2_Varga_Roettig.indd 121Varga, Gy. et al. Hungarian Geographical Bulletin 67 (2018) (2) 121–141.DOI: 10.15201/hungeobull.67.2.2 Hungarian Geographical Bulletin 67 2018 (2) 121–141. Introduction Global mineral dust particle emission from arid-semiarid areas can be set into the range between 2 and 3 billion tons per year. Wind- blown particles are playing important role in several climatic and other environmen- tal processes, while the accumulated eolian dust deposits are valuable climatic archives (Harrison, S.P. et al. 2001; Kohfeld, K. and Tegen, I. 2007; Maher, B.A. et al. 2010; Pós- fai, M. and Buseck, P.R. 2010; Shao, Y. et al. 2011). Huge amount of dust is deposited far from the source area, contributing to local sedimentary units as exotic mineral material and has an influence on parent material for soils. Examples for atmospheric dust addi- tion with significantly different geochemical fingerprint were reported from several plac- es, e.g. quartz-rich dust addition to basaltic environments: Hawaii (Jackson, M.L. et al. 1971); Cheju (Lim, J. et al. 2005); San Clem- ente Island (Muhs, D.R. et al. 2007b); Canary Islands (Coude-Gaussen, G. et al. 1987); clay- rich dust delivery to Caribbean soils (Pros- pero, J.M. and Lamb, P.J. 2003) and Florida (Muhs, D.R. et al. 2007a) and dust contribu- tion to the formation of red soils in the Medi- 1 Geographical Institute, Research Centre for Astronomy and Earth Sciences, Hungarian Academy of Sciences, Budaörsi út 45, H-1112, Budapest, Hungary. Corresponding author’s e-mail: varga.gyorgy@csfk.mta.hu 2 Institute of Geography, Dresden University of Technology, Helmholtzstraße 10. 01069 Dresden, Germany. E-mail: christopher-bastian.roettig@tu-dresden.de Identification of Saharan dust particles in Pleistocene dune sand- paleosol sequences of Fuerteventura (Canary Islands) György VARGA 1 and Cristopher-Bastian ROETTIG 2 Abstract Automated static image analysis and newly introduced evaluation techniques were applied in this paper to identify Saharan dust material in the unique sand-paleosol sequence of Fuerteventura (Canary Islands). Measurements of ~50,000 individual mineral particles per samples provided huge amount of granulometric data on the investigated sedimentary units. In contrast to simple grain size and shape parameters of bulk samples, (1) parametric curve-fitting allowed the separation of different sedimentary populations suggesting the presence of more than one key depositional mechanisms. Additional (2) Raman-spectroscopy of manually targeted individual particles revealed a general relationship among grain size, grayscale intensity and miner- alogy. This observation was used to introduce the (3) intensity based assessment technique for identification of large number of quartz particles. The (4) cluster and (5) network analyses showed that only joint analysis of size, shape and grayscale intensity properties provided suitable results, there is no specific granulometric parameter to distinguish Saharan dust due to their irregular shape characteristics. The presented methods al- lowed the separation of Saharan dust-related quartz grains from local sedimentary deposits, but due to the lack of robust granulometric characterization of coarsest fractions and due to the diverse geochemical properties of North African sources, exact volumetric amount of deposited dust material and sedimentation rates could not be determined from these data. Keywords: Saharan dust; Canary Islands; grain size; grain shape; automated image analysis Varga, Gy. et al. Hungarian Geographical Bulletin 67 (2018) (2) 121–141.122 terranean (Yaalon, D.H. and Ganor, E. 1973; MacLeod, D.A. 1980; Yaalon, D.H. 1997) or in Australia (Mee, A.C. et al. 2004). North African regions are responsible for 50–70 per cent of the global dust budget and are regarded as the most intense and domi- nant sources of atmospheric eolian dust (Tegen, I. et al. 1996; Mahowald, N.M. et al. 1999, 2006; Ginoux, P.M. et al. 2001; Miller, R.L. et al. 2004). This dust is transported into the direction of Europe, Middle East and the Atlantic Ocean (Israelevich, P.L. et al. 2002; Barkan, J. et al. 2005; Engelstaedter, S. et al. 2006; Stuut, J.-B.W. et al. 2009). The largest amount is transported westward by high al- titude Saharan Air Layer towards the Canary Islands, Caribbean and North America, and by the so-called ‘Harmattan’ winds of the northeasterly trade winds to South Atlantic and South America (Prospero, J.M. et al. 1970; Swap, R. et al. 1992; Prospero, J.M. 1996). Fuerteventura is the second largest mem- ber of the archipelago of the Canary Islands located closest to the African continent, 100 kilometres west of Morocco. The basaltic Eastern Canary Islands are influenced by Saharan dust events, locally called ‘Calima’ (Figure 1). The silt, clay and very fine-sand sized mineral particles are deposited on the widespread, shelf-originated carbon- ate eolianites of the island. Cyclic nature of Quaternary climates, changing amount of transported mineral dust, sea-level vari- ations and related sand availability created a unique carbonate sand dune-paleosol se- quence on the basaltic island, making it an excellent natural laboratory to study the com- plex Quaternary eolian dynamics (Roettig, C-B. et al. 2017). The sedimentary deposits are excellent archives of past environmental changes and landscape evolution history. It is especially true for relatively isolated areas, where to some extent limited transport and depositional mechanisms have played a role in the formation of sedimentary deposits. The present study aims to (1) provide in- formation on granulometric character of vari- ous windblown deposits of Fuerteventura; (2) present a set of new methods to identify Saharan dust material in the carbonate eoli- anite-paleosol sequences of the island. Both of these proposed aims will be discussed by using the results of automated static im- age analysis technique. Determination of granulometric parameters is standing in the focal point of sedimentary studies and it is of growing interest in the Earth sciences. Accurate grain size and shape data of sedi- mentary deposits provide valuable informa- tion on entrainment, transport and accumu- Fig.1. Location of Fuerteventura (Canary Islands) and satellite images of Saharan dust events on 4th February 2013 (NASA Aqua MODIS) (on the left), and on 8th March 2012 (NASA Terra MODIS) (on the right). Source: https://worldview.earthdata.nasa.gov/ 123Varga, Gy. et al. Hungarian Geographical Bulletin 67 (2018) (2) 121–141. lation mechanisms of sedimentary particles, and makes us able to gain insights into soil erodibility (Centeri, Cs. et al. 2015). Investigation area Geological setting of Fuerteventura can be characterized by the products of Neogene and Quaternary alkali basaltic volcanism, unconsolidated carbonate eolianites deliv- ered from exposed shelf areas of the island and admixtured Saharan dust material. Paleo-dune