Influence of Trade Winds on the Detection of Trans-Hemispheric Swells near the Canary Islands

Abstract

Trade winds are common in the Canary Islands archipelago and affect not only the weather of the islands but also the local wave climate. On the other hand, the arrival in the Canaries of swells from the Southern Hemisphere is little known, but usual. The records of these swells arriving in the Canary Islands have two clear peaks throughout the year, one in spring and the other in autumn. In this work, how the trade winds influence the detection of this type of swells is studied. It is estimated that only approximately half of this type of wave that reaches the Canary Islands could be adequately recorded in the buoy output data tables by the action of these winds. Therefore, their effects may be underestimated in local wave climate studies.

1. Introduction

The Canary Islands are a volcanic archipelago located off the coast of Africa, in the North Atlantic Ocean, specifically, at a latitude between 27 and 29 degrees north (Figure 1). This archipelago, together with those of Azores, Madeira, and Cape Verde, is part of the Macaronesia Region.

Sea waves are produced by the wind. This phenomenon depends on the speed and duration of the wind and the distance it acts over the sea surface (fetch) [1]. The waves generated by the wind, once they leave the generation zone, evolve into swell-type waves and start to decay as they travel across the ocean.

Despite the great travel distance, waves produced by extratropical cyclones in the South Atlantic Ocean are detected in the southern shoreline of Canary Islands [2], mainly in the westernmost islands of the archipelago. These swells have their most plausible origin in the southwest of the South Atlantic, off the Latin American coast.

There are three well-defined regions where the largest number of extratropical cyclones production are concentrated: on the southern coast of Brazil, off the coast of Uruguay, and, perhaps the most important, off the coast of Argentina [3,4]. The associated winds of these cyclones can produce important sea waves if the right conditions exist (enough wind intensity and duration, sufficient effective fetch, etc.), and eventually evolve into swell-type waves (swells). The ability of swells to reach long distances across the oceans, potentially crossing the Equator to reach coastlines thousands of kilometers away, has been known about and studied for decades [5,6]. This type of waves travel following along geodesic paths (great circles) [7]. Swell waves generated by extratropical cyclones in the Southern Hemisphere, as a result of the long distance traveled, reach the Canary Islands with low wave heights (usually less than 2 m) and long periods (sometimes exceeding 23 s).

On the other hand, most of the Canarian coastline has a narrow and steep coastal platform [8]. The combination of the characteristics of the shorelines and the aforementioned swell waves often results in plunging wave break, which leads to extraordinary wave run-up and overtopping events on the exposed coastline [9].

Nevertheless, the low heights of the swell under study make it difficult to detect, and it is usually masked by the rest of the waves, in spite of the high energy load they carry due to their long periods. Visibility will therefore depend on the sea state. This is formed by the confluence of all the different waves at that point at a given time, both wind waves and swells [10].

The highest wind waves in the Canary Islands are generated by trade winds, the prevailing winds in the Macaronesian region [11]. The aim of this work is to prove that the wind waves generated by trade winds influence the detection of the trans-hemispheric swells, masking the latter and making them difficult to characterize.

2. The Climate of the Canary Islands in the General Context of the Macaronesian Region

Let us point out the main characteristics of the climate of the Macaronesian region in general, and the Canary Islands in particular, in order to understand how trade winds can influence the wave climate.

As a general rule, the Canary Islands enjoy a stable and mild climate, which is why it is a top tourist destination all year round [12]. In detail, there are significant variations between islands, mainly between the geologically younger, rougher, western islands and the much more eroded eastern islands. In addition, within each island, there are major differences according to orientation and altitude [13].

A joint work of the Spanish Meteorological Agency (AEMET) and the Portuguese Institute of Meteorology (IM) concluded with a very detailed description of the climate of Azores, Madeira, and Canary Islands [14]. In short, and focusing on the major areas of the Canary Islands only, according to the Köppen classification, the BWh variety (hot desert) is the most widespread on the easternmost islands (Lanzarote and Fuerteventura) and in the southern bands of Gran Canaria, Tenerife, La Gomera, and El Hierro. The Csb variety (temperate with dry and warm summers) is dominant in the rest of Tenerife, La Gomera, La Palma, El Hierro, and in the higher areas of Gran Canaria. The Csa variety (temperate with hot and dry summers) is prevalent on the north and east coast of La Palma, while on the southwest coast of this island the most widespread variety is BSh (hot steppe).

