Assessing Heatwaves and Their Association with North African Dust Intrusions in the Algarve (Portugal)

: Heatwaves are an extreme meteorological event in which affected populations may also be exposed to deteriorated air quality conditions due to the increase in air pollutant concentrations, such as PM 10 (particulate matter < 10 µ g/m 3 ). In order to identify heatwaves (1973–2019) in the region of Faro (Algarve) during the hot season (April–September), the Excess Heat Factor (EHF) index was applied. The Mann–Kendall test revealed an upward trend in three heatwave metrics in Faro, and the trend of accumulated heat load (EHF load) was also positive as would be expected, but its signal was not statistically signiﬁcant. An inventory of North African dust events (2006–2019) was made, and their simultaneous occurrence with heatwaves was assessed, pointing to only 20% of dust events of the Sahara occurring simultaneously with heatwave days. A cluster analysis was conducted on daily geopotential height ﬁelds at 850 hPa level over the 2006–2019 period, and four distinct patterns were identiﬁed as the most prominent synoptic circulations promoting both heatwave conditions and North African dust over the Algarve. scientiﬁc study has been conducted to date to assess the simultaneous occurrence of heatwaves and SDE in Portugal by means of a systematic analysis. Therefore, this study aims to (i) identify heatwaves in the Faro (Algarve) district using the EHF method and ascertain whether their frequency, duration, and EHF load trends increased in Faro between 1973 and 2019; (ii) analyse an inventory of North African dust events to establish any simultaneous occurrences with the detected heatwaves from 2006 to 2019; and (iii) identify the main atmospheric circulation types that lead to the simultaneous occurrence of heatwaves and SDE. pollutants during


Introduction
Heatwaves are a phenomenon known for their impacts not only on human health and mortality but also on the economy and environment [1]. There is no consensual definition of this extreme meteorological event to date despite a better understanding and more accurate heatwave assessments in addition to more characteristics in terms of their description and duration [2]. The effects of abnormally high temperatures depend on both the level of exposure (frequency, duration, intensity) [3] and vulnerability of the population, which may be influenced by several factors, such as physiological aspects, age, gender [4], or preexisting health conditions [5]. The impacts on mortality and morbidity have been described in many parts of the globe (e.g., [4,[6][7][8][9]). Likewise, Portugal has experienced heatwaves that have led to an increase in mortality, as was the case in June 1981 [10], July 1991 [10], and during the European-wide heatwave in 2003 [11,12]. In this study, a more in-depth analysis of heatwaves in the southernmost Portuguese region (Algarve) was conducted using the Excess Heat Factor (EHF). This index was established primarily for assessing Australian heatwaves [13] and combines both the statistical and human-impact features of the identified heatwaves [14]. More recently, studies from other countries, such as the Czech Republic [15] and Greece [14], have provided additional and useful information on the impact of these extreme heat conditions on human health.
During a heatwave, the affected population may suffer as a result of exposure to heat stress and also due to air quality deterioration, as air pollutant concentrations, such as PM 10 (particulate matter < 10 µg/m 3 ) and ozone, are frequently exacerbated under anticyclonic weather conditions, increased temperatures, and light winds [5]. Recent studies have shown the relationship between temperature, ozone, and particulate matter in premature mortality and morbidity. Therefore, it is important to reduce the concentrations of these pollutants during heatwaves [16]. The Sahara Desert is a major source of