fields and sand sheets of coarse grained biogenetic sandy shelf material with intercalated silty paleosols provide insight into the complex eolian dynamics of the Qua- ternary. Main phases of sand accumulation are dependent on sand availability and geo- morphic stability determined by humidity- driven soil formation. However, as it was reported by Criado, C. et al. (2012), not all reddish layers are in-situ soils, but rather they are formed by higher admixture of silt- sized Saharan dust material with quartz and calcite with some additional feldspar, illite, kaolinite and hematite during periods of reduced sand supply (Roettig, C-B. et al. in press). Nowadays, sand availability has also been a key-issue at the island as the demand for sand has grown tremendously by road and building constructions The identification of past Saharan dust par- ticles and the assessment of their admixture into terrestrial archives play a key role in (1) the understanding of past climate-driven atmospheric circulations; (2) recognition of periods with stable geomorphic evolution and soil formation. Recent observations and measurements allow us to get an overview on dust transportation, deposition and gen- eral characteristics of Saharan dust particles. Three different synoptic meteorological situ- ations have to be taken into account regard- ing the dust availability on Fuerteventura: (1) low altitude easterly winds dominant all year long with winter-spring maximum; (2) sum- mertime dust-bearing Saharan Air Layer as a results of northward migration of the inter- tropical convergence zone (although the main dust transport route is between N15°-21°, a southerly component of flow occur in the lee of the easterly wave); (3) low-level conti- nental trade winds. Modern annual Saharan dust deposition rate is around 20 to 80 g/m2/ year in the region, the reported grain sizes are primarily in the medium and coarse silt frac- tions (Menéndez, I. et al. 2007). The amount of deposited dust in the past could even be sig- nificantly higher (Tsoar, H. and Pye, K. 1987). Methods Granulometric characterization of eolian deposits Samples were taken from 24 silty units con- sidered as paleo-surfaces of stable geomor- phic periods with reduced sand movements and relatively enhanced Saharan dust influ- ence, additional dune sand and sand sheet samples were also investigated as refer- ences for intense sand transportation inter- vals. Detailed description of the units and stratigraphic analysis of selected sites can be found in the works of Faust, D. et al. (2015) and Roettig, C-B. et al. (2017, in press). Air- dried and 2 mm sieved samples were meas- ured by Malvern Morphologi G3-ID instru- ment in the Laboratory for Sediment and Soil Analysis (Geographical Institute, Research Centre for Astronomy and Earth Sciences, Hungarian Academy of Sciences). The applied automated static image analysis technique is a new, innovative mode of grain size and shape analyses completed with chem- ical identity assessments of Raman spectrome- try. In contrast to widely used laser diffraction measurements, image analysis provides direct observational data of particle size, and due to the automatic measurement technique large number of particles are characterized allowing us a more robust and objective granulometric description of particles compared to manual microscopic approaches (Figure 2). 7 mm3 of mineral particles per samples were dispersed by 4 bar compressed air onto a glass slide with 60 s settling time. The used Varga, Gy. et al. Hungarian Geographical Bulletin 67 (2018) (2) 121–141.124 20 x objective lens provide a 960 x magnifica- tion, suitable for detailed characterization of particles in the size range between fine silt and fine sand fractions. Two-dimensional imaging was completed with the usage of additional vertical focal planes, two addi- tional layers were applied above and two other ones below the focus, equivalent to a total of 27.5 µm. The captured high-resolution grayscale im- ages of ~50,000 individual mineral particles were automatically analysed by the device soft- ware to get a raw granulometric data-matrix. Each row of the table represents one sedimen- tary particle (with its own identity number), while the columns are various size and shape parameters, completed with light transmissiv- ity data and Raman correlation scores. Circle-equivalent (CE) diameter is the key size descriptor, calculated as the diameter of a circle with the same area as the projected two- dimensional image of a given mineral grain. Beside several other simple size properties (e.g. length, width, perimeter, sphere-equiv- alent volume), various shape parameters are derived from these sizes. Aspect ratio is the ratio of width and length, circularity describes the proportional relationship between circum- ference of a circle equal to the projected area of the particle and perimeter. Convexity (and solidity) parameters are measures of edge roughness by using the ratio of particle and convex hull perimeter (and area). Circularity and convexity values are also suitable to filter out stacked particles and aggregated particles, in this study particles with lower than 0.65 cir- cularity and convexity values were excluded from further calculations. Intensity mean and standard deviation parameters are determined from the gray- scale images as a results of light transmissiv- ity of particles. These values are dependent on mineralogy, particle thickness, chemical homogeneity and surface roughness (for de- tailed description of the method, see: Varga, Gy. et al. 2018). Intensity values together with chemical identity analyses of the build-in Raman spectrometer provide useful addi- tional information for separation of granu- lometrically similar particles. Identification of Saharan dust material Based on the fact that the Saharan dust de- posited at Fuerteventura is mainly (1) silt- sized and (2) contains a lot of quartz parti- cles (regarded as exotic in the basaltic and carbonate-rich environment of the island), these two deterministic factors were evaluat- ed separately to identify North African dust particles. Three different assessment meth- ods were applied to determine the amount of Saharan dust material of the samples. An indirect approach was applied to theo- retically discriminate the silt-sized sedimen- Fig. 2. Key grain size and shape parameters of mineral particles (modified after Varga, Gy. et al. 2018). – Aspect Ratio = Width/Length; CE Diameter = diameter of a circle with the same area as the projected 2D particle image; Circularity = (2 x Π0.5 x Area0.5)/Perimeter; Convexity = PerimeterConvex hull/Perimeter; Elongation = 1 – Width/ Length, the same as 1 – Aspect Ratio; SE Volume = volume of a sphere with the same CE Diameter as the projected 2D particle image; Solidity = AreaConvex hull/Area 125Varga, Gy. et al. Hungarian Geographical Bulletin 67 (2018) (2) 121–141. tary subpopulations which were mathemati- cally