Download pdf - Saharan dust storms: nature and consequences

Ž .Earth-Science Reviews 56 2001 179–204www.elsevier.comrlocaterearscirev

Saharan dust storms: nature and consequences

A.S. Goudie), N.J. MiddletonSchool of Geography and the EnÕironment, UniÕersity of Oxford, Mansfield Road, Oxford OX1 3TB, UK

Received 14 December 2000; accepted 14 May 2001

Abstract

This paper reviews recent work on the role of Saharan dust in environmental change, the location and strength of sourceareas, the transport paths of material away from the desert, the rates of Saharan dust deposition, the nature of that materialŽ .including PeriSaharan loess and the changing rates of dust activity in response to long and short-term climatic changes.The Sahara produces more aeolian soil dust than any other world desert, and Saharan dust has an important impact onclimatic processes, nutrient cycles, soil formation and sediment cycles. These influences spread far beyond Africa, thanks tothe great distances over which Saharan dust is transported. The precise locations of Saharan dust source areas are not well

Ž .known, but data from the Total Ozone Mapping Spectrometer TOMS suggest two major source areas: the Bodele´ ´depression and an area covering eastern Mauritania, western Mali and southern Algeria. Trajectories of long-distancetransport are relatively well documented, but the links between source areas and seasonal Saharan dust pathways are not.However, it is possible that Harmattan dust from the Bodele depression may not be the source of the prominent winter plume´ óver the tropical North Atlantic, as is often suggested in the literature. Few of the data on particle size characteristics ofSaharan dust are derived from major source areas or from Africa itself. Saharan dusts sampled from the Harmattan plumeand over Europe are dominated by SiO and Al O , a characteristic they share with North American and Chinese dusts. The2 2 3

concentrations of these two major elements are similar to those found in world rocks. PeriSaharan loess is conspicuous by itsrelative absence, considering the Sahara’s dominance of the global desert dust cycle both in the contemporary era andthrough the geological past. In recent decades, the frequency of Saharan dust events has varied markedly in response toclimatic factors such as drought and anthropogenic disturbance of desert marginal surfaces. Nonetheless, the Sahara’s twomajor dust sources are little affected by human activities and are in fact located in areas that receive very low rainfall totals.Hence, the Sahara does not fit the postulated global picture of a peak in dust storm activity in the 100–200-mm mean annualrainfall zone. q 2001 Elsevier Science B.V. All rights reserved.

Keywords: dust; Sahara; desert; geomorphology; atmospheric transport; TOMS

1. Introduction

The Sahara is the world’s largest source of aeo-Žlian soil dust Schutz et al., 1981; D’Almeida, 1987;¨

) Corresponding author. Tel.: q44-1865-271-921; fax: q44-1865-271-940.

Ž .E-mail address: andrew.goudie @geog.ox.ac.uk A.S. Goudie .

.Swap et al., 1996 and probably accounts for almosthalf of all the aeolian material supplied to the world’soceans. This large quantity of dust generation indi-cates the importance of the geomorphological pro-cesses of aeolian deflation and abrasion in mouldingthe landscape of parts of the Sahara. In this paper,we review recent work that has investigated the roleof Saharan dust in environmental change, the source

0012-8252r01r$ - see front matter q 2001 Elsevier Science B.V. All rights reserved.Ž .PII: S0012-8252 01 00067-8

( )A.S. Goudie, N.J. MiddletonrEarth-Science ReÕiews 56 2001 179–204180

areas from which the dust is derived, the strength ofthe Saharan dust source, the transport paths of mate-rial away from the desert, the rates at which Saharandust is being deposited, the nature of that materialŽ .including PeriSaharan loess and the changing ratesof dust provision in response to long and short-termclimatic changes.

2. Dust and environmental change

One environmental consequence of atmosphericdust loadings is their significance for climate througha range of possible influences and mechanisms. Theymay affect air temperatures through the absorption

Žand scattering of solar radiation Li et al., 1996;Moulin et al., 1997; Alpert et al., 1998; Miller and

.Tegen, 1998 . Saharan dust modifies short-wave so-lar radiation transmitted through to the earth’s sur-face and long-wave infra-red radiation emitted tospace. However, the balance between these two ten-dencies determines whether this creates cooling orwarming, and this in turn, depends in part upon suchvariables as the size distribution of dust particles and

Ž .their chemical composition. Fouquart et al. 1987found that either warming or cooling could takeplace in a Saharan dust event largely dependent uponthe numberrsize distribution of the mineral particlepopulation. Other important factors in this equationare cloud cover and the albedo of the underlying

Ž .surface Nicholson, 2000 . In the case of clouds,their altitude and optical depth are important deter-

Žminants of the direct radiative impact of dust Quijano.et al., 2000 .