Study Area
The Faro district is located in the south of Portugal and is characterized by a Mediterranean climate, with a predominantly cool wet season and dry summer ( Figure 1). sions into the Mediterranean Sea and neighbouring countries, causing health impacts th have been demonstrated in cities such as Madrid, Barcelona, Rome, and Athens (e [17,18]). However, African dust from the Sahara and Sahel regions has also been observ in other global locations, for example, the Caribbean and Southern United States [1 Some studies have shown that PM10 and ozone concentrations during hot summ weather, and particularly during the most extreme heatwaves, can lead to an increase mortality [4,16,20,21]. Several authors state that in some European cities, heatwave effe are more related to respiratory diseases than to cardiovascular problems, e.g., in the su mer of 2003 (June-August), around 1460 deaths were attributable to ambient PM10 exp sure and 1400 deaths to ambient ozone exposure in the Netherlands [20]. Stedman [ also estimated that between 225 and 593 additional deaths were the result of ozone exp sure and 207 due to PM10 in the United Kingdom during the same heatwave episode.
Dust transport from the Sahara Desert is one of the most significant sources of p ticulate matter in the southern Mediterranean [22]. From April to September (warm s son), the Faro district (Algarve, Portugal) is frequently affected by heatwaves and d intrusions [23]. Its proximity to the African continent and the Sahara Desert, the latte source of dust emission and hot and dry continental air mass formation, leads to grea exposure of the population to these phenomena. During these months, there is a sign cant population increase across the Algarve due to the domestic and international infl of tourists. As a region characterized by beach and sun tourism, the Algarve is curren the main tourism destination in Portugal, recording 30.9% of the total overnight stays 2017, followed by the Lisbon Metropolitan Area, with 25.5% [24]. In view of these i portant seasonal changes in the demography of the Algarve, it is crucial to assess a study the occurrence of heatwaves and Saharan Dust Events (SDE), as both are potentia responsible for the negative effects on health [25][26][27]. No scientific study has been co ducted to date to assess the simultaneous occurrence of heatwaves and SDE in Portu by means of a systematic analysis. Therefore, this study aims to (i) identify heatwaves the Faro (Algarve) district using the EHF method and ascertain whether their frequen duration, and EHF load trends increased in Faro between 1973 and 2019; (ii) analyse inventory of North African dust events to establish any simultaneous occurrences w the detected heatwaves from 2006 to 2019; and (iii) identify the main atmospheric circu tion types that lead to the simultaneous occurrence of heatwaves and SDE.

Study Area
The Faro district is located in the south of Portugal and is characterized by a Me terranean climate, with a predominantly cool wet season and dry summer ( Figure 1).  Mainly in June, July, and August, Portugal and the Algarve region in particular is exposed to heatwave events [28][29][30] that typically occur when the subtropical anticyclone is in an abnormally northern position [31]. A strong subtropical ridge covers the Iberian Peninsula, moving north and transporting Saharan tropical air (very hot and dry) which, after crossing the barrier of the Atlas Mountains, hits the Iberian thermal depression [28]. On the other hand, Ferreira and Marques [32] claimed that the predominant meteorological conditions for the occurrence of dust events can be divided into two sets of months: April to June, when the movement of particles from the Sahel to the Sahara occurs, with the influence of a low pressure system in the SW of Portugal and anticyclones in the Mediterranean and Northern Africa; and from July to August, when the transport of particles from the Sahara intensifies, associated with the presence of anticyclones in the Mediterranean and North Africa. The relationship between the Iberian heatwaves and Saharan warm air intrusions is observed to be closer in the southern sectors of the Iberian Peninsula [31]; thus, due to its proximity to North Africa, the Faro ( Figure 2) district is highly sensitive to these phenomena. Mainly in June, July, and August, Portugal and the Algarve region in particular is exposed to heatwave events [28][29][30] that typically occur when the subtropical anticyclone is in an abnormally northern position [31]. A strong subtropical ridge covers the Iberian Peninsula, moving north and transporting Saharan tropical air (very hot and dry) which after crossing the barrier of the Atlas Mountains, hits the Iberian thermal depression [28] On the other hand, Ferreira and Marques [32] claimed that the predominant meteorological conditions for the occurrence of dust events can be divided into two sets of months April to June, when the movement of particles from the Sahel to the Sahara occurs, with the influence of a low pressure system in the SW of Portugal and anticyclones in the Mediterranean and Northern Africa; and from July to August, when the transport of particles from the Sahara intensifies, associated with the presence of anticyclones in the Mediterranean and North Africa. The relationship between the Iberian heatwaves and Saharan warm air intrusions is observed to be closer in the southern sectors of the Iberian Peninsula [31]; thus, due to its proximity to North Africa, the Faro ( Figure 2) district is highly sensitive to these phenomena.

Daily Air-Temperature Data
Series of daily maximum and minimum temperatures (°C) from the Faro/Airport weather station (Figure 2) for the period 1973 to 2019 were collected at the NOAA database, https://www.ncdc.noaa.gov/cdo-web/, accessed on 15 March 2021). These are the longest air temperature series available for the Algarve region, hence allowing for an almost up-to-date and pluri-decadal identification of the heatwaves. Daily mean temperatures were then computed as required for the method of heatwave detection (Excess Heat Factor, explained further ahead in this paper).