separated. The polymodal grain size dis- tribution curves were partitioned into several unimodal Weibull-distributions by applying parametric curve-fitting technique (Sun, D. et al. 2002, 2004; Varga, Gy. et al. in press). According to the applied parametric curve fitting technique the polymodal particle size curves can be interpreted as sum of several, in this case three overlapping Weibull-functions which represent three sediment populations: where, shape (α1-3), location (β1-3) and weight- ing (c1-3) parameters of the three Weibull- functions were modified by an iterative nu- merical method as a least-square problem to assess the appropriate goodness of fit of measured data and calculated size distribu- tions of constructed subpopulations (Varga, Gy. et al. 2012, in press). According to pub- lished data on recent dust events from the area (Criado, C. and Dorta, P. 2003; von Suchodoletz, H. et al. 2009; Menéndez, I. et al. 2013) and measurements of other far- travelled North African dust material (Var- ga, Gy. et al. 2016), the subpopulation with smallest particles are regarded as the product of long-ranged dust transport. Raman-spectroscopy (at 785 nm wave- length with 3 µm spot) was also applied to directly identify the quartz grains as an indicator of Saharan dust contribution. The acquired spectra of targeted particles were compared to Raman spectral reference librar- ies using KnowitAll® software from Bio-Rad to identify the minerals present. The third applied technique was based on the grayscale intensity mean values of par- ticles, the relatively high values were used as a proxy for quartz grains as it was found that there is a strong correlation between light transmissivity and chemical identity (especially in this special case of an environ- ment characterized with the overwhelming majority of carbonate and quartz particles). Cluster and networks analysis techniques were also applied to differentiate various mineral particle populations based on their general normalized shape (aspect ratio, cir- cularity, convexity, solidity) and grayscale intensity (mean, standard deviation) values. Hierarchical cluster trees were created by using the Euclidean distance pairs of the selected parameters of separated quartz and carbonate size fractions (fine, medium, coarse silt and sand). For network analysis 192 x 192 ([24 sam- ples x 2 minerals x 4 size fractions] x [24 x 2 x 4]) matrix was compiled, where coefficient of determination was calculated for each pair of records based on the normalized shape and grayscale intensity parameters. This matrix was transformed into an adjacency matrix with values of 0, if r2 < 0.99 and 1, if r2 ≥ 0.99, in this way all of the similar mineral grains were coupled and the whole database can be handled as a network or a finite graph, where the similar records (nodes) are con- nected (edges) to each other. The Gephi network visualization software was used to analyse the compiled network by applying the ForceAtlas2 continuous graph layout al- gorithm (Jacomy, M. et al. 2014). Results General granulometric character of sedimentary samples from Fuerteventura Grain size distribution curves of samples from the paleo surface units were poly- modal, mostly with two-three distinct peaks in coarse silt, fine sand and medium sand fractions. Samples taken additionally from sand members of the sedimentary sequence showed a pure unimodal distribution almost entirely in the sand fractions. It is also worth noting, that even these medium and coarse , Varga, Gy. et al. Hungarian Geographical Bulletin 67 (2018) (2) 121–141.126 sand dominated samples contained small amount (< 0.5 vol.%) of silt-sized particles. General granulometric characteristics by size fractions are presented in Table 1. and Figure 3. The dominance of sand-sized frac- tion is clearly visible on circle-equivalent diameter, length and width box-plots; this fraction determines also the bulk grain size values. Shape parameters of samples showed a more diverse picture. Mean aspect ratio val- ues were between 0.75 and 0.80 for all frac- tions, but higher standard deviation scores could be observed in case of sand-sized par- ticles. Circularity and convexity parameters of silt grains were relatively high, especially compared to sand particles, which had a more irregular shape character. Particles with highest solidity values were from the medium and coarse silt-sized fractions, so- lidity parameters of fine silt- and sand-sized grains were lower, but in case of clay parti- cles it could be the result of the small number of pixels on the acquired images of individu- al clay-sized grains. Parametric curve-fitting: the mathematical- statistical approach All of the measured volume-based grain size distribution curves of the measured samples showed a polymodal character. Be- side the sand-sized modes, a clear medium and coarse silt-sized peak is present on the diagrams. By using three three-parameter Weibull-distribution functions, proper good- ness-of-fit values were reached among the constructed and measured distributions (r2 values were 0.98±0.2) during the parametric curve-fitting (Figure 4). Samples could be represented by diverse grain size distributions; the amount of the coarsest fraction was especially various. During the measurements an average of ~50,000 individual mineral particles were scanned, so even a few sand-sized grains could have a significant effect on volume- based grain size distributions. As the result of polydisperse grain size of samples (parti- cle sizes covering several orders of magni- tude: submicron to some few hundred mi- crons) to get a more robust representation several millions of scanned mineral particles would be necessary (Varga, Gy. et al. 2018). Mean modal value of circle-equivalent diameter was 62.3 µm (±12.1 µm standard deviation), while the average median was calculated as 49.2 µm (±9.5 µm standard de- viation) for the 24 samples. Direct differentiation of quartz grains via Raman-spectroscopy The applied measurement system of Mal- vern Morphology G3-ID enables the chemical characterization of dispersed mineral parti- cles with the use of integrated Raman probe. Due to the relatively low number of interpret- able spectra, special focus was given to the medium silt to fine sand-sized components of the samples studied. Only few hundreds of 30–120 μm quartz (Raman shift ~464 cm-1 and carbonate (Raman shift ~1,086 cm-1) grains were identified, and this low number of par- ticles did not allow a mathematically robust, quantitative determination of quartz content. However, some general, broad conclusions could be drawn based on the whole mass of measured samples. Two distinct clusters of quartz and carbonate particles were visible on the circle-equivalent diameter and mean grayscale intensity scatterplots. As mean in- tensity scores are primarily dependent on particle thickness and mineralogy, with the assumptions of high proportion of