Figure 3 takes an average of the temperature, precipitation, and wind speed at each station for each month over the 60 year period. It shows that the general annual trends in temperature follow the same pattern at all stations, with the Northeast colder than the Southwest, and the Southwest colder than the South, though with Kirkjubæjarklaustur temperatures more in line with Southwestern stations than Southern ones. A line is drawn at 5°C, above which the months will be referred to as “warm season” and below which will be “cold season.” Months falling between 5 ± 0.7°C will be considered “transitional,” sometimes acting as a warm season month while other times acting as a cold season month.

Like temperature, precipitation ranges are distinguishable by region, though here, Kirkjubæjarklaustur aligns with the Southern stations instead of the Southwestern ones. Vík í Mýrdal precipitation records also far exceed the other stations. Grímsstaðir is the only station with warm season precipitation rising over cold season precipitation. Note that these precipitation records do not distinguish between types, such as rain or snow. No consistent, long-term record of snowfall exists for this time period.

Unlike temperature and precipitation, wind speed does not segregate by region, but rather by proximity to the coast or mountains. Coastal stations Raufarhöfn, Reykjavik, Eyrarbakki, and Vatnsskarðshólar tend to group together with a trend of higher wind speeds, while inland stations of Egilsstaðir, Hæll, and Kirkjubæjarklaustur group together with lower wind speeds. Grímsstaðir, despite also being an inland station, has moderate wind speeds, while Vík í Mýrdal, another coastal station, has wind speeds that are more similar to the inland stations, likely due to its location being sheltered by surrounding mountain terrain. Stórhöfði, the only separate island station, stands as an outlier, with its average wind speed exceeding the other stations’ maximums.

Figure 4 shows the average number of dust days per month over the 60 year period at each station, with the final plot showing the average for the whole of Iceland. Each station peaks in dust day frequency (DDF) at the beginning of the warm season. Overall, dust activity is greatest in the summer, May through July, and lowest in the winter, December and January. Shown in Figure 5, Stórhöfði leads in average number of dust days per year with over 40, while also leading in wind speed at around 11 m/s, which is about 5 m/s above the next highest wind speed. Vík í Mýrdal, with the greatest amount of precipitation averaging at almost 20 cm/year, is among the three lowest DDFs, next to Raufarhöfn and Hæll.

Figure 6 depicts the annual DDF for each region across the 60-year period. Included are vertical lines indicating volcanic eruptions and major jökulhlaups with the potential of affecting annual dust activity. Three major jökulhlaups are indicated in the figure, two from Mýrdalsjökull (located in South Iceland) in 1955 and 1999, and one in 1996 at Grimsvötn Gjalp (from Vatnajökull, which is bordered by Dyngjusandur on its north end and Skeiðarársandur on its south end). In South stations, the 1999 Mýrdalsjökulhlaup is followed by an increase in DDF the next year, but the 1955 Mýrdalsjökulhlaup is followed by a drop in DDF the next year. There is an increase in DDF seen in several stations the year following the Grimsvötn jökulhlaup, but greater peaks are seen throughout the 60 year period. However, the nearest weather stations from the jökulhlaup-affected areas are not always on the main dust plume paths and many dust events remain unreported. This occurs especially in South Iceland where dust plumes from the Mýrdalssandur and Skeiðarársandur areas head directly toward the ocean. The number of dust events after the jökulhlaups can therefore be underestimated.

The most consistent dust activity occurs in South stations, but stations in other regions, such as Grímsstaðir and Eyrarbakki, experienced sharp increases in DDF that at times matched that of Stórhöfði. As seen in Figure 5, this contributes to their high number of average dust days per year. Overall though, Stórhöfði had the highest number of dust days, with a total of just under 2500 dust days from 1950 to 2009, as compared to 118 dust days previously reported at Stórhöfði from 1949 to 2011 (Dagsson-Waldhauserová et al., 2014).

In each station, during the cold season, DDF increases as average temperature rises toward the warm season. It peaks at the start of the warm season, but dips before the return of the cold season. The drop that occurs after the start of the warm season may be because the increased temperature is accompanied by decreased wind speeds. There is likely a threshold temperature around 5°C, below which temperature acts as the limiting factor, as warmer temperatures are necessary in order to evaporate moisture between sediments and reduce particle adhesion for entrainment. Above 5°C, the rate of dessication may be sufficient for the frequency of entrainment to be determined mostly by wind speeds.

Other variance within the warm season may be due to precipitation. If DDF was only determined by wind speeds, it would increase as wind speed picks up after July. Instead, DDF drops even more sharply, coinciding with the sharp increase in precipitation after July. Though precipitation increases up through October, wind speeds also steadily increase, surpassing a threshold at which it outcompetes the effects of heavy precipitation. Once the cold season starts after September or October, entrainment is once again limited by desiccation rate, despite continually increasing wind speeds.