The climate of the Macaronesian region is strongly affected by the following elements [15]:

• The semi-permanent Azores subtropical high-pressure system.
• Trade winds.
• The African continental thermal low and monsoon.
• African easterly jet/waves.
• Extra/sub-tropical storms.
• Coastal upwelling.
• Saharan dust advection (calima).
• Ocean currents (Canary current).

Many of these factors have an important influence on the sea state of the Canary waters, some on a one-off basis, others on a much more continuous basis, but perhaps the first two are the most important for wave climate:

• The semi-permanent Azores subtropical high-pressure system, and its relation with the Icelandic low determines the North Atlantic Oscillation (NAO). NAO is a fundamental parameter in North Atlantic wave climate [16].
• Northeasterly trade winds with cycles of increased intensity and calm periods are, as already mentioned in the introduction, key players in the production and evolution of sea waves, but they also influence swells coming from other regions.

These two determining factors (NAO and trade winds) are two actors of climate on a global scale. They are intimately related to other important climatic phenomena on a planetary level. There are remarkable works that study the relationship between both or with other phenomena, such as the El Niño–Southern Oscillation (ENSO) or the Atlantic warm pool (AWP) [17,18,19].

3. Materials and Methods: The REDEXT Buoy of Tenerife Sur Data

The research was carried out with 12-year records (period 2009–2020) of wind and waves taken by the buoy that the Spanish Port Administration (Puertos del Estado) installed in the south of Tenerife (Tenerife Sur buoy).

The Tenerife Sur buoy is part of the Spanish deep-sea buoy network REDEXT, which consists of 16 deep-water buoys: two of them are located in the Canary Islands (the second one is moored off the northwest of the island of Gran Canaria) [20]. Some of its characteristics are [21]:

• Mooring position: longitude: 16.61° W, latitude: 28.00° N. Mooring depth: 710 m.
• Technical information: Data sampling frequency: hourly. Type of sensor: directional Oce-Met. Model: SeaWatch.

A time series with 12 years of hourly data is sufficient for a robust statistical study. The buoy is optimally located for recording the swell coming from the south, as the island of Tenerife itself shelters it from the northwest swell, which is the most important swell reaching the Canary Islands (Figure 2). In contrast, the Gran Canaria buoy is exposed to the north swell and is much more sheltered from the south swell, so its series of data is not suitable for this research.

As the buoy is located in deep water even for the longest sea waves, the recorded swell is not affected by the seabed. In addition, the distance to the coast ensures that wind records are not affected by local turbulence due to topographical features or obstacles.

For wind data (hourly data values, acquired at 3 m high over the sea level, calculated for an interval of 10 min), the parameters used were:

• Vv_md: mean wind speed (m/s).
• Dv_md: mean wind direction (°, 0 = N, 90 = E).

For wave data (hourly data values calculated for a time interval of 26 min), the parameters used were:

• Hm0: spectral significant height, hereinafter referred to as significant height Hs (m).
• Tp: peak period (s).
• Dmd_P: mean wave direction at spectral peak (°, 0 = N, 90 = E).

A screening process of the dataset was needed to eliminate null and wrong data. Afterwards, the necessary processing was carried out to match the wind and wave data according to date and time of registration.

All these data have been kindly provided by Puertos del Estado.

4. Results

In this section, the results concerning the relationship between the wave and wind data, and the extratropical cyclones of the Southern Atlantic Ocean, are shown.

4.1. Trade Wind Records

After studying the wind records taken by the buoy, it can be seen that most of the registers come from the northeast (NE) and east–northeast (ENE). These are the directions from which the trade wind comes, and Figure 3 shows how this pattern is repeated throughout the months of the year.

The maximum speed recorded is 21.5 m/s, and the annual average is 5.7 m/s. The monthly average wind speed is not maintained throughout the year but, as can be seen in Figure 4, there are average wind speed peaks in winter and summer.

The mean annual wind speed remains fairly stable over the 12 years of the study, in the range of 5.4 to 5.9 m/s. As for the proportion of wind records with the usual trade wind directions (north–northeast (NNE) to east–northeast (ENE)) of the total, this is fairly constant. This is true except in the years 2011 and 2012, where it is slightly lower, but the direction records for those two years have a large proportion of null data, so they are not significant (Figure 5).

4.2. Swells from the Southern Hemisphere Records

Assuming that the swells move along geodesic paths, a screening of the available data was carried out to ensure that the data come from the swell under study. This filtering was applied to two parameters: the direction the waves come from and the value of the peak period of the swell.