isotropic particles, lower circle-equivalent diameter re- sults a higher mean grayscale intensity values (Figure 5). This relationship is clearly visible in case of carbonate particles, but quartz grains are lighter in colour, these can be character- ized by higher grayscale intensity values. General grain size properties of the identi- fied quartz and carbonate particles are also presented. As the selection of mineral grains for manually targeted chemical identity analysis cannot be regarded as representa- tive due to the relatively high number of not 127Varga, Gy. et al. Hungarian Geographical Bulletin 67 (2018) (2) 121–141. Ta bl e 1 . M ea n gr ai n siz e a nd sh ap e p ar am et er s o f i nv es tig at ed sa m pl es Sa m pl e na m e C E, µm Fi ne s ilt M ed iu m si lt C oa rs e si lt Sa nd Le ng th W id th Pe ri m et er A sp ec t ra tio C ir cu la ri ty C on ve xi ty So lid ity In te ns ity m ea n In te ns ity ST D vo l. % µm ID -E nc -1 ID -E nc -2 ID -E nc -3 ID -E nc -4 ID -E nc -5 ID -E nc -6 ID -E nc -7 ID -E nc -8 ID -E nc -9 ID -E nc -1 0 ID -E nc -1 1 ID -E nc -1 2 ID -E nc -1 3 ID -E nc -1 4 ID -E nc -1 5 ID -E nc -1 6 ID -E nc -1 7 ID -E nc -1 8 ID -E nc -1 9 ID -E nc -2 0 ID -E nc -2 1 ID -E nc -2 2 ID -E nc -2 3 ID -E nc -2 4 15 0. 70 22 9. 30 18 3. 90 28 6. 60 17 7. 00 22 7. 40 20 4. 10 95 .9 6 17 8. 10 19 7. 40 19 8. 30 19 7. 70 14 8. 20 22 0. 30 10 6. 10 15 0. 50 25 0. 80 16 5. 00 20 6. 20 15 5. 90 98 .3 9 15 4. 70 19 1. 50 14 5. 50 0. 08 04 6 0. 04 73 5 0. 04 33 7 0. 01 92 0 0. 04 31 0 0. 06 08 5 0. 03 40 1 0. 16 38 0 0. 05 21 0 0. 26 67 0 0. 05 10 3 0. 12 90 0 0. 04 12 6 0. 09 19 2 0. 13 03 0 0. 16 05 0 0. 13 63 0 0. 23 43 0 0. 45 11 0 0. 19 41 0 0. 13 86 0 0. 12 26 0 0. 29 80 0 0. 28 09 0 1. 78 4 1. 02 0 2. 03 6 0. 26 0 1. 34 8 1. 24 7 0. 97 7 3. 78 6 0. 34 5 1. 70 8 0. 59 5 1. 69 0 0. 64 6 1. 01 5 1. 57 7 1. 44 1 1. 71 1 2. 46 8 3. 66 9 0. 57 0 2. 21 0 1. 75 9 3. 29 7 3. 76 3 25 .6 10 14 .8 20 21 .1 20 6. 17 8 24 .1 20 14 .3 20 18 .3 70 24 .9 10 9. 00 6 9. 86 4 8. 28 0 12 .4 50 12 .5 30 15 .9 20 21 .2 40 15 .1 60 10 .0 40 21 .4 00 18 .2 90 9. 39 9 27 .0 00 16 .5 00 19 .7 60 20 .2 90 72 .5 3 84 .1 1 76 .8 0 93 .5 4 74 .4 9 84 .3 8 80 .6 2 71 .1 4 90 .6 0 88 .1 6 91 .0 7 85 .7 3 86 .7 8 82 .9 7 77 .0 5 83 .2 4 88 .1 1 75 .9 0 77 .5 9 89 .8 4 70 .6 5 81 .6 1 76 .6 5 75 .6 7 18 7. 8 28 1. 0 23 6. 1 36 1. 5 21 7. 0 25 6. 3 24 6. 8 11 2. 7 20 8. 1 24 0. 9 24 0. 2 25 5. 4 17 9. 6 26 0. 2 12 8. 8 18 0. 1 33 3. 5 19 3. 4 24 5. 8 19 5. 9 12 2. 8 18 9. 8 23 7. 5 18 6. 7 14 0. 1 20 4. 9 16 7. 5 26 8. 0 15 4. 8 21 5. 3 18 6. 4 89 .4 16 1. 8 18 6. 1 18 7. 6 17 8. 8 13 7. 7 20 8. 0 99 .1 4 14 4. 0 22 7. 6 14 5. 0 17 8. 5 13 6. 1 91 .5 14 0. 6 18 4. 0 13 6. 4 45 3. 9 76 9. 1 45 9. 3 98 0. 9 65 4. 6 80 8. 5 63 7. 9 33 7. 9 67 1. 9 60 1. 6 63 5. 1 77 7. 4 47 2. 6 68 3. 6 39 1. 1 55 8. 3 84 6. 5 59 9. 6 78 4. 4 59 6. 2 36 0. 6 50 3. 2 36 7. 7 54 0. 7 0. 79 30 0. 74 88 0. 80 58 0. 77 31 0. 81 05 0. 80 68 0. 74 95 0. 78 05 0. 78 71 0. 80 32 0. 77 69 0. 77 82 0. 78 30 0. 76 40 0. 78 97 0. 77 57 0. 78 62 0. 73 87 0. 73 37 0. 73 37 0. 75 09 0. 78 71 0. 77 14 0. 70 93 0. 86 66 0. 85 65 0. 86 89 0. 83 42 0. 86 7 0. 82 59 0. 82 87 0. 90 62 0. 84 07 0. 82 19 0. 81 98 0. 81 10 0. 84 30 0. 82 29 0. 87 42 0. 84 75 0. 84 59 0. 88 89 0. 85 70 0. 83 92 0. 85 51 0. 83 14 0. 78 54 0. 84 67 0. 93 91 0. 92 92 0. 92 43 0. 88 28 0. 94 21 0. 87 30 0. 89 63 0. 95 18 0. 89 66 0. 87 65 0. 87 89 0. 87 67 0. 90 68 0. 87 26 0. 93 24 0. 90 68 0. 91 54 0. 93 26 0. 90 66 0. 91 45 0. 92 46 0. 89 55 0. 88 69 0. 92 16 0. 95 48 0. 95 92 0. 97 20 0. 96 24 0. 97 13 0. 96 18 0. 95 32 0. 97 52 0. 95 60 0. 96 11 0. 95 73 0. 94 69 0. 96 08 0. 95 68 0. 96 24 0. 95 86 0. 97 45 0. 97 77 0. 97 70 0. 94 96 0. 96 34 0. 96 71 0. 90 48 0. 96 10 27 .6 4 24 .6 2 26 .1 1 22 .5 3 26 .8 4 29 .5 7 29 .0 8 37 .2 0 29 .4 0 29 .8 3 29 .0 6 29 .2 1 32 .2 9 28 .2 1 36 .0 9 32 .0 2 24 .9 9 30 .3 0 27 .6 8 32 .2 0 37 .1 0 31 .1 5 27 .0 9 31 .4 0 25 .0 2 24 .0 7 25 .8 8 24 .4 4 25 .6 5 28 .2 7 27 .0 5 26 .9 1 27 .3 8 28 .2 5 28 .3 2 27 .4 2 27 .6 7 28 .0 7 27 .6 1 27 .2 8 24 .3 6 26 .1 8 26 .3 9 26 .9 7 28 .0 2 27 .6 6 25 .0 8 25 .9 0 Varga, Gy. et al. Hungarian Geographical Bulletin 67 (2018) (2) 121–141.128 Fi g. 3 . B ox -p lo ts o f v ar io us g ra nu lo m et ri c p ar am et er s b y si ze fr ac tio ns . – C E di am et er = ci rc le -e qu iv al en t d ia m et er ; F -s ilt = fi ne si lt (2 .0 –6 .5 μ m ); M -s ilt = m ed iu m si lt (6 .5 –2 0. 0 μm ); C -s ilt = c oa rs e si lt (2 0. 0– 62 .5 μ m ). 129Varga, Gy. et al. Hungarian Geographical Bulletin 67 (2018) (2) 121–141. Fig. 4. Results of parametric curve-fitting of the measured grain size distributions Varga, Gy. et al. Hungarian Geographical Bulletin 67 (2018) (2) 121–141.130 Fig. 4. Continued 131Varga, Gy. et al. Hungarian Geographical Bulletin 67 (2018) (2) 121–141. Fig. 4. Continued Varga, Gy. et al. Hungarian Geographical Bulletin 67 (2018) (2) 121–141.132 interpretable acquired Raman spectra, it is assumed that the grain size distributions are somewhat higher (quartz mode: ~70 μm, median: 68.5 µm; carbonate mode: ~520 µm, median: ~460 µm) than the real particle sizes. Grayscale intensity-based indirect assessment of quartz particles The automatically recorded dimensionless grayscale intensity values served as proxies of optical properties of mineral grains. As it was above shown above, two distinct groups of particles could be separated on the grain size - grayscale intensity scatter plots, and the acquired Raman spectra confirmed our hypothesis that in a given size-class the gray- scale intensities of quartz particles were high- er, meaning brighter (lighter in colour) grains. By using this observation, we applied spe- cific grayscale intensity threshold values (+5%; +1σ [standard deviation] above size-class mean intensities) for every size-classes to identify possible quartz particles (Figure 6). While +1σ thresholds provided too many outliers (prob- ably caused by the low number of scanned particles), the + 5 per cent filtering resulted an average of modal value around 65.1 µm (± 12.7 µm standard deviation) and mean median of the 24 samples was 48.3 µm (± 6.7 µm stand- ard deviation). These values were very simi- lar to the results of fine-grained populations of parametric curve-fittings (W1s), but grain size mode and median of the directly identified quartz particles were slightly higher. Discussion Irregular shape character of the quartz particles The obtained results of the applied various methods demonstrated that quartz and car- bonate particles could be distinguished by simultaneous analysis of size, shape and grayscale intensity values of the investigated samples. Cluster analysis of shape param- eters of medium and coarse silt-sized quartz and carbonate particles showed that the two Fig. 5. Relationship among grain size, grayscale intensity and mineral composition of particles on a grains size vs. intensity scatter plot of quartz and carbonate particles. 