Saharan Dust Events (APA Inventory)
In this study, the dust events from northern Africa were identified. The events that occurred between April and September (warm months) were identified through an analysis of the Identification and Evaluation Reports of Natural Events in Portugal by the APA

Daily Air-Temperature Data
Series of daily maximum and minimum temperatures ( • C) from the Faro/Airport weather station (Figure 2) for the period 1973 to 2019 were collected at the NOAA database, https://www.ncdc.noaa.gov/cdo-web/, accessed on 15 March 2021). These are the longest air temperature series available for the Algarve region, hence allowing for an almost up-todate and pluri-decadal identification of the heatwaves. Daily mean temperatures were then computed as required for the method of heatwave detection (Excess Heat Factor, explained further ahead in this paper).

Saharan Dust Events (APA Inventory)
In this study, the dust events from northern Africa were identified. The events that occurred between April and September (warm months) were identified through an analysis of the Identification and Evaluation Reports of Natural Events in Portugal by the APA (Portuguese Environmental Agency), available since 2006. These reports are published annually and contain the identification of Saharan dust events by Portuguese regions and with a monthly temporal resolution. All the events that reached national territory and particularly the Algarve region were considered even if they did not exceed the legally required air-quality limits.

Air-Quality Station Data: PM 10 Daily/Hourly Legal Exceedances
In order to identify the Saharan dust days (SD) that exceeded the PM 10 daily limit values (under Directive 2008/50 on ambient air quality and cleaner air for Europe), data from two air quality stations were analysed: the Joaquim de Magalhães urban station (2011-2016) and the Cerro rural station (2012-2016), both located in the Faro municipality ( Figure 2). The air-quality data from these two stations have the legally required efficiency, i.e., one station should have 85% of observations with continuous measurements, as below this value, the measurements are indicative only [33].
Portuguese law considers that, in the period of a year, the PM 10 value should not exceed 40 µg/m 3 . The daily value of this pollutant shall not exceed 50 µg/m 3 , which may not be exceeded more than 35 times a year [33]. Therefore, days with PM 10 values above 50 µg/m 3 are deemed to correspond to poor air-quality days.

Excess Heat Factor (EHF)
There are a great diversity of indices and methodological approaches to identify heatwaves and assess their duration and intensity (e.g., [2]. This diversity causes limitations and uncertainties on the comparative assessment of heatwaves trends and metrics across the world. As stated in the introduction, the authors decided to adopt the EHF methodology. The Excess Heat Factor-EHF [13,34]-is a heatwave-intensity measure based on the average of the daily mean temperature (DMT) of a three-day period (TDP) for the hot semester (April to September). The two ingredients in the EHF calculation are the significance index-EHI sig (1)-and the acclimatization index-EHF accl (2). In the EHI sig , a three-day-averaged DMT is compared against the 95th percentile for DMT using the following calculation: In the second index, the three-day-averaged DMT is compared with the average daily temperature of the previous 30 days based on the following calculation: Following the calculation of the two indices, the product-EHF (3)-between them is calculated: The highest value between 1 • C and the EHI is selected, as only positive EHF values obtained with the two positive partial indices are considered for the identification of heatwaves.
As the values of the two partial indices increase, so does the intensity of the heatwaves. To be considered a heatwave, there must be a positive EHF for at least three consecutive days, and a negative EHF value signifies the end of the identified heatwave. The Excess Heat Factor load, or cumulative/total EHF daily values across an entire heatwave, was also assessed for each heatwave. The Excess Heat Factor load can be used to assess an event and is useful for agencies, community groups, or business sectors wishing to evaluate their resilience thresholds [13].

Heatwave Trends (Frequency, Duration, and EHF Load)
The Mann-Kendall test is a non-parametric test that is used to determine the existence and significance of trends in data collected in meteorological stations [35]. Using the XLSTAT software, the Mann-Kendall test was applied to identify climate trends for the variables: number of heatwave days (HWF), which means the number of days where EHF > 0, i.e., days with abnormally hot conditions; number of heatwaves (HWN); longest heatwave duration (HWD); and also for the EHF load accumulated for each year under study.

