6 As Table 5 but for the SB gusts
Full size table
Table 7 As Table 5 but for the SB days
Full size table

For the 26 JC weather types, the highest positive correlations occurred for the SB days (Fig. 12c), where the A regime enhanced the occurrence of SB events regionally from autumn till spring (+ 0.6 > r <  + 0.8, p < 0.05), followed by the E, NE and AE types (p < 0.05), which exerted a strong influence throughout the year. It is worth mentioning that at coastal level, the summer SB occurrence is clearly inhibited by westerlies, although a frequent anomalous easterly advection may enhance it (p < 0.10; Fig. S15). The most remarkable results for the SB speeds (Fig. 12a) and gusts (Fig. 12b) were their significant negative relationship with the A regime (p < 0.05; except in summer), denoting that a stable atmosphere may weaken SB circulations. Other weather types (i.e., E, N, NE and NW) partly exerted some positive and significant relationships with both parameters, but these results should be carefully examined at the local level (Fig. S12).

Fig. 12
figure 12

Monthly Pearson’s correlation coefficients (r) between the SB a speeds, b gusts and c occurrence, and the 26 Jenkinson and Collison’s (JC) weather types for 1961–2019. Squares mask the r significance at p < 0.05, and circles at p < 0.10. Bar colors refer to r magnitude and sign. Coastal-based relationships between JC types and SB were also explored in supplemental material (Figs. S12–S16)

Full size image

4 Discussion

This study investigated for the first time long-term changes in the SB speeds, gusts and occurrence across Eastern Spain for 1961–2019. We proposed an automated method to identify past SB events by applying six alternative test criteria, with the goal to create the longest (58-years) available database of potential SB episodes across the Spanish Mediterranean coast, and to provide new knowledge about the multidecadal climate and variability of this local wind. Our method represents an alternative technique to identify potential events of SB in areas in which long-climate records over land and sea surfaces are limited. By using a minimum number of variables (Laird et al. 2001), it can be easily applied worldwide in regions with a regional sea-level pressure difference (Azorin-Molina and Lopez-Bustins, 2008).

Our main findings reveal new evidence regarding the changes in SB in the Eastern Spain, a region in which SB represent the most frequent local wind circulation throughout the year (Olcina-Cantos and Azorin-Molina 2004; Azorin-Molina and Martin-Vide 2007). Our results also represent new insights into the stilling vs. reversal phenomena (Roderick et al. 2007; Zeng et al. 2019; Zhang et al. 2020), as it is focused on improving our understanding of seasonal changes of the near-surface wind speeds and gusts driven by local winds. Here we showed a decoupled variability and trends between SB speeds and gusts, which represents a finding never reported before for local winds. Actually, our results also showed that the understanding of SB changes is challenging because drivers may differ between mean and gust speeds (Miller et al. 2003; Shen et al. 2019, 2021b): e.g., we found large-scale circulation being largely responsible for the SB occurrence changes and in lesser extent for SB speeds, but not for SB gusts. A specific discussion of the principal findings of this research follows:

4.1 Increased SB occurrence due to anticyclonic activity and other related large-scale dynamics

The overall increase in the annual SB occurrence is in agreement with some global analyses (Jiang et al. 2010; Perez et al. 2017; Shen and Zhao 2020), but this trend may not be a widespread phenomenon (Pazandeh-Masouleh et al. 2019; Shen et al. 2019). Specifically, the tendency of SB in Eastern Spain suggests that it is becoming a less common phenomenon in summer, one result similar to that found in Shangai (Shen et al. 2019) but in discordance with the hypothetical increase induced by the air-temperature rise (Lebassi-Habtezion et al. 2011; Perez et al. 2017). In contrast, the increased winter SB days have never been reported in long climatological studies, as most research has focused on summer SB. Based on our correlations with large-scale atmospheric circulation, this agrees with the recent finding of a twentieth-century Azores High expansion, in which anticyclonic circulations are significantly more frequent across the western Mediterranean and the Iberian Peninsula in winter (Cresswell-Clay et al. 2022) and could be behind the increase of SB days during cold months (Giorgi and Lionello 2008; Zhang et al. 2012; Rojas et al. 2013; Otero et al. 2018; Seager et al. 2019). This pattern is also projected for the future and is consistent with anthropogenically driven warming (Cresswell-Clay et al. 2022), so we could expect an increase of winter SB occurrence during the twenty-first century. The increase of the SB occurrence is also strongly related to the E, while other regimes such as the AE and NE types partly favor it, agreeing with previous studies (Azorin-Molina et al. 2011a); although with few differences depending on the region (Figs. S10–S14). The most evident pattern found seasonally is the negative effect of westerlies and NW on inhibiting the summer occurrence (Fig. S13), although anomalous easterly advection enhance it (Haarsma et al. 2009; Zhang et al. 2012; Fernández-González et al. 2012). However, a next logical step of this research is to quantify trends of synoptic-scale flows from mean layer vector wind between 1000 and 700 hPa to assess the impact of synoptic flows changes on SB features.