133Varga, Gy. et al. Hungarian Geographical Bulletin 67 (2018) (2) 121–141. Fig. 6. Intensity-based assessment of quartz particles based on filtering of + 5 per cent above size-class means of intensity populations were well separated from each other, suggesting main shape properties of the two clusters could be determined to serve as granulometric fingerprint to identify external quartz particles in the investigated deposition- al environment (Figure 7). However, detailed Varga, Gy. et al. Hungarian Geographical Bulletin 67 (2018) (2) 121–141.134 Fig. 6. Continued analysis of different parameters indicated that the homogeneous-inhomogeneous shape characters of carbonate and quartz particles were the main drivers of clustering. All shape parameters in all size fractions were falling into a wider range in case of quartz particles, 135Varga, Gy. et al. Hungarian Geographical Bulletin 67 (2018) (2) 121–141. Fig. 6. Continued while the standard deviation of shape val- ues of carbonates were significantly smaller (Table 2). These phenomena were also recog- nizable on the more platy shape distribution curves of quartz particles compared to the lep- tokurtic carbonate shape distrubutions. Varga, Gy. et al. Hungarian Geographical Bulletin 67 (2018) (2) 121–141.136 Fi g. 7 . C lu st er a na ly si s of m ed iu m a nd c oa rs e si lt- si ze d qu ar tz a nd c ar bo na te p ar tic le s of th e sa m pl es b as ed o n di ffe re nt s ha pe p ar am et er s S o lid it y Medium silt-sized fraction C o n v e x it y Medium silt-sized fraction C ir c u la ri ty Medium silt-sized fraction A s p e c t R a ti o Medium silt-sized fraction A s p e c t R a ti o Coarse silt-sized fraction C ir c u la ri ty Coarse silt-sized fraction C o n v e x it y Coarse silt-sized fraction S o lid it y Coarse silt-sized fraction 0 0 .1 0 .2 0 .3 0 .4 0 .5 0 .6 0 0 .1 0 .2 0 .3 0 .4 0 .5 0 .6 0 0 .1 0 .2 0 .3 0 .4 0 .5 0 .6 0 0 .1 0 .2 0 .3 0 .4 0 .5 0 .6 0 0 .1 0 .2 0 .3 0 .4 0 .5 0 .6 0 0 .1 0 .2 0 .3 0 .4 0 .5 0 .6 0 0 .1 0 .2 0 .3 0 .4 0 .5 0 .6 0 0 .1 0 .2 0 .3 0 .4 0 .5 0 .6 ID -E n c -1 Q ID -E n c -2 Q ID -E n c -4 Q ID -E n c -1 2 Q ID -E n c -1 5 Q ID -E n c -1 6 Q ID -E n c -6 Q ID -E n c -8 Q ID -E n c -1 4 Q ID -E n c -1 1 Q ID -E n c -1 0 Q ID -E n c -7 Q ID -E n c -1 3 Q ID -E n c -3 Q ID -E n c -9 Q ID -E n c -5 Q ID -E n c -1 C ID -E n c -2 C ID -E n c -9 C ID -E n c -1 2 C ID -E n c -4 C ID -E n c -3 C ID -E n c -8 C ID -E n c -5 C ID -E n c -6 C ID -E n c -1 3 C ID -E n c -7 C ID -E n c -1 1 C ID -E n c -1 5 C ID -E n c -1 4 C ID -E n c -1 0 C ID -E n c -1 6 C II D -E n c -1 Q ID -E n c -2 Q ID -E n c -4 Q ID -E n c -3 Q ID -E n c -5 Q ID -E n c -6 Q ID -E n c -8 Q ID -E n c -9 Q ID -E n c -1 6 Q ID -E n c -1 1 Q ID -E n c -1 4 Q ID -E n c -1 5 Q ID -E n c -1 2 Q ID -E n c -7 Q ID -E n c -1 0 Q ID -E n c -1 3 Q ID -E n c -1 C ID -E n c -9 C ID -E n c -1 4 C ID -E n c -2 C ID -E n c -1 6 C ID -E n c -1 2 C ID -E n c -4 C ID -E n c -1 0 C ID -E n c -3 C ID -E n c -8 C ID -E n c -6 C ID -E n c -1 5 C ID -E n c -7 C ID -E n c -1 1 C ID -E n c -1 3 C ID -E n c -5 C ID -E n c -1 Q ID -E n c -4 Q ID -E n c -2 Q ID -E n c -5 Q ID -E n c -1 C ID -E n c -2 C ID -E n c -1 2 C ID -E n c -7 C ID -E n c -1 0 C ID -E n c -1 6 C ID -E n c -9 C ID -E n c -1 1 C ID -E n c -1 5 C ID -E n c -4 C ID -E n c -3 C ID -E n c -5 C ID -E n c -6 C ID -E n c -1 3 C ID -E n c -1 4 C ID -E n c -8 C ID -E n c -3 Q ID -E n c -9 Q ID -E n c -6 Q ID -E n c -1 4 Q ID -E n c -1 0 Q ID -E n c -1 3 Q ID -E n c -1 5 Q ID -E n c -1 1 Q ID -E n c -8 Q ID -E n c -1 2 Q ID -E n c -1 6 Q ID -E n c -7 Q ID -E n c -1 Q ID -E n c -4 Q ID -E n c -2 Q ID -E n c -5 Q ID -E n c -1 C ID -E n c -1 2 C ID -E n c -4 C ID -E n c -2 C ID -E n c -1 0 C ID -E n c -1 6 C ID -E n c -7 C ID -E n c -1 1 C ID -E n c -3 C ID -E n c -5 C ID -E n c -6 C ID -E n c -1 3 C ID -E n c -1 4 C ID -E n c -9 C ID -E n c -1 5 C ID -E n c -8 C ID -E n c -3 Q ID -E n c -9 Q ID -E n c -1 0 Q ID -E n c -1 1 Q ID -E n c -1 3 Q ID -E n c -7 Q ID -E n c -8 Q ID -E n c -1 5 Q ID -E n c -6 Q ID -E n c -1 4 Q ID -E n c -1 2 Q ID -E n c -1 6 Q ID -E n c -1 Q ID -E n c -7 Q ID -E n c -1 4 Q ID -E n c -1 3 Q ID -E n c -1 6 Q ID -E n c -2 Q ID -E n c -8 Q ID -E n c -4 Q ID -E n c -6 Q ID -E n c -1 5 Q ID -E n c -5 Q ID -E n c -1 1 Q ID -E n c -3 Q ID -E n c -9 Q ID -E n c -1 0 Q ID -E n c -1 2 Q ID -E n c -1 C ID -E n c -2 C ID -E n c -3 C ID -E n c -8 C ID -E n c -1 5 C ID -E n c -4 C ID -E n c -1 0 C ID -E n c -5 C ID -E n c -1 3 C ID -E n c -1 1 C ID -E n c -1 4 C ID -E n c -1 6 C ID -E n c -6 C ID -E n c -9 C ID -E n c -1 2 C ID -E n c -7 C ID -E n c -1 Q ID -E n c -2 Q ID -E n c -8 Q ID -E n c -1 4 Q ID -E n c -4 Q ID -E n c -5 Q ID -E n c -1 5 Q ID -E n c -1 2 Q ID -E n c -9 Q ID -E n c -1 1 Q ID -E n c -1 0 Q ID -E n c -1 6 Q ID -E n c -1 3 Q ID -E n c -7 Q ID -E n c -6 Q ID -E n c -3 Q ID -E n c -1 C ID -E n c -2 C ID -E n c -1 2 C ID -E n c -7 C ID -E n c -9 C ID -E n c -4 C ID -E n c -1 5 C ID -E n c -1 3 C ID -E n c -1 4 C ID -E n c -1 6 C ID -E n c -1 0 C ID -E n c -1 1 C ID -E n c -3 C ID -E n c -5 C ID -E n c -8 C ID -E n c -6 C ID -E n c -1 Q ID -E n c -2 Q ID -E n c -1 4 Q ID -E n c -1 6 Q ID -E n c -8 Q ID -E n c -4 Q ID -E n c -6 Q ID -E n c -7 Q ID -E n c -1 2 Q ID -E n c -1 5 Q ID -E n c -9 Q ID -E n c -1 3 Q ID -E n c -5 Q ID -E n c -1 0 Q ID -E n c -1 1 Q ID -E n c -1 C ID -E n c -1 2 C ID -E n c -9 C ID -E n c -2 C ID -E n c -1 6 C ID -E n c -1 5 C ID -E n c -7 C ID -E n c -1 3 C ID -E n c -1 0 C ID -E n c -1 1 C ID -E n c -8 C ID -E n c -1 4 C ID -E n c -4 C ID -E n c -6 C ID -E n c -3 C ID -E n c -5 C ID -E n c -3 Q ID -E n c -1 Q ID -E n c -2 Q ID -E n c -1 4 Q ID -E n c -8 Q ID -E n c -1 3 Q ID -E n c -5 Q ID -E n c -6 Q ID -E n c -7 Q ID -E n c -1 6 Q ID -E n c -4 Q ID -E n c -9 Q ID -E n c -1 5 Q ID -E n c -1 2 Q ID -E n c -1 0 Q ID -E n c -1 1 Q ID -E n c -1 C ID -E n c -2 C ID -E n c -1 5 C ID -E n c -4 C ID -E n c -1 3 C ID -E n c -1 4 C ID -E n c -7 C ID -E n c -1 0 C ID -E n c -1 1 C ID -E n c -1 6 C ID -E n c -9 C ID -E n c -1 2 C ID -E n c -3 Q ID -E n c -3 C ID -E n c -8 C ID -E n c -5 C ID -E n c -6 C 137Varga, Gy. et al. Hungarian Geographical Bulletin 67 (2018) (2) 121–141. Ta bl e 2 . S ha pe a nd g ra ys ca le in te ns ity m ea ns a nd st an da rd d ev ia tio ns o f q ua rtz (Q ) a nd ca rb on at e ( C) p ar tic les b y siz e f ra ct io ns * In di ca to r A sp ec t ra tio C ir cu la ri ty C on ve xi ty So lid ity M ea n in te ns ity In te ns ity S ta nd ar d de vi at io n M ea n St an da rd de vi at io n M ea n St an da rd de vi at io n M ea n St an da rd de vi at io n M ea n St an da rd de vi at io n M ea n St an da rd de vi at io n M ea n St an da rd de vi at io n FS (C ) FS (Q ) M S (C ) M S (Q ) C S (C ) C S (Q ) Sd (C ) Sd (Q ) 0. 80 9 0. 76 0 0. 78 2 0. 68 0 0. 77 7 0. 65 5 0. 77 7 0. 64 0 0. 01 9 0. 03 8 0. 01 2 0. 04 1 0. 01 1 0. 02 9 0. 03 7 0. 04 9 0. 95 8 0. 94 4 0. 96 2 0. 90 7 0. 94 0 0. 