K-Means
In this study, the K-means clustering algorithm [36] was used to classify the atmospheric circulation patterns for the days with simultaneous occurrence of heatwave and SDE. This non-hierarchical method provides excellent separability among the clusters, good temporal and spatial stability [37], and aims to classify data into different groups or clusters, performing an iterative adjustment of data centroids and classifying the observations of a variable into homogeneous K groups [38,39], which should be specified by the user at first, showing the subjectivity of the methodology [37], and then each point is allocated to the cluster to the closest centre, and each center is recalculated as the centroid of the points in it [36]. The cluster analysis was applied to daily geopotential height fields at 850 hPa level for the specified days over the 2006-2019 period from R1 NCEP reanalysis available on the National Centres for Environmental Prediction (NCEP) website. These data were extracted for the geographical domain defined by 20-55 • N latitude and 25 • W-15 • E longitude limits, encompassing a total of 255 points (2.5 • × 2.5 • spatial resolution). SPSS was the software used to carry out the K-means classification, applied to a matrix with 174 rows (heatwave and SD days) × 255 columns (grid points in geopotential height fields). The composites or atmospheric circulation patterns (average of all days included in each cluster obtained) were computed using the CDO program (Climate Data Operators, https://code.mpimet.mpg.de/projects/cdo, accessed on 11 April 2021) and plotted in Panoply software (https://www.giss.nasa.gov/tools/panoply/, accessed on 29 April 2021).

Heatwaves: Trend Analyses
The EHF method enabled the identification of 722 heatwave days and 113 heatwaves between 1973 and 2019, occurring mostly in July (53 heatwaves) and August (41 heatwaves), with an average of approximately two heatwaves per year. According to the EHF method, the only month of the hot semester that did not experience a heatwave was April. The mean duration of the heatwaves was around five days, and the longest lasted 18 days (30 July to 16 August 2003).
The statistics of the non-parametric test on temporal trends are presented in Table 1. to heatwaves having increased in number and duration in the Algarve over the last five decades.
The magnitude of an upward trend in the HWF parameter (number of heatwave days) is the strongest, indicating an increase of 2.5 heatwave days per decade (Figure 3a), while the frequency of heatwave events (HWN) reveals a slower growth (+0.2 heatwaves per decade, Figure 3b). In addition, the duration of the longest heatwave (HWD) increased by 3.1 days over the 47-year period (Figure 3c). The trend of the accumulated heat load (EHF load) is also positive, as would be expected; however, its magnitude is not statistically significant ( Figure 4). The magnitude of an upward trend in the HWF parameter (number of heatwave days) is the strongest, indicating an increase of 2.5 heatwave days per decade (Figure 3a), while the frequency of heatwave events (HWN) reveals a slower growth (+0.2 heatwaves per decade, Figure 3b). In addition, the duration of the longest heatwave (HWD) increased by 3.1 days over the 47-year period (Figure 3c). The trend of the accumulated heat load (EHF load) is also positive, as would be expected; however, its magnitude is not statistically significant (Figure 4).   The outstanding heatwave of 23-29 July 2004 was remarkable for its severity, as can be seen in Figure 5, with an accumulated heat load surpassing several other even longe heatwaves. Figure 5 clearly illustrates how heatwaves have become longer and more se vere in terms of accumulated heat load in Faro (Algarve).
In this respect, it should be noted that the three more intense heatwaves in terms o EHF load (°C 2 ) occurred in the current century: 23-29 July 2004 (~266 °C 2 ), 30 July-16 Au gust 2003 (~146 °C 2 ), and 1-9 August 2018 (133 °C 2 ). The temporal trend of the EHF load over the 1973-2019 period in Faro was positive, in line with the trends detected in th HWF, HWN, and HWD but without statistical significance (Table 1).  The outstanding heatwave of 23-29 July 2004 was remarkable for its severity, as can be seen in Figure 5, with an accumulated heat load surpassing several other even longer heatwaves. Figure 5 clearly illustrates how heatwaves have become longer and more severe in terms of accumulated heat load in Faro (Algarve).  The outstanding heatwave of 23-29 July 2004 was remarkable for its severity, as can be seen in Figure 5, with an accumulated heat load surpassing several other even longer heatwaves. Figure 5 clearly illustrates how heatwaves have become longer and more severe in terms of accumulated heat load in Faro (Algarve).
In this respect, it should be noted that the three more intense heatwaves in terms of EHF load (°C 2 ) occurred in the current century: 23-29 July 2004 (~266 °C 2 ), 30 July-16 August 2003 (~146 °C 2 ), and 1-9 August 2018 (133 °C 2 ). The temporal trend of the EHF load over the 1973-2019 period in Faro was positive, in line with the trends detected in the HWF, HWN, and HWD but without statistical significance (Table 1). Figure 5. Relationship between heatwaves duration and EHF load (°C 2 ). The size of the bubbles depends on the EHF load of each heatwaves, i.e., the larger their size, the greater the accumulated Excess Heat load. Figure 5. Relationship between heatwaves duration and EHF load ( • C 2 ). The size of the bubbles depends on the EHF load of each heatwaves, i.e., the larger their size, the greater the accumulated Excess Heat load.
In this respect, it should be noted that the three more intense heatwaves in terms of EHF load ( • C 2 ) occurred in the current century: 23-29 July 2004 (~266 • C 2 ), 30 July-16 August 2003 (~146 • C 2 ), and 1-9 August 2018 (133 • C 2 ). The temporal trend of the EHF load over the 1973-2019 period in Faro was positive, in line with the trends detected in the HWF, HWN, and HWD but without statistical significance (Table 1).