Dagsson-Waldhauserová et al. (2014) reported 1965 to be the driest year for SW Iceland in 100 years, and it is during that same year that Reykjavik peaks in DDF, further indicating a relationship between moisture and entrainment frequency. Furthermore, Dagsson-Waldhauserová et al. (2014) reported that almost half of the dust events in the south part of Iceland occurred in the winter or at subzero temperatures. Though including the dust haze synoptic code increased the number of dust days observed during warmer temperatures, the presence of cold-weather dust activity indicates greater connection of dust events to moisture as opposed to temperature.

Volcanic eruptions that took place within this time period do not appear to have a significant long-term effect extending more than a year beyond the time of eruption, though they may have sizeable impacts on short-term dust activity. For instance, the Eyjafjallajökull eruption in 2010 had increased the dust storm frequency in South Iceland significantly during the 6–9 months following the start of the eruptive explosion (Arnalds et al., 2013; Dagsson-Waldhauserová et al. (2014), though this eruption year is not included in the presented analysis, as it is outside the examined time frame.

While records of minor seasonal jökulhlaups are not present for the observation of seasonal impacts, major jökulhlaups do not appear to have a consistent impact on annual DDF. Prospero et al. (2012) attributes the increase of dust concentration at Stórhöfði in 1997 potentially to the 1996 jökulhlaup, but the total dust days for Stórhöfði in 1997 was 40, which is just below Stórhöfði’s 60 year annual mean of 41.6 (standard deviation of 15.8). They additionally linked the increase in dust concentration at Stórhöfði in 2000 to the 1999 Mýrdalsjökulhlaup, and though the number of dust days in 2000 was relatively high at 74 (two standard deviations above the mean), there are two other spikes in annual dust day number (1981 and 2002) that are at least two standard deviations above the mean and not temporally proximal to a major jökulhlaup. It may be relevant to note that the year 2000 coincides with an eruption of Hekla, located in Southwest Iceland just north of Landeyjasandur. The DDF in 1956, the year following the 1955 Mýrdalsjökulhlaup, was 17, which is greater than a standard deviation below Stórhöfði’s 60 year mean. Remaining South stations also did not reflect DDFs above their stations’ annual mean in the year following each of the three jökulhlaups.

Variations in overall DDF between stations may be due to sediment availability and the downwind location of stations from the dust plume sources. While Stórhöfði’s spot as the highest DDF can be attributed to its extraordinarily high wind speeds, other stations are not as simple. Vík í Mýrdal’s spot among the lowest three DDFs can be attributed to its extraordinarily high precipitation, as well as the fact that synoptic codes for this station were only measured at 9 a.m. and 9 p.m., missing long periods during the day at which dust activity may have occurred. The placement of Hæll and Raufarhöfn among the bottom three alongside Vík í Mýrdal may be due to their location in more vegetative regions, as can be seen in Figure 1. Vegetation serves to prevent erosion by holding together ground sediment and acting as a buffer between entrainable particles and the atmosphere. Egilsstaðir, with unremarkable temperature and wind speeds, but somewhat low precipitation, also has some vegetative cover and, like Raufarhöfn, is a fair distance from a dust plume source. However, it manages a moderate DDF compared to Hæll and Raufarhöfn, mainly due to dust flowing in the northeast direction from the Dyngjusandur dust plume source (Arnalds et al., 2016).

Vatnsskarðshólar has moderate wind speeds and is located in proximity to the Mýrdalsjökull floodplain. Despite this, it experiences a fair amount of precipitation, which is likely responsible for limiting its resulting DDF. Kirkjubæjarklaustur, which has similar wind speeds and precipitation to Vatnsskarðshólar and is likewise situated near major floodplains (Skeiðarársandur and Mýrdalssandur), follows with a similar resulting DDF.

The remaining stations Reykjavik, Eyrarbakki, and Grímsstaðir may owe their DDFs to a combination of moderate meteorological factors and unremarkable locational factors. Reykjavik has moderate temperature, precipitation, and wind speed, and is located a moderate distance from the Hagavatn dust plume source, but also receives dust from the South coast (Landeyjasandur) (Arnalds et al., 2016). As a result, its DDF falls in between the median three (Egilsstaðir, Vatnsskarðshólar, Kirkjubæjarklaustur) and the top three (Eyrarbakki, Grímsstaðir, Stórhöfði). Eyrarbakki, though also with fairly moderate meteorological factors, is located more proximally to an intersection zone of dust plume sources from Hagavatn, Landeyjasandur, and Mælifellssandur. Finally Grímsstaðir, with moderate wind speeds but extraordinarily low precipitation, is located within the dust plume region of Dyngjusandur, surrounded by barren desert.

Grímsstaðir in Northeast Iceland was originally the station with the highest reported frequency of dust storms in Iceland (Dagsson-Waldhauserová et al., 2013, 2014). Inserting all the dust codes into the analyses shows that the highest frequency of dust codes is now reported from South Iceland, at Stórhöfði. This describes the general situation in South Iceland, where many dust plume sources are located close to the seashore and dust plumes can only be captured on satellite images (Dagsson-Waldhauserová et al., 2014). Monitoring stations floating on the sea south of Iceland could possibly record the highest rates of airborne dust observations.