87 6 0. 82 8 0. 80 6 0. 00 8 0. 02 1 0. 00 5 0. 03 8 0. 00 9 0. 02 3 0. 02 7 0. 04 1 0. 99 3 0. 98 8 0. 99 5 0. 98 6 0. 98 9 0. 98 1 0. 89 2 0. 93 7 0. 00 2 0. 00 4 0. 00 1 0. 00 5 0. 00 3 0. 00 7 0. 02 1 0. 02 1 0. 97 8 0. 97 1 0. 99 5 0. 97 6 0. 98 3 0. 96 1 0. 95 7 0. 91 3 0. 00 6 0. 00 6 0. 00 1 0. 01 4 0. 00 4 0. 01 3 0. 01 5 0. 02 7 11 0. 74 2 11 5. 44 2 88 .3 96 10 5. 13 8 54 .7 08 65 .2 39 26 .8 73 49 .3 95 2. 10 3 1. 40 3 2. 00 5 1. 56 8 2. 32 2 2. 40 9 3. 28 8 1. 69 4 11 .4 13 8. 76 4 20 .4 08 13 .0 67 28 .5 10 24 .5 71 26 .4 40 26 .8 10 1. 05 1 0. 89 9 0. 78 6 0. 50 2 0. 40 5 0. 79 7 1. 53 8 1. 24 1 *F S = fin e si lt; M S = m ed iu m s ilt ; C S = co ar se s ilt ; S d = sa nd . Network analysis showed similar results. More connections among different size frac- tions of different particle groups marked similar shape properties in Figure 8. It is clearly visible that majority of carbonates were clustered into distinct groups, except for fine silt-sized particles as a result of anal- ogous shapes of smallest particles caused by the relatively low-resolution of the acquired images in this scale. Medium silt-sized quartz particles showed some clustering in case of more than a half of the samples, but overwhelming majority of coarse sized fractions of quartzes appeared as individual samples without any connec- tion to other ones. As the largest volumetric proportion of supposed Saharan dust mate- rial is falling into the coarse silt-sized frac- tion and even this size population showed the least clustered shape-determined structure, it could be ascertained, that we were not able to find any simple shape parameters to distin- guish Saharan dust-related sedimentary pop- ulations. However, the joint characterization of size, shape and grayscale intensity prop- erties allowed us to differentiate the North African dust material from local carbonates. Potential source areas of identified dust particles The lack of specific shape properties of depos- ited Saharan dust particles in Fuerteventura can be traced back to diverse geological and pedological character of dust source areas. Previous satellite based studies (e.g. Pros- pero, J.M. et al. 2002; Washington, R. et al. 2003; Engelstaedter, S. et al. 2006; Goudie, A.S. and Middleton, N.J. 2006; Varga, Gy. 2012) on Saharan dust source areas allowed us to identify potential source areas of miner- al dust particles deposited on Fuerteventura. While comprehensive reviews on geochemi- cal characteristics of North African sources provided information on geochemical, min- eralogical and isotopic composition of the dust material (Scheuvens, D. et al. 2013). One of the potential source areas is located in the southern part of the Taoudenni Basin Varga, Gy. et al. Hungarian Geographical Bulletin 67 (2018) (2) 121–141.138 Fi g. 8 . N et w or k an al ys is o f q ua rtz (Q ) a nd c ar bo na te (C ) p ar tic le s by s am pl es a nd s iz e fr ac tio ns . – F S = fin e si lt; M S = m ed iu m s ilt ; C S = co ar se s ilt ; S d = sa nd F in e s ilt -s iz e d q u a rt z a n d c a rb o n a te g ra in s S a n d -s iz e d c a rb o n a te g ra in s M e d iu m s ilt -s iz e d c a rb o n a te g ra in s M e d iu m s ilt -s iz e d q u a rt z g ra in s C o a rs e s ilt -s iz e d c a rb o n a te g ra in s C o a rs e s ilt -s iz e d q u a rt z g ra in s 01 _F S ( Q ) 01 _F S ( C ) 01 _M S ( Q ) 01 _M S ( C ) 01 _C S ( Q ) 01 _C S ( C ) 01 _S d ( Q ) 01 _S d ( C ) 02 _F S ( Q ) 02 _F S ( C ) 02 _M S ( Q ) 02 _M S ( C ) 02 _C S ( Q ) 02 _C S ( C ) 02 _S d ( Q ) 02 _S d ( C ) 03 _F S ( Q ) 03 _F S ( C ) 03 _M S ( Q ) 03 _M S ( C ) 03 _C S ( Q ) 03 _C S ( C ) 03 _S d ( Q ) 03 _S d ( C ) 04 _F S ( Q ) 04 _F S ( C ) 04 _M S ( Q ) 04 _M S ( C ) 04 _C S ( Q ) 04 _C S ( C ) 04 _S d ( Q ) 04 _S d ( C ) 05 _F S ( Q ) 05 _F S ( C ) 05 _M S ( Q ) 05 _M S ( C ) 05 _C S ( Q ) 05 _C S ( C ) 05 _S d ( Q ) 05 _S d ( C ) 06 _F S ( Q ) 06 _F S ( C ) 06 _M S ( Q ) 06 _M S ( C ) 06 _C S ( Q ) 06 _C S ( C ) 06 _S d ( Q ) 06 _S d ( C ) 07 _F S ( Q ) 07 _F S ( C ) 07 _M S ( Q ) 07 _M S ( C ) 07 _C S ( Q ) 07 _C S ( C ) 07 _S d ( Q ) 07 _S d ( C ) 08 _F S ( Q ) 08 _F S ( C ) 08 _M S ( Q ) 08 _M S ( C ) 08 _C S ( Q ) 08 _C S ( C ) 08 _S d ( Q ) 08 _S d ( C ) 09 _F S ( Q ) 09 _F S ( C ) 09 _M S ( Q ) 09 _M S ( C ) 09 _C S ( Q ) 09 _C S ( C ) 09 _S d ( Q ) 09 _S d ( C ) 10 _F S ( Q ) 10 _F S ( C ) 10 _M S ( Q ) 10 _M S ( C ) 10 _C S ( Q ) 10 _C S ( C ) 10 _S d ( Q ) 10 _S d ( C ) 11 _F S ( Q ) 11 _F S ( C ) 11 _M S ( Q ) 11 _M S ( C ) 11 _C S ( Q ) 11 _C S ( C ) 11 _S d ( Q )11 _S d ( C ) 12 _F S ( Q ) 12 _F S ( C ) 12 _M S ( Q ) 12 _M S ( C ) 12 _C S ( Q ) 12 _C S ( C ) 12 _S d ( Q ) 12 _S d ( C ) 13 _F S ( Q ) 13 _F S ( C ) 13 _M S ( Q ) 13 _M S ( C ) 13 _C S ( Q ) 13 _C S ( C ) 13 _S d ( Q ) 13 _S d ( C ) 14 _F S ( Q ) 14 _F S ( C ) 14 _M S ( Q ) 14 _M S ( C ) 14 _C S ( Q ) 14 _C S ( C ) 14 _S d ( Q ) 14 _S d ( C ) 15 _F S ( Q ) 15 _F S ( C ) 15 _M S ( Q ) 15 _M S ( C ) 15 _C S ( Q ) 15 _C S ( C ) 15 _S d ( Q ) 15 _S d ( C ) 16 _F S ( Q ) 16 _F S ( C ) 16 _M S ( Q ) 16 _M S ( C ) 16 _C S ( Q ) 16 _C S ( C ) 16 _S d ( Q ) 16 _S d ( C ) 17 _F S ( Q ) 17 _F S ( C ) 17 _M S ( Q ) 17 _M S ( C ) 17 _C S ( Q ) 17 _C S ( C ) 17 _S d ( Q ) 17 _S d ( C ) 18 _F S ( Q ) 18 _F S ( C ) 18 _M S ( Q ) 18 _M S ( C ) 18 _C S ( Q ) 18 _C S ( C ) 18 _S d ( Q ) 18 _S d ( C ) 19 _F S ( Q ) 19 _F S ( C ) 19 _M S ( Q ) 19 _M S ( C ) 19 _C S ( Q ) 19 _C S ( C ) 19 _S d ( Q ) 19 _S d ( C ) 20 _F S ( Q ) 20 _F S ( C ) 20 _M S ( Q ) 20 _M S ( C ) 20 _C S ( Q ) 20 _C S ( C ) 20 _S d ( Q ) 20 _S d ( C ) 21 _F S ( Q ) 21 _F S ( C ) 21 _M S ( Q ) 21 _M S ( C ) 21 _C S ( Q ) 21 _C S ( C ) 21 _S d ( Q ) 21 _S d ( C ) 22 _F S ( Q ) 22 _F S ( C ) 22 _M S ( Q ) 22 _M S ( C ) 22 _C S ( Q ) 22 _C S ( C ) 22 _S d ( Q ) 22 _S d ( C ) 23 _F S ( Q ) 23 _F S ( C ) 23 _M S ( Q ) 23 _M S ( C ) 23 _C S ( Q ) 23 _C S ( C ) 23 _S d ( Q ) 23 _S d ( C ) 24 _F S ( Q ) 24 _F S ( C ) 24 _M S ( Q ) 24 _M S ( C ) 24 _C S ( Q ) 24 _C S ( C ) 24 _S d ( Q ) 24 _S d ( C ) 139Varga, Gy. et al. Hungarian Geographical Bulletin 67 (2018) (2) 121–141. NW from the large bend of Niger River and west from the Adrar des Ifoghas. This area is covered by the deposits of Pleistocene Lake Araouane, one of the largest pluvial lakes in Africa (Bridges, E.M. 1990). The salt and diatomite deposits of the enclosed basin are visible also on satellite images. The prevail- ing NE trade winds formed an extensive system of barchanoid dunes, partly covering the surface of the ancient lakebed. The dust emission mechanism of the region and in- tensive deflation of fine-grained particles of lacustrine deposits are enhanced by the bom- bardment energy of saltating sand particles. A long narrow band of dust sources is lo- cated at the western part of the Sahara at the eastern slopes of gently rolling hills running parallel to the Atlantic coast. Couple of sea- sonal streams (with frequent flash floods in the spring) and sebkhas (e.g. Sebkha