Air-Quality Deterioration during Heatwaves SDE Inventory
The analysis of the Identification and Evaluation Reports of Natural Events in Portugal made it possible to identify 174 dust events from northern Africa in the warm months (AMJ-JAS) between 2006 and 2019 ( Figure 6). The year 2012 had the lowest number of SDE (8), while in 2017, the highest number of events was recorded (20). The year 2017 was classified as extremely hot (especially April, June, and October) and extremely dry (meteorological drought between April and December) [40], with a very severe fire season [40]. According to Figure 3b, 2017 was the year in which the highest number of heatwaves was detected (6); however, in this year, short-duration heatwaves occurred, with less accumulated EHF load.

SDE Inventory
The analysis of the Identification and Evaluation Reports of Natural Events in Portuga made it possible to identify 174 dust events from northern Africa in the warm month (AMJJAS) between 2006 and 2019 ( Figure 6). The year 2012 had the lowest number of SD (8), while in 2017, the highest number of events was recorded (20). The year 2017 wa classified as extremely hot (especially April, June, and October) and extremely dry (mete orological drought between April and December) [40], with a very severe fire season [40 According to Figure 3b, 2017 was the year in which the highest number of heatwaves wa detected (6); however, in this year, short-duration heatwaves occurred, with less accumu lated EHF load. SDE were most frequent in July and August, and most of them had a shor duration (under five days). The average SDE duration was around four days.  Unlike the heatwaves identified by means of the EHF method, the SDE were detecte in all the months under study. A similarity observed was that both phenomena have higher occurrence in the months of July and August.
In the analysis of the number of SDE corresponding to heatwaves, 42 of the SDE wer found to occur simultaneously with heatwaves between 2006 and 2019, in other words less than half.
However, it is important to evaluate the number of days for which SDE were detecte in the Algarve air-quality stations and their co-occurrence with heatwave days (EHF > 0 According to Figure 7, 2010 was the year with the most SDE corresponding to heatwav days (38% simultaneous occurrence). Ferreira et al. [41] stated that in 2010, the Algarv was the region with the most SDE in Portugal, and the summer was hot and dry. Th heatwave of 2018 (1-9 August), considered one of the most intense of the last decade, als corresponds to an SDE and occurred between 2 and 7 August of the same year.
On average, only 20% of the SDE occurred simultaneously with heatwave days (Fig  ure 7). Unlike the heatwaves identified by means of the EHF method, the SDE were detected in all the months under study. A similarity observed was that both phenomena have a higher occurrence in the months of July and August.
In the analysis of the number of SDE corresponding to heatwaves, 42 of the SDE were found to occur simultaneously with heatwaves between 2006 and 2019, in other words, less than half.
However, it is important to evaluate the number of days for which SDE were detected in the Algarve air-quality stations and their co-occurrence with heatwave days (EHF > 0). According to Figure 7, 2010 was the year with the most SDE corresponding to heatwave days (38% simultaneous occurrence). Ferreira et al. [41] stated that in 2010, the Algarve was the region with the most SDE in Portugal, and the summer was hot and dry. The heatwave of 2018 (1-9 August), considered one of the most intense of the last decade, also corresponds to an SDE and occurred between 2 and 7 August of the same year.
On average, only 20% of the SDE occurred simultaneously with heatwave days (Figure 7).