It is also possible that dust may affect climatethrough its influence on marine primary productivityŽ .Jickells et al., 1998 , and there is some evidence

Žthat dust may cause ocean cooling Schollaert and.Merrill, 1998 . Changes in atmospheric temperatures

and in concentrations of potential condensation nu-clei may affect convectional activity and cloud for-

Žmation, thereby modifying rainfall amounts Bryson.and Barreis, 1967; Maley, 1982 . Dust particles in

storms reaching Israel from the Hoggar, for example,are typically coated with sea salts after their longjourney across the eastern Mediterranean, as well as

Ž .anthropogenic sulphates Levin et al., 1996 . Theseparticles are thought to act as ice and cloud conden-

sation nuclei, hence playing a role in the formationof rainfall.

Secondly, dust deposition provides considerablequantities of nutrients to ocean surface waters andthe seabed, particularly in the North Atlantic Ocean.

Ž .Talbot et al. 1986 have determined the percentageof a number of nutritive species in western African

Ž . Ž .aerosols Table 1A , while Swap et al. 1996 havecalculated the westward mass flux of these same

Ž .nutrients Table 1B . In addition to these four species,aeolian dust contains appreciable quantities of ironŽ .Zhu et al., 1997 , the addition of which to oceanwaters may increase plankton productivity by stimu-lating N2-fixation and thus acting to reduce nitrogenlimitation stresses on the phytoplankton communityŽ .Gruber and Sarmiento, 1997 . Saharan dust aerosolsinfluence the nutrient dynamics and biogeochemicalcycling of both terrestrial and oceanic ecosystems.Moreover, because of the thousands of kilometersover which the dust is transported, its influence

Žextends as far afield as Northern Europe Franzen et. Ž .al., 1994 , Amazonia Swap et al., 1992 , and the

coral reefs of the Caribbean. In the case of Caribbeancoral reefs, Saharan dust has also been implicated by

Ž .Shinn et al. 2000 as an efficient substrate fortransporting disease-spreading spores, which on oc-casion, can cause Caribbean-wide epidemics thatdiminish coral reef vitality, a good match having

Table 1Nutrient contributions from the Sahara to the North Africanaerosol

Ž .A % Contribution

Mean Minimum MaximumqK 0.17 0.1 0.33

qNH 0.18 0.01 0.274yNO 0.51 0.21 1.533yPO 0.037 0.017 0.0714

Ž .after Talbot et al., 1986; Swap et al., 1996

Ž . Ž .B Westward mass flux mean values for 1989–199211Ž .10 gryear

Mean Minimum MaximumqK 4.9 2.9 9.5

qNH 5.2 0.3 7.84yNO 14.7 6.1 43.133yPO 1.1 0.5 2.14

Ž .modified from Table 4B in Swap et al., 1996

( )A.S. Goudie, N.J. MiddletonrEarth-Science ReÕiews 56 2001 179–204 181

been found between times of coral-reef die-off andpeak dust deposition.

Thirdly, dust loadings may change substantially inresponse to climatic changes, such as the North

ŽAtlantic Oscillation and to drought phases Moulin et.al., 1997; Middleton, 1985; Littmann, 1991a and in

Žresponse to land cover changes Tegen and Fung,.1995 .

Fourthly, additions of dust to land surfaces mayaffect soil formation. This has been proposed in the

Ž .context of West Africa by Vine 1987 and for theformation of terra rossa soils in southern Europe and

Žthe Levant Yaalon and Ganor, 1973; MacLeod,.1980; Rapp, 1984 . It may also contribute to the

character of soils on Caribbean and West AtlanticŽ .Islands Muhs et al., 1990 . Dust that has a high

carbonate content may be a factor in the formation ofcalcretes.