Ijil) ly- ing on the pedimented surface of the Adrar Souttouf and Zemmour Massif are the main sources of fine-grained material in this region. Several other sources are associated with the large alluvial fans and extensive wadi- system located at the western and north- western slopes of the Ahaggar. The Tidikelt Depression at northern part of the region, sur- rounded by plateaus (the Tanezrouft to the south and Plateau du Tademait to the north), by mountains (Ahaggar and Tassili-n-Ajjer to the east) and by the sand sea of Erg Chech to the west has an extensive ephemeral drain- age system including several wadis from el- evated regions, seasonal marshes and mud flats (Glaccum, R.A. and Prospero, J.M. 1980). Conclusions Automated static image analysis provided huge amount of granulometric (size and shape) data on sedimentary deposits of Fuerteventura. The presented set of meth- ods provided new data on the granulometric character, depositional mechanisms and ad- mixture of dust material to sandy units. Para- metric curve-fitting suggested the presence of more than one key depositional mechanisms, Raman-spectroscopy of manually targeted individual particles revealed a general rela- tionship among grain size, grayscale intensity and mineralogy, while intensity based assess- ment technique was introduced for identi- fication of large number of quartz particles. All three presented evaluation methods have their own advantages and drawbacks. Para- metric curve fitting is a relatively fast tech- nique, it is only based on single one grain size distribution, but it does not take any shape or mineralogy-related information into consid- eration. Raman spectroscopy provide direct chemical identity data on the selected parti- cles, but generally the number of character- ized grains is several orders of magnitude smaller than the whole investigated particle population. The grayscale intensity-based assessment is an indirect technique, but it is suitable for identification of large number of exotic particles. According to our results, there is no specific granulometric fingerprint parameter for iden- tification of Saharan dust material in the de- posits of Fuerteventura, but joint applications of several size, shape and grayscale intensity values and mathematical techniques allowed the separation of quartz particles from other local sedimentary units. However, the lack of robust granulometric characterization of largest fractions, caused by the wide polydispersity of the mineral deposits, did not allow a stable quantitative assessment of volumetric amount of quartz material. Determination of the total mass of deposited Saharan dust material is a more dif- ficult question in the area, as in most of dust and sediment samples from North African source areas, quartz was found to be the most dominant mineral, but the carbonate-content of dust is also relevant, what can be also in- teresting in case of Fuerteventura. The cal- cite content and consequently the (Ca+Mg)/ Fe ratio of western sources are among the highest in North Africa. The reason for high carbonate content of other mentioned sources is unclear at the moment, but it can be stated that almost all discussed emission regions had relatively high carbonate content. Varga, Gy. et al. Hungarian Geographical Bulletin 67 (2018) (2) 121–141.140 Acknowledgement: Support of the National Research, Development and Innovation Office NKFIH K120620 is gratefully acknowledged. REFERENCES Barkan, J., Alpert, P., Kutiel, H. and Kishcha, P. 2005. Synoptics of dust transportation days from Africa toward Italy and Central Europe. Journal of Geophysical Research. Atmospheres 110. D07208. Bridges, E.M. 1990. World Geomorphology. Cambridge, Cambridge University Press. Centeri, Cs., Szalai, Z., Jakab, G., Barta, K., Farsang, A., Szabó, Sz. and Bíró, Zs. 2015. Soil erodibility cal- culations based on different particle size distribution measurements. Hungarian Geographical Bulletin 64. (1): 17–23. Coude-Gaussen, G., Rognon, P., Bergametti, G., Gomes, L., Strauss, B., Gros, J.M. and Le Coustumer, M.N. 1987. Saharan dust on Fuerteventura Island (Canaries): Chemical and mineralogical characteristics, air mass trajectories, and probable sources. Journal of Geophysical Research 92. D8. Criado, C. and Dorta, P. 2003. An unusual‚ ‘blood rain’ over the Canary Islands (Spain). The storm of January 1999. Journal of Arid Environments 55. 765–783. Criado, C., Torres, J.M., Hansen, A., Lillo, P. and Naranjo, A. 2012. Intercalaciones de polvo sahariano en paleodunas bioclásticas de Fuerteventura (Islas Canarias). Cuaternario y Geomorfología 26. (1–2): 73–88. Engelstaedter, S., Tegen, I. and Washington, R. 2006. North African dust emissions and transport. Earth- Science Reviews 79. (1–2): 73–100. Faust, D., Yurena, Y., Willkommen, T., Roettig, C., Richter, Dan., Richter, Dav., von Suchodoletz, H. and Zöller, L. 2015. A contribution to the understand- ing of late Pleistocene dune sand-paleosol-sequences in Fuerteventura (Canary Islands). Geomorphology 246. 290–304. Ginoux, P.M., Chin, I., Tegen, I., Prospero, J., Holben, M., Dubovik, O. and Lin, S.J. 2001. Global simulation of dust in the troposhere: model description and assess- ment. Journal of Geophysical Research 106. 20255–20273. Glaccum, R.A. and Prospero, J.M. 1980. Saharan aerosols over the tropical North Atlantic: mineralogy. Marine Geology 37. 295–321. Goudie, A.S. and Middleton, N.J. 2006. Desert Dust in the Global System. Springer. Harrison, S.P., Kohfeld, K.E., Roelandt, C. and Claquin, T. 2001. The role of dust in climate changes today, at the last glacial maximum and in the future. Earth-Science Reviews 54. 43–80. Israelevich, P.L., Levin, Z., Jospeh, J.H. and Ganor, E. 2002. Desert aerosol transport in the Mediterranean region inferred from the TOMS aerosol index. Journal of Geophysical Research. Atmospheres 107: D21. Jackson, M.L., Levelt, T.W.M., Syers, J. K., Rex, R.W., Clayton, R.N., Sherman, G.D. and Uehara, G. 1971. Geomorphological relationships of tropospherically- derived quartz in the soils of the Hawaiian Islands. Soil Science Society of America, Proceedings 35. 515–525. Jacomy, M., Venturini, T., Heymann, S. and Bastian, M. 2014. ForceAtlas2, a Continuous Graph Layout Algorithm for Handy Network Visualization Designed for the Gephi Software. PLoS ONE 9. (6): e98679. Kohfeld, K.E. and Tegen, I. 2007. Record of Mineral Aerosols and Their Role in the Earth System. Treatise on Geochemistry 4. 1–26. Lim, J., Matsumoto, E. and Kitagawa, H. 2005. Eolian quartz flux variations in Cheju Island, Korea, during the last 6,500 yr and a possible Sun-monsoon linkage. Quaternary Research 64. 12–20. MacLeod, D.A. 1980. The origin of the red Mediterranean soils in Epirus, Greece. Journal of Soil Science 31. 125–136. Maher, B.A., Prospero, J.M., Mackie, D., Gaiero, D., Hesse, P.P. and Balkanski, Y. 2010. Global con- nections between eolian dust, climate and ocean biogeochemistry at the present day and at the last glacial maximum. Earth-Science Reviews 99. 