PM10 Legal Exceedances during SDE
A monthly assessment of the PM10 exceedances during SDE was conducted in orde to evaluate the air quality during these events between 2011 and 2019. According to th Joaquim de Magalhães urban air quality station (Table 2), there was no daily average valu above the legal limit of 50 µg/m 3 on any day during SDE in May, July, and September. Th highest annual hourly average above 50 µg/m 3 and the highest number of consecutiv hours with values above 50 µg/m 3 were recorded (21 h) in August. However, it was in April that the highest value of the data series was recorded, namely 135.7 µg/m 3 (19 April) At the Cerro rural air quality station (Table 3), April saw the highest number of day with an average value ≥ 50 µg/m 3 as well as the highest average hours of days with ≥5 µg/m 3 and highest hourly µg/m 3 value, occurring on 9 April of 2015. In both stations, Au gust was the month with the most SDE, and this may have increased air-quality deterio ration during this month. However, it is for April that both stations show a higher averag of hours that exceed the legal PM10 levels. In fact, the 2015 APA report demonstrates tha the Algarve region recorded the highest number of African intrusions in April, which ma have contributed to greater air-quality deterioration.
Although it would be necessary to evaluate the effect of SDE on the legal exceedance of PM10 (in addition to the African intrusions, this value can be influenced by wildfires for example), particular attention should be given to the month of August, which present a greater frequency of events and whose contribution, although not accounted for, ma

PM 10 Legal Exceedances during SDE
A monthly assessment of the PM 10 exceedances during SDE was conducted in order to evaluate the air quality during these events between 2011 and 2019. According to the Joaquim de Magalhães urban air quality station (Table 2), there was no daily average value above the legal limit of 50 µg/m 3 on any day during SDE in May, July, and September. The highest annual hourly average above 50 µg/m 3 and the highest number of consecutive hours with values above 50 µg/m 3 were recorded (21 h) in August. However, it was in April that the highest value of the data series was recorded, namely 135.7 µg/m 3 (19 April). At the Cerro rural air quality station (Table 3), April saw the highest number of days with an average value ≥ 50 µg/m 3 as well as the highest average hours of days with ≥50 µg/m 3 and highest hourly µg/m 3 value, occurring on 9 April of 2015. In both stations, August was the month with the most SDE, and this may have increased airquality deterioration during this month. However, it is for April that both stations show a higher average of hours that exceed the legal PM 10 levels. In fact, the 2015 APA report demonstrates that the Algarve region recorded the highest number of African intrusions in April, which may have contributed to greater air-quality deterioration.
Although it would be necessary to evaluate the effect of SDE on the legal exceedances of PM 10 (in addition to the African intrusions, this value can be influenced by wildfires, for example), particular attention should be given to the month of August, which presents a greater frequency of events and whose contribution, although not accounted for, may be observed in the increase of PM 10 levels in the Algarve region. It is during this month that the population increases substantially due to tourism, and therefore, the deterioration of air quality should be considered in light thereof.

Atmospheric Patterns
The K-means classification of the atmospheric patterns associated with heatwaves and SD in the Algarve region is depicted in Figure 8. All the clusters exhibit anticyclonic circulations promoting airmass advections from Northern Africa to the southwestern regions of the Iberian Peninsula. The prevalent influence of the subtropical high (Azores anticyclone) in this geographic domain during the summer months explains the relevance of high-pressure centres within the atmospheric circulation associated with heatwaves and SDE in the Algarve region.
Cluster 1 groups 25% of the heatwaves and SD identified over the study period. An anticyclone cantered over the Atlantic area is its most prominent feature, with its eastern margin encompassing the Iberian Peninsula and the north of Morocco and Algeria ( Figure 8). This synoptic pattern promotes the south-easterly fluxes responsible for warm air advection and dust transport into the Algarve region.
Cluster 2 includes 38.5% of the heatwaves and SD, thus the most represented of the four. The presence of a high-pressure centre is visible over northern Algeria and the western Mediterranean, ensuring south-easterly or southerly winds that may explain the occurrence of very hot weather and Saharan dust in the study region.
The third identified pattern (Cluster 3) is apparently quite similar to the Cluster 1 scenario, characterized by an Atlantic anticyclone, with its eastern margin overarching the Iberian Peninsula and the northwestern area of Africa. This cluster includes 47 days (27% of the heatwaves and SD days). The configuration of the isohypses clearly suggests the presence of a thermal low located over northern Morocco and southwestern Iberia, which may reinforce the pressure gradient and increase the speed of southerly air fluxes, hence contributing to African warm airmass advection and Saharan dust.
The less frequent pattern (10% of the events) corresponds to Cluster 4. This synoptic type is characterized by the presence of a strong ridge over Western Europe, with two high pressure cores located over the British Islands and northwest Africa (Morocco and Algeria). The position of this great ridge assures the influence of a southerly circulation over southern Portugal, apparently responsible for warm airmass advection (heatwave conditions) and Saharan dust transport.