Fifthly, dust additions play a major role in thedelivery of sediments to the oceans. For example,

Ž .Guerzoni et al. 1999, p. 147 have suggested thatABoth the magnitude and the mineralogical composi-tion of atmospheric dust inputs indicate that eolian

Ž .deposition is an important 50% or even dominantŽ .)80% contribution to sediments in the offshorewaters of the entire Mediterranean basin.B

3. Source areas of Saharan dust

One of the most important advances required tounderstand the production of Saharan dust is to beable to identify the major source areas. Herrmann

Ž .et al. 1999 attempted to survey previous studiesŽ .of Saharan dust sources, but remarked p. 142 ,

AAstonishingly, the results with regard to sourceareas are totally different. No overlapping can bedetected which could serve as a confirmation ofresults.B One explanation for this unsatisfactory situ-ation is the range of source identification methodsthat have been used by different workers: remotesensing, analysis of surface dust observations, backtrajectory analysis of isobar data, and the use ofmineral tracers.

However, in recent years, some progress has beenmade in identifying source areas by measurements ofinfra-red radiances such as those acquired by ME-TEOSAT. These can be used to produce the Infra-red

Ž . ŽDifference Dust Index IDDI Brooks and Legrand,.2000 . This method has highlighted the Bodele De-´ ´

pression between Tibesti and Lake Chad as an im-portant source region, together with a large swathe ofcountry covering portions of Mauritania, Mali andsouthern Algeria. It also suggests that the Horn ofAfrica and the Nubian Desert in southern Egypt andNorthern Sudan are important sources. The impor-tance of the Bodele region was also shown by Kalu´ ´Ž . Ž .1979 and Herrmann et al. 1999 , but the status ofthe other regions is less clear.

Another recently developed method to detect dustsource regions is the Total Ozone Mapping Spec-

Ž .trometer TOMS , which is currently carried on boardŽthe Earth Probe Satellite Middleton and Goudie, in

.press . TOMS can detect UV-absorbing aerosols inthe atmosphere from the spectral contrast betweenthe 340- and 380-nm channels, and an Aerosol IndexŽ .AI has been developed which is linearly propor-

Žtional to the aerosol optical thickness Herrmann et.al., 1999; Chiapello et al., 1999; Hsu et al., 1999 .

AI values indicate the intensity of dust content, nottotal dust flux.

Ž .The TOMS data Fig. 1 confirm that the Bodele´ ´Region is the most intense source region not only inthe Sahara, but also in the world, with AI values thatexceed 30. It also demonstrates the presence of a

Ž .large but less intense area AI values over 24 in theWest Sahara. This extends through to the Atlanticcoast of Mauritania. Relatively high AI values arealso observed in the interior of Libya.

The importance of Bodele as a dust source relates´ ´to various factors. Firstly the region is very dryŽFaya Largeau receiving an average annual rainfall

.of just 17 mm , but is fed with silty alluvium bystreams draining from the Tibesti Massif. There mayalso be susceptible silty materials that were laiddown in an expanded Lake Chad during earlyHolocene and Pleistocene pluvials, and Mainguet

Ž .and Chemin 1990 have argued that deflational ac-tivity downwind from Tibesti may be substantial andhelp to explain the excavation of Lake Chad itself.The reasons for the importance of the West Saharandust source in Mali, Mauritania and Algeria are lesswell-understood. However, it is an area of low reliefbounded on the north and east by uplands. Whilesuch upland areas are not themselves major dustsource regions, wadis draining from them will have


Fig. 1. Annual mean Aerosol Index for the Sahara, derived from TOMS.

transported silt-rich alluvium into the area. Likewise,in the past, the southern part of the region may havereceived alluvial inputs from the Niger river prior toits capture by southeast-trending drainage near

Ž .Tosaye Urvoy, 1942 . It also contains an enormousclosed depression some 900 km long and variousergs that could provide a dust source through win-nowing. The depression contains many ancientlakebeds that show signs of intense deflation in the

Ž . Ž .Holocene Petit-Maire, 1991 . Dubief 1953 maps itas an area of high aeolian activity, and it is also

Table 2Estimates of the source strength of the Sahara

Ž .Author s Annual quantityŽ .millions of tonnesryear

Ž .Jaenicke 1979 260Ž .Schutz et al. 1981 260¨

Ž .Prospero 1996a,b 170Ž .Swap et al. 1996 130–460Ž .d’Almeida 1986 630–710

Ž .Marticorena and Bergametti 1996 586–665Ž .Callot et al. 2000 760

rather dry, with annual precipitation levels of be-tween 5 and 100 mm.