61–97. Mahowald, N.M., Kohfeld, K., Hansson, M., Balkanski, Y., Harrison, S.P., Prentice, I.C., Schulz, M. and Rodhe, H. 1999. Dust sources and deposition during the last glacial maximum and current climate: a comparison of model results with paleodata from ice cores and marine sediments. Journal of Geophysical Research 104. 15895–15916. Mahowald, N.M., Muhs, D.R., Levis, S., Rasch, P.J., Yoshioka, M., Zender, C.S. and Luo, C. 2006. Change in atmospheric mineral aerosols in response to climate: Last glacial period, preindustrial, modern, and doubled carbon dioxide climates. Journal of Geophysical Research 111. D10202. Mee, A.C., Bestland, E.A. and Spooner, N.A. 2004. Age and origin of Terra Rossa soils in the Coonawarra area of South Australia. Geomorphology 58. 1–25. Menéndez, I., Díaz-Hernández, J.L., Mangas, J., Alonso, I. and Sánchez-Soto, P.J. 2007. Airborne dust accumulation and soil development in the North-East sector of Gran Canaria (Canary Islands, Spain). Journal of Arid Environments 71. 57–81. Menéndez, I., Pérez-Chacón, E., Mangas, J., Tauler, E., Engelbrecht, J.P., Derbyshire, E, Cana, L. and Alonso, I. 2013. Dust deposits on La Graciosa Island (Canary Islands, Spain): Texture, mineralogy and a case study of recent dust plume transport. Catena 117. 133–144. Miller, R.L., Tegen I. and Perlwitz, J. 2004. Surface radiative forcing by soil dust aerosols and the hydrologic cycle. Journal of Geophysical Research. Atmospheres 109. D04203. Muhs, D.R., Budahn, J.R., Prospero, J.M. and Carey, S.N. 2007a. Geochemical evidence for African dust 141Varga, Gy. et al. Hungarian Geographical Bulletin 67 (2018) (2) 121–141. inputs to soils of western Atlantic islands: Barbados, the Bahamas, and Florida. Journal of Geophysical Research 112. F02009. Muhs, D.R., Budahn, J.R., Reheis, M., Beann, J., Skipp, G. and Fisher, E. 2007b. Airborne dust transport to the eastern Pacific Ocean off southern California: Evidence from San Clemente Island. Journal of Geophysical Research 112. D13203. Pósfai, M. and Buseck, P.R. 2010. Nature and climate effects of individual tropospheric aerosol particles. Annual Review of Earth and Planetary Sciences 38. 17–43. Prospero, J.M. 1996. Saharan dust transport over the north Atlantic Ocean and Mediterranean: An overview. In The impact of desert dust across the Mediterranean. Eds.: Guerzoni, S. and Chester, R., Environmental Science and Technology Library 11, Dordrecht and London, Kluwer, 133–152. Prospero, J.M. and Lamb, P.J. 2003. African droughts and dust transport to the Caribbean: Climate change implications. Science 302. 1024–1027. Prospero, J.M., Bonatti, E., Schubert, C. and Carlson, T.B. 1970. Dust in the Caribbean atmosphere traced to an African dust storm. Earth and Planetary Science Letters 9. (3): 287–293. Prospero, J.M., Ginoux, P.M., Torres, O., Nicholson, S.E. and Gill, T.E. 2002. Environmental charac- terization of global sources of atmospheric soil dust identified with the Nimbus-7 Total Ozone Mapping Spectrometer (TOMS) absorbing aerosol product. Reviews of Geophysics 40. 1–31. Roettig, C-B., Kolb, T., Wolf, D., Baumgart, P., Richter, C., Schleicher, A., Zöller, L. and Faust, D. 2017. Complexity of aeolian dynamics (Canary Islands). Palaeogeography, Palaeoclimatology, Palaeoecology 472. 146–162. Roettig, C-B., Varga, G., Sauer, D., Kolb, T., Wolf, D., Makowsky, V., Recio Espejo, J.M., Zöller, L. and Faust, D. Characteristics, nature and formation of palaeosurfaces within dunes on Fuerteventura. Quaternary Research (in press) Scheuvens, D., Schütz, L., Kandler, K., Ebert, M. and Weinbruch, S. 2013. Bulk composition of northern African dust and its source sediments — A compila- tion. Earth-Science Reviews 116. 170–194. Shao, Y., Wyrwoll, K.H., Chappell, A., Huang, J., Lin, Z., McTainsh, G.H., Mikami, M., Tanaka, T.Y., Wangh, X. and Yoon, S. 2011. Dust cycle: An emerging core theme in Earth system science. Aeolian Research 2. 181–204. Stuut, J-B.W., Smalley, I. and O’Hara-Dhand, K. 2009. Aeolian dust in Europe: African sources and European deposits. Quaternary International 198. 234–245. Sun, D., Bloemendal, J. , Rea, D.K., An, Z., Vandenberghe, J., Lu, H., Su, R. and Liu, T.S. 2004. Bimodal grain-size distribution of Chinese loess, and its paleoclimatic implications. Catena 55. 325–340. Sun, D., Bloemendal, J., Rea, D.K., Vandenberghe, J., Jiang, F., An, Z. and Su, R. 2002. Grain-size distribu- tion function of polymodal sediments in hydraulic and aeolian environments, and numerical partition- ing of the sedimentary components. Sedimentary Geology 152. 263–277. Swap, R., Garstang, M., Greco, S., Talbot, R. and Kallberg, P. 1992. Saharan dust in the Amazon basin. Tellus B 44. (2): 133–149. Tegen, I., Lacis, A.A. and Fung, I. 1996. The influence of mineral aerosols from disturbed soils on climate forcing. Nature 380. 419–422. Tsoar, H. and Pye, K. 1987. Dust transport and the question of desert loess formation. Sedimentology 34. 134–153. Varga, Gy, Cserháti, Cs., Kovács, J. and Szalai, Z. 2016. Saharan dust deposition in the Carpathian Basin and its possible effects on interglacial soil formation. Aeolian Research 22. 1–12. Varga, Gy. 2012. Spatio-temporal distribution of dust storms - a global coverage using NASA Total Ozone Mapping Spectrometer aerosol measure- ments (1979–2011). Hungarian Geographical Bulletin 61. (4): 275–298. Varga, Gy., Kovács, J. and Újvári, G. 2012. Late Pleistocene variations of the background aeolian dust concentration in the Carpathian Basin: an es- timate using decomposition of grain-size distribu- tion curves of loess deposits. Netherlands Journal of Geosciences – Geologie en Mijnbouw 91. (1–2): 159–171. Varga, Gy., Kovács, J., Szalai, Z., Cserháti, Cs. and Újvári, G. 2018. Granulometric characterization of paleosols in loess series by automated static image analysis. Sedimentary Geology 370. 1–14. Varga, Gy., Újvári, G. and Kovács, J. Interpretation of sedimentary (sub)populations extracted from grain size distributions of Central European loess- paleosol series. Quaternary International (in press) von Suchodoletz, H., Kühn, P., Hambach, U., Dietze, M., Zöller, L. and Faust, D. 2009. Loess- like and palaeosol sediments from Lanzarote (Canary Islands/Spain). – Indicators of palaeoen- vironmental change during the Late Quaternary. Palaeogeography, Palaeoclimatology, Palaeoecology 278. (1–4): 71–87. Washington, R., Todd, M., Middleton, N.J. and Goudie, A.S. 2003. Dust-storm source areas deter- mined by the Total Ozone Monitoring Spectrometer and surface observations. Annals of the Association of American Geographers 93. (2): 297–313. Yaalon, D.H. 1997: Soils in the Mediterranean region: what makes them different? Catena 28. 157–169. Yaalon, D.H. and Ganor, E. 1973. The influence of dust on soils during the Quaternary. Soil Science 116. 146–155. Varga, Gy. et al. Hungarian Geographical Bulletin 67 (2018) (2) 121–141.142