Heatwaves
The Mann-Kendall test showed that the HWF, HWN, and HWD have positive trends in Faro (Algarve), which points to an increase in these variables between 1973 and 2019. These results are in line with studies that have indicated an increase in heatwave frequency and duration in Europe over the last decades [42][43][44][45][46]. Regarding heatwave days (HWF), an increase of 2.5 days/decade was estimated, which is very close to the result obtained by Sánchez-Benítez et al. [45]: +2.6 heatwave days/decade, assessing the 1979-2019 period in the Iberian Peninsula) but lower than the 3.8 days/decade estimated by Lorenzo et al. [46] as an average spatial trend for the Iberian Peninsula territory during the period 1971-2000, also using the EHF. The study of Lorenzo et al. [46] provides the most suitable results for a comparison with the present heatwave analysis, considering its geographical context (the Iberian Peninsula) and methodological approach (EHF and Mann-Kendall test). The trend detected in the number of heatwave events (HWN) in Faro (Algarve) was also statistically significant and evaluated at +0.2 events/decade. This heatwave metric was not assessed by Lorenzo et al. [46]. In the present study, an upward trend in the duration of the longest heatwave (HWD) was also found, assessed at +3.1 days over the 1973-2019 period.
In comparison with other approaches, the EHF method has the advantage of quantifying accumulated excess heat load and is more focused on the effects of heat on human health. However, it is apparent that a longer lasting heatwave is not always more intense ( Figure 5) when compared with other longer lasting heatwaves. An example of this is the 2003 heatwave, which is noteworthy for its duration (18 consecutive days in Faro, the longest duration) despite being known for its duration and mortality impacts across Europe [47]. Its intensity was lower in the areas closer to the sea [30], which may explain its sixth position in the ranking of accumulated EHF load values (146 • C 2 ). The longest heatwave of 2003 hit the entire territory of mainland Portugal, but its impacts were greater on inland areas than those close to the Atlantic Ocean [30], which may justify the fact that Faro does not have a higher accumulated EHF load, while the southern region of Alentejo (interior of Portugal) registered an outstanding sequence of extremely hot days [30].
As may be observed in Figure 5, it is the 2004 heatwave that stands out for its intensity or higher integrated EHF (266 • C 2 ). The 2004 heatwave particularly affected the Algarve region [48] not only due to its duration but most notably its intensity, recording the highest maximum temperature of the long-term data under study: 45 • C (25 July). In fact, the highest EHF values were recorded in 2004 ( Figure 4). Conversely, despite the 2018 heatwave being considered the third most intense, studies have revealed that this event did not have a significant impact on mortality in the Algarve region [49].

Saharan Dust Events (SDE)
From April to September, SDE were detected in Faro, mostly with a duration of less than five consecutive days. In the Mediterranean, the relationship between heatwaves and SDE is acknowledged, especially in the summer months [31,50]. Our results show that, on a yearly basis, heatwaves and SDE occur simultaneously only on an average of 20% of the days. Other studies demonstrated that air masses from the North of Africa transport dust that can lead to an increase of PM 10 in the atmosphere in Portugal and other Mediterranean countries [17,23,50]. During the identified SDE, around 20 days exceeded the legal values of PM 10 between 2011 and 2019 at both air quality stations. However, a further study in order to determine the weight of SDE in the increase of PM 10 in the atmosphere is necessary as is further research into the role of the heatwave's mechanisms in this possible increase and concentration of particles.