In addition, it is worth to mention that winter anticyclonic circulations over the region (and therefore SB activity) may be favored by the NAOI + (Fernández-González et al. 2012). The strong correlation found between NAOI and SB days provide plausible insights about the dominant relationship between NAOI + , A regime and the increased SB activity over the region in winter, despite this teleconnection index is thought to not represent well the atmospheric variability of the eastern Iberian Peninsula (Martin-Vide and Lopez-Bustins 2006). However, the non-significant trend of NAOI towards positive phases in winter versus the significant negative one in summer-autumn difficult the understanding of its positive correlation with the SB occurrence in all time scales. Other atmospheric oscillations such as MOI also contribute positively the SB activity, but it is negatively correlated in summer. Given that NAO and MO have shown to be strongly linked in winter (Angulo-Martínez and Beguería, 2012), it is possible that coupled modes of atmospheric circulation or their interaction with local mechanisms that take place in the warm months could have a greater weight in explaining the occurrence of the summer SB.

4.2 Uncertainties in the mechanisms driving decoupled SB speeds and gusts

Different studies indicate that global warming is enhancing stability in mid-latitudes due to subsidence that may cause more frequent but weaker SB circulations (Giorgi and Lionello 2008; Zappa et al. 2015; Deng et al. 2021). This is consistent with our findings of higher SB activity but weakened SB speeds throughout the year. Our results suggest that more frequent anticyclonic circulations (Cresswell-Clay et al. 2022) are partly behind the winter’s decline of the SB speeds and gusts, though spatial trends of the last parameter suggest a local behavior that could be due to local mechanisms. Regarding this, recent works suggest that the weakening of summer local winds in Western Europe is due to the effect of air-temperature rise and ocean warming (e.g., the Atlantic warm pool or the Mediterranean hot spot) on reducing the land-sea thermal gradient (Misra et al. 2011; Tuel and Eltahir 2020; Real et al. 2021). Other authors suggested the effect of urban expansion and irrigation patterns on SB speeds declining (Jiang et al. 2010; Vahmani et al. 2016; Shen et al. 2019; Shen and Zhao 2020), being mechanisms that may further explain local-based trends (especially in autumn–winter), as well as the opposing ones between rural and urban stations. Finally, the decline in the SB speeds is contrary to the generalized hypothesis about a reinforced SB driving intensified coastal summer winds (Lebassi‐Habtezion et al. 2011; Vahmani and Ban-Weiss 2016; Azorin-Molina et al. 2018a; Zhang et al. 2020), and specifically converse with the increasing summer near-surface wind speeds trend reported for Spain (Azorin-Molina et al. 2014a, 2016).

The opposite positive trends of the SB gusts introduce complexity to the explanation of the variability of local-winds. These were spatially heterogeneous in sign and magnitude, suggesting the influence of local-based drivers. However, our findings are not conclusive in regards to the mechanisms behind the SB gusts despite having analyzed their relationship with large-scale atmospheric circulation. For SB, the influence of large-scale synoptic winds in specific locations (e.g., offshore flows produce stronger SB gusts; Azorin-Molina and Chen 2009) could be behind the decoupled tendency between the SB speeds and gusts, but probably coupled to local mechanisms such as (i) land degradation and topography (Miao et al. 2003; Marshall et al. 2004); (ii) air temperature warming (Pazandeh-Masouleh et al. 2019); (iii) soil depletion (Diffenbaugh et al. 2005; Grau et al. 2021); (iv) and onshore synoptic flows and their interaction with orography (Zecchetto and De Biasio 2007; Azorin-Molina and Chen 2009) have been suggested to strongly influence wind extremes (e.g., SB gusts and fronts). Besides, it is widely addressed that climate change is enhancing extreme events (Chen et al. 2020), being these driven by local-scale features and processes amplifying their responses and causing stronger gusts (Diffenbaugh et al. 2005; Tuel and Eltahir 2020; Azorin-Molina et al. 2021). Nevertheless, a comprehensive analysis of local effects and global warming is beyond the scope of this research, and it could be addressed in future works through historical observed data, modeling and sensitivity studies. Future attribution research is strongly needed for unrevealing the causes behind the opposite trends in SB speeds and gusts.

4.3 Overview, possible implications and final remarks

To summarize, although some few works have reported opposite trends in SB, e.g., the decreasing summer SB speed in China (Jiang et al. 2010; Shen et al. 2019) or the strengthening of SB gusts in Australia (Pazandeh-Masouleh et al. 2019), opposite trends in regional SB speeds and gusts have never been reported in the literature for Eastern Spain. It is worth mentioning that the strengthening of the SB gusts corresponds with the increase of the near-surface wind gusts in the Iberian Peninsula during the warm months for specific stations (Azorin-Molina et al. 2016). However, the comparison with other results is not straightforward due to the differences in spatial and temporal scales, being the trend analysis of wind speeds sensitive to the period and region taken into account (McVicar et al. 2010; Troccoli et al. 2012). Furthermore, the local-based nature of SB makes the trend comparison between regions even more complex, and SB research has not reported a pattern on global variability and trends of this local wind (Shen et al. 2021b). Likewise, the drivers of these changes may be different (Shen et al. 2019, 2021b; Pazandeh-Masouleh et al. 2019). Among the large-scale mechanisms investigated here, we did not find the processes behind the increasing SB gusts and its decoupling variability with the SB speeds, an uncertainty that still need to be fully understood. With this in mind, one robust approach to better attribute this phenomenon would be to develop sensitivity studies with regional-local climate models; i.e., by integrating all natural and anthropogenic forcing such as fine-scale land-sea air temperature difference, horizontal pressure gradient, land-use changes, soil drying, large-scale synoptic flows, orography, urbanization, among many others (Zhang and Wang 2021).

Our results are based on the largest available dataset of observed near-surface wind speed data in Eastern Spain, even though a longer period or a larger number of stations might help to confirm our findings or to answer some remaining questions. We used low-temporal resolution data (i.e., daily averages from 3 subdaily observations) to calculate daily means of SB speed, one approach that may not capture the whole picture of SB changes but detects regional trends and, at the same time, distinguishes local peculiarities. It is worth mentioning that early in the morning offshore winds (land breezes) do not usually blow or are weak in Eastern Spain, thus SB speeds may be underestimated because of the 0700 UTC observation, though our wind speed dataset is mainly weighted by winds at 1300 and 1800 UTC, reflecting winds associated with SB. We also used daily wind peak gusts recorded as the highest 3-s mean over 24 h, nevertheless the SB days filtered by our method ensure that the wind gusts analyzed here are associated with local wind circulations. We emphasize that the limited availability of reliable measurements at a 10-min, hourly or even at sub-daily scale is still challenging the SB community addressing long-climatological studies. In addition, the limited spatial and temporal resolution of observed weather data over ocean (buoy observation network) and land made difficult to analyze trends of SB characteristics (e.g., inland penetration, onset, duration, etc.) but also physical mechanisms controlling it such as the land-sea air temperature difference.

We end by emphasizing the increasing frequency of anticyclonic circulation (Cresswell-Clay et al. 2022), inducing persistent SB activity (especially in winter) but weaker speeds. This has been previously suggested in previous works, with indications that it could favor the wind industry (Jiang et al. 2010). It is also very interesting that a decrease in the SB speeds does not exclude a winter increase in its extremes, e.g., SB gusts and fronts (Laurila et al. 2021), possibly due to the direct or collateral effect of global warming on changes in atmospheric circulation and land uses (Miao et al. 2003). However, the weakening of summer SB speeds, and the declining of its activity may be explain reductions in summer inland precipitation with implications in amplifying droughts, land aridity and exacerbating wildfires (Millán et al. 2005; Millán 2014; Pastor et al. 2015; Pausas and Millán 2019). In specific locations, stronger but drier gusts may occur as response of the combined effect of air-temperature rise and soil depletion, urbanization and land degradation on reducing the moisture available for SB (Pausas and Millán 2019; Guion et al. 2021). The foregoing shows the implications of filling a research niche of changes in SB on various socioeconomic and environmental spheres.

5 Conclusions

The main findings of the long-term trends and variability of sea breeze occurrence, speeds and gusts in Eastern Spain for 1961–2019 are:

  1. 1.

    The SB occurrence increased annually (+ 0.83 days per decade, p > 0.10) and in winter (+ 1.70 days per decade, p < 0.05). On the contrary, summer SB days decreased at a rate of − 0.58 days decade−1 (p < 0.05).

  2. 2.

    Opposite trends were observed between the SB speeds (− 0.08 m s−1 dec−1, p < 0.05) and the SB gusts (+ 0.01 m s−1 dec−1, p > 0.10) with distinct seasonality, i.e., the SB speeds significantly decreased in spring–summer (− 0.11 m s−1 dec−1, p < 0.05), while the SB gusts increased in autumn (p < 0.05), but not significantly in winter.

  3. 3.

    Trends in the winter SB occurrence are thought to be exerted by frequent anomalous anticyclonic conditions and the NAOI + and MOI + , while inhibiting the SB speeds. The winter’s declining of the SB speeds is mainly due to the influence of the A regime, while the SB gusts reinforcement remain misunderstood. Future sensitivity studies using climate models could help to better understand the drivers controlling their variability in a warming climate.