• With respect to direction, all readings that do not come from the window bounded by the shape of the Atlantic Ocean basin were removed. Given the location of the buoy, the range of directions that guarantee the Southern Hemisphere provenance is 184°–208° (Figure 6).
• With respect to the peak period, to ensure that the records come from waves formed thousands of kilometers away, limiting their origin to the Southern Hemisphere, all records with periods shorter than 14 s were eliminated. It should be noted that the periods do not have a clear general distribution, even more so when sea and swell waves coexist in the same area [1]. There are methods to separate these two types of waves from the spectra [22,23], but this is not the case here, since only the buoy-processed output data described above are available. This is why a value limit of wave period is used for partitioning.

With the resulting data, the necessary statistical study was carried out in order to characterize this type of swell and its distribution throughout the year. A total of 4674 hourly records passed the screening process covering the 12-year period, which represents around 5% of the total number of registrations.

The average significant wave height of the records was 67 cm, and the maximum significant height 207 cm. With regard to the peak period, the maximum record was 25.6 s, with the median value being 15.8 s.

With respect to the number of filtered records per month, there are clearly two periods of concentration, in spring and autumn (Figure 7). The shape of the two lines (mean and median) are very close, which leaves no doubt as to the existence of these two peaks in the records. The fact that the mean values are slightly larger than the median indicates some skewness in the distribution (right-skewed distribution).

It can also be clearly seen how the values of the monthly mean and maximum significant heights show a very marked seasonality. In this case, the peak values are recorded in summer, coinciding with the austral winter (Figure 8). The significant cyclonic activity in the Southwest Atlantic during the austral winter is evident [24].

4.3. Possible Correlation of Wind Speed to Southern Swell Records

There is a clear coincidence between the seasonality of wind speed and the number of recorded swells, as can be seen in Figure 9. The minimum monthly average wind speeds in spring and autumn coincide with the peaks of recorded swells. The most pronounced minimum in wind speed coincides with the well-known autumn calm periods in the Canary Islands. The maximum wind speed for which records were obtained after the screening process was 12.1 m/s.

Therefore, the influence of locally recorded wind speed on the number of south swell records is clear. Let us quantify this influence.

To this end, a third parameter was included in the process of screening the swell records: the wind speed measured by the buoy at the time the wave is registered.

For this, a new filtering was carried out by limiting the maximum wind speed, and again counting the number of records for different speed limit values. The threshold values taken were those of the Beaufort anemometric scale for levels 1 (light air) to 5 (fresh breeze) [25]. The total number of records obtained is shown in

Table 1. Records of swells from Southern Hemisphere for different wind velocity threshold values.

Wind Speed (m/s)	Total Number of Records	Swells Records	Annual Mean Ratio of Swell Records
<1.6	7709	644	0.084
<3.4	27,101	2149	0.079
<5.5	47,318	3192	0.067
<8	71,303	4019	0.056
<10.8	90,519	4340	0.048
≤12.1 *	93,549	4372	0.047

* Maximum wind velocity registered at the time of southern swells.

.

The values in Table 1 are plotted in Figure 10. It can be seen that a linear relationship with a good fit is obtained.

The expression of the annual mean proportion of south swell records y is shown as follows:

$$y=0.0895-0.0038 x,$$

(1)

where x is the wind speed threshold in m/s.

The coefficient of determination obtained for the linear regression is R² = 0.9789.

5. Discussion

According to Figure 10, if the wind conditions were completely calm, around 9% of the records taken by the buoy would correspond to swells coming from the Southern Hemisphere.

Since the annual proportion of records of southern swells is 4.6%, it is not unreasonable to think that around only half of the Southern Hemisphere swells are recorded.

The maximum heights of the wave records are obtained in the northern summer, as can be seen in Figure 8. This can be explained by two reasons: first, the important cyclonic activity during the austral winter in the area of origin of this trans-hemispheric swell. Secondly, as the wind speed recorded by the buoy is higher during this season, the wind sea will be larger, and it is more probable for small swell heights to be masked.

In view of the results obtained, where the presence of highly transformed swells, due to the long distances traveled (swell decay), may be masked by local factors, it is necessary to take into account this possible underestimation in their detection, and in the development and validation of models, through the use of bias correction methods as deemed necessary.

At the same time, it is possible that there are other factors influencing the record of the studied swells, such as coincidence with swells generated at higher latitude in the North Atlantic. Further study is needed.

6. Conclusions

In view of the results obtained, we can state that there is a clear relationship between the intensity of the local winds and the proportion of records of swells coming from the third quadrant generated in the extratropical region of the South Atlantic, registered by the deep-water buoy moored off the south of the island of Tenerife.