Atmospheric Patterns of Heatwave and Dust Days
Sousa et al. (2019) [31] already showed that dust intrusions can be a relevant mechanism to trigger heatwaves in Southern Iberia. However, so far, no systematic study had been conducted in this region to identify the main atmospheric patterns responsible for the Atmosphere 2021, 12, 1090 13 of 16 simultaneous occurrence of dust and warm air advections. The present analysis brought clearer evidence about the atmospheric patterns that promote this combination of anomalous weather conditions. Four distinct patterns were identified as the most prominent synoptic circulations that promoted both heatwave conditions and North African dust over the Algarve during the hot season (April-September). An alternative number of clusters (three and five) was assessed, but it did not provide a better distinction of the atmospheric patterns. All four atmospheric types (850 hPa level) are associated with strong subtropical ridges, inhibiting the presence of cloudiness, hence assuring high insolation and hot weather persistence. The analysis of these atmospheric patterns (Figure 8) also revealed the importance of the synoptic circulations responsible for southerly winds in the lower troposphere (850 hPa), which are essential to drive the transport of dust plumes from northern Africa to the Algarve region. These findings are clearly in line with the results obtained in the study by Huerta-Viso et al. [39] on the dust intrusions over the Valencia region (Spain). Considerable similarities were found between the identified patterns and the prevalent meteorological scenarios during the summer months (C1, C3, and C5) determined by Huerta-Viso et al. [39]. Furthermore, it should also be noted that the four atmospheric patterns identified in the present study may be seen as different variants or types of synoptic circulation associated with the influence of a North African high at the upper levels, described as "Scenario 4" in the research of Escudero et al. [51] and also described by these authors as "the most frequent scenario causing dust outbreaks over Iberia ( . . . ) produced by intense heating of the Sahara and the consequent development of the North African thermal low" (p. D18S08). Salvador et al. [37] also highlighted this general synoptic situation (referred to as "type IV" by the authors in their classification) as the most typical in summer, characterized by an "intense North African high over northeastern Algeria and Tunisia, advecting warm African air masses onto the Iberian Peninsula from southern and southeastern areas" (p. 6765).

Conclusions
Heatwaves in Southwestern Europe where the Algarve region is included may sometimes be induced by the transport of hot air masses from North Africa, as highlighted by Theoharatos et al. [52] in their study on heatwaves in Greece in 2007. Papanastasiou et al. [50] stated that the exceedance of air-quality standards correlates well with the annual number of heatwave days and the accumulation of particles favoured by the atmospheric conditions that prevail during this phenomenon.
Although there are other heatwaves definitions, the EHF method was chosen to identify heatwaves at Faro (Algarve). Between 1973 and 2019, the EHF method enabled the identification of 722 heatwave days and 113 heatwaves. The Mann-Kendall test revealed that three heatwave metrics (HWF, HWN, HWD) exhibited positive and statistically significant trends over the 1973-2019 period, i.e., for the last 47 years, these variables have been on an upward trend in Faro. The trend of the accumulated heat load (EHF load) is also positive as would be expected, but its magnitude is not statistically significant. The heatwaves with the highest accumulated EHF load occurred in the 2000-2009 and 2010-2019 decades. The EHF method identified the most intense and durable heatwaves in the Algarve region with precision.
The APA reports made it possible to identify 174 SDE between 2006 and 2019, mostly in August, which corresponds to the second month with the most heatwaves. Around 20% of SDE occur simultaneously with heatwave days. Additionally, in August, both air-quality stations recorded daily and hourly exceedances of the legal values of PM 10 during the SDE. These exceedances may be related to the African dust intrusions identified, but a more in-depth analysis should be conducted in order to ascertain the relevance of SDE in explaining the daily exceedances of PM 10 . On the other hand, air quality during heatwaves can be affected by wildfires, but this contribution was not assessed in this study, which is also a limitation. Finally, in a novel classification, four distinct atmospheric scenarios promoting heatwave conditions and Saharan dust intrusions in the Algarve were identified and assessed. Warm air advections and Saharan dust transport affecting this region are both possible under different synoptic circulation patterns mainly controlled by strong subtropical anticyclonic ridges over the Northwestern Africa.
Author Contributions: R.F. and M.F.; R.F. developed all the research on heatwaves and Saharan dust intrusions, including the statistical analysis and illustration (Figures 1-7); M.F. carried out the section on the atmospheric circulation patterns, including the illustration (Figure 8). R.F. wrote the first draft version of the paper; R.F. and M.F. worked together on the revisions to produce the final version of the paper. Both authors have read and agreed to the published version of the manuscript.