4. The strength of the Saharan dust source

Various attempts have been made to estimate andmodel the source strength of the Sahara using dataon mineral loadings in the atmosphere, surface mate-

Ž .rial characteristics Callot et al., 2000 and transportŽ .models Table 2 . The estimates show a wide range

of values that may reflect differences in modelling

Table 3Estimates of dust emissions to the atmosphere at a global scale

Ž .Author s Annual quantityŽ .millions of tonnes

Ž .Schutz 1980 up to 5000¨Ž .Peterson and Junge 1971 500

Ž .Andreae 1995 1500Ž .Duce 1995 1000–2000

Ž .d’Almeida 1986 1800–2000Ž .Tegen and Fung 1994 3000


Table 4Maximum mean AI values for major global dust sources deter-mined from TOMS

Location

Bodele Depression of Central Sahara )30´ ´West Sahara in Mali and Mauritania )24

Ž .Arabia Southern OmanrSaudi border )21Ž .Eastern Sahara Libya )15Ž .Southwest Asia Makran coast )12

TaklamakanrTarim basin )11Ž .Etosha Pan Namibia )11

Lake Eyre Basin )11Ž .Mkgadikgadi Basin Botswana )8

Ž .Salar de Uyuni Bolivia )7Great Basin of the USA )5

procedures, differences in the time scales considered,and differences in the areal extent of the source.

There are few data available which allow a com-parison with other major source areas. An exception

Ž .to this is provided by the work of Zhang et al. 1997on the Taklamakan Desert. For this region, they

estimate an annual dust production of 800 milliontonnes. On this basis, they propose that this may bearound half of the global production of dust.

Estimates of soil dust emissions to the atmosphereŽ . Žon a global scale Table 3 show a large range see

.the excellent review by Prospero, 1996a , largelybecause of differences in the assumptions made indifferent models with regard to such factors as therate of scavenging of particles from the air. Thismakes it difficult to evaluate the contribution madeby the Sahara.

An alternative method that can be used to com-pare relative source strengths is the TOMS data. Bylooking at the Aerosol Index intensity and its arealextent, it is possible to gain an indication of thepredominance of the Sahara in comparison with otherdesert areas.

As Table 4 shows, three of the world’s four mostimportant dust sources occur within the Sahara, andFig. 2 shows the area and intensity of the SaharanAerosol Index compared with those for Arabia, NWChina and the Thar. Fig. 2 confirms the importanceof the Sahara on the global scale.

Ž .Fig. 2. The extent and intensity of the Aerosol Index AI derived by TOMS for four major desert areas: Sahara, Arabia, China and Thar.Ž 2 3.The figure shows the areas in km =10 covered by different intensities of the AI.


5. Trajectories of Saharan dust transport

Saharan dust is regularly transported from itssource areas along three main transport paths: west-

Ž . Žward over the North Atlantic Ocean NAO Carlson.and Prospero, 1972; Moulin et al., 1997 , to North

Ž .America Perry et al., 1997 and South America

Ž .Swap et al., 1992 ; northward across the Mediter-Ž .ranean Loye-Pilot et al., 1986 to southern Europe¨

Ž .Avila et al., 1997; Rodriguez et al., 2001 andŽsometimes as far north as Scandinavia Franzen et

.al., 1994 ; and along easterly trajectories across theŽeastern Mediterranean Herut and Krom, 1996; Kubi-

. Žlay et al., 2000 , to the Middle East Ganor et al.,

Table 5Ž .Local names for Saharan dust-bearing winds after Middleton, 1986

Name Area affected Season Direction Meteorological ReferenceŽ Ž .derivation from conditions

.where known

Ž .Irifi Western Sahara SE Frontal Morales 1946Ž .Ghibli Tripolitania Pre-frontal Sivall 1957

Žfree translation: ‘wind.from south Mecca’

Guebli Tunisia and Algeria All year but S Pre-frontal with Naval IntelligenceŽ . Ž . Ž .south wind northern parts most prevalent katabatic effects Division 1943

May–October from interioruplands tocoastal plains

Ž .Sahel Morocco SW Frontal Mainguet 1980Ž Ž .Harmattan Fantee: BilmarFaya Largeau October–April ENE Pressure surge after Kalu 1979

‘aharaman’ to blow area plus much cold air outbreaksand ‘ta’ grease of West Africa from mid-latitudeslocally used to south of 208N

.cover skinŽ .Brume seche West Africa October–April Harmattan haze Bertrand et al. 1979`

Ž .French: ‘dry haze’ in light windsŽ Ž Ž .Haboob Arabic: Sudan but has May–July Single cell Freeman 1952.‘to blow’ become almost thunderstorm

.generic in its use downdraftŽ Ž .Khamsin Arabic: Egypt Spring Pre-frontal Fisher 1978

a.‘fifty’Chili Tunisia and Algeria Spring SW Pre-frontal Naval Intelligence

Ž . Ž .southern parts Division 1943Ž .Shekheli Algeria Spring Borushko 1972

Chergui Moroccan Sahara Summer NE Naval IntelligenceŽ .Division 1943Ž .Dschani Southern Sahara Goudie 1978Ž .Kharif Somalia June–September SW Brooks 1920Ž .Gobar Ethiopia Goudie 1978

Ž .Sirocco Southern Europe Spring S Frontal Conte et al. 1996Ž .Leveche Spanish Spring SE–SW Frontal Tout and Kemp 1985

Mediterraneancoast Malaga–Alicante

Ž .Leste Madeira Frontal Goudie 1978Ž .Levanto Canary Islands Nalivkin 1983

a Ž . Ž .Variously taken to refer to the average duration of the wind 50 h , its annual frequency 50 times and its season of maximum onsetŽ .the 50 days either side of the spring solstice .


.1991 . The winds that entrain and transport this dustare regular features of local climates and have a

Ž .large number of local names Table 5 .

6. North Atlantic trajectories

The westward flow of material over the NAO isthe most voluminous, accounting for 30–50% of

Ž .output Schutz et al., 1981; D’Almeida, 1986 . Nu-¨merous papers have documented the transport anddeposition of Saharan dust to distant regions of the

ŽNAO and to the Americas see Duce, 1995; Pros-.pero, 1996a for reviews . Large dust outbreaks dur-

ing the summer appear to be associated with strongconvective disturbances that develop over WestAfrica at about 15–208N and move westward, carry-ing material entrained in Saharan and Sahelian lati-tudes. Resultant dust plumes over the NAO areusually associated with easterly waves that emergefrom the African coast every 3–4 days. Their com-plex structure produces intricate distribution patterns,including northward branches that can transport ma-terial to Western Europe. Remote sensing imagesover the NAO have also demonstrated the impor-tance of the development of the Azores–Bermudahigh-pressure system in summer in drawing dust-laden air from the tropical North Atlantic into the

Ž .subtropical region Jickells et al., 1998 .Saharan dust outbreaks over the NAO commonly

persist for several days and can last for tens of daysŽ .Carlson, 1979 . The longer-lived plumes transportmaterial the furthest. Transport to the Caribbean,where an estimated 20 million tonnes of Saharan

Ž .dust is deposited annually Schlatter, 1995 , typicallyŽ .takes 5–7 days Prospero and Carlson, 1981 . The

duration of individual Saharan dust events monitoredat Trinidad in the West Indies can vary between 3and 5 days, and sometimes back-to-back episodes

Žcan last as long as 20 days Rajkumar and Chang,.2000 . The latitudinal pathways of transatlantic

transport vary with the seasons. Hence, maximumSaharan aerosol concentrations monitored at Barba-

Ždos and Miami are in July and August Prospero and.Carlson, 1981; Prospero et al., 1987; Prospero, 1999 ,

while the highest concentrations monitored atŽ . ŽCayenne Prospero et al., 1981 are in March Fig.

.3 . Sal Island lies in a zone that is affected by both

of these seasonal pathways, displaying a bimodalŽ .peak March and AugustrSeptember in atmospheric

Ž .turbidity Schutz, 1979 . TOMS analysis shows this¨clear seasonal pattern of dust export over the NAOŽ .Table 6 . The zone of dust export is most intensebetween 108 and 258N, but it migrates seasonally. InJFM, the zone of maximum AI is between 58 and108N. By AMJ, it is between 108 and 208N, whereasin JAS, it is between 158 and 258N. By OND, asouthward retreat has begun, but dust export is rela-tively modest in amount. This seasonal pattern iscomparable to that obtained from AVHRR aerosol

Ž .optical thickness data Swap et al., 1996 and fromship observations of haze made prior to the 1930sŽ .McDonald, 1938 .

Specific sources for transatlantic dust plumes arenot well known. They are perhaps most likely to bein the Mauritania–Mali area, and further north inWestern SahararSouthern Morocco, although theclear seasonal signals found in dust concentrationson the western side of the Atlantic are not simplyrelated to the seasonality of dust events recorded on

Ž .the West African coast Fig. 3 . At Nouakchott, duststorms are a feature of the first 6 months of the year,before the annual rains, and hence it is unlikely thatthis station lies in the pathway of the strong summerflow that reaches Miami and Barbados. Further north,dust event frequencies at Nouadhibou and Dahklaare much less obviously seasonal, although the monthof maximum dust activity at both stations is Febru-ary. Harmattan dust blown from the Bodele depres-´ ´sion tends not to travel far over the Gulf of Guinea,as it is efficiently scavenged by rainfall associatedwith the Intertropical Convergence Zone, which typi-

Žcally descends no further that 58N Afeti and Resch,.2000 . The implication in Table 6 and various re-

Žmote sensing studies e.g. Swap et al., 1996; Husar.et al., 1997 that dust reaching South America may

be from the Bodele depression is investigated further´ ín Fig. 4 using data from the only period, in 1977–1979, when regular monitoring of mineral dust was

Ž .carried out at Cayenne Prospero et al., 1981 . Thelink with Bodele dust is not confirmed by comparing´ ´the seasonality of mineral dust concentrations insurface level air at Cayenne with that of thick dusthaze at Maiduguri in Northern Nigeria, a stationdirectly within the Harmattan trajectory. A better,though still far from complete, link to potential


Fig. 3. Seasonality of dust events over the North Atlantic and the west coast of Africa.

sources is made with Nouakchott. However, thenumber of sites and the length of data considered aretoo limited for firm conclusions to be drawn and it

Table 6ŽDust over the North Atlantic from 1999 TOMS data percentage

.of days with AI)19

Ž .Latitude 8N Season

JFM AMJ JAS OND

45–50 0 0 0 040–45 0 0 0 035–40 0 0 0 030–35 0.44 0 2.24 025–30 0.22 8.64 9.78 020–25 0 24.58 29.52 0.6815–20 0.68 33.10 29.12 2.2610–15 3.44 30.00 8.7 0.685–10 11.9 11.7 1.32 0.440–5 3.44 0.68 0 0.22

may be that a combination of sources contributes tothe Cayenne record.

7. European trajectories

Saharan dust is often deposited in precipitationover southern Europe and has been reported since

Ž .ancient times Bucher and Lucas, 1984 . Less fre-¨quently, deposition occurs further north, on the

Ž . ŽBritish Isles Wheeler, 1986 , the Netherlands Reiff. Ž .et al., 1986 , Germany Littmann, 1991b and North-

Ž .ern Scandinavia Franzen et al., 1994 . Individualevents can be large, such as the dustfall in March1991, which covered at least 320,000 km2, stretching

Žfrom Sicily to Sweden and Finland Burt, 1991a;.Bucher and Dessens, 1992; Franzen et al., 1995 .¨

The Saharan source strength for dust transport toEurope was estimated at 80–120 million tonnesryear


Ž . Ž . Ž .Fig. 4. A Monthly mean mineral dust concentrations at Cayenne, French Guiana after Prospero et al., 1981 . B Monthly numbers ofŽ .days with thick dust haze at Maiduguri, Nigeria. C Monthly numbers of dust storm days at Nouakchott, Mauritania.

Ž .by D’Almeida 1986 , based on sunphotometer read-ings taken in the early 1980s, while the annualaeolian flux to the Western Mediterranean basin hasbeen put at 3.9 million tonnes by Loye-Pilot et al.¨Ž .1986 who extrapolated from their monitoring ofdeposits at Corsica. A major source area for transportto Western Europe was identified in southernmostAlgeria, between Hoggar and Adrar des Iforhas.

Ž .Another source Molinaroli, 1996 , is in WesternSahara–Southern Morocco. These sources have beenconfirmed by back trajectory analysis for dust de-

Ž .posited over Northeastern Spain. Avila et al. 1997traced deposition events back to three main areas:Western Sahara, the Moroccan Atlas, and centralAlgeria. These sources have also been identified for

Žtransport to the British Isles Tullet, 1978; Wheeler,


.1986 . A common trajectory for transport to BritainŽis over the Bay of Biscay Wheeler, 1986; Coude-´

.Gaussen et al., 1988 .Transport to southern Europe is more frequent,

however. A year of monitoring on Corsica, for ex-Žample, revealed 20 dust events Bergametti et al.,

.1989 originating in three source areas: eastern Alge-riarTunisiarWestern Libya; MoroccorWestern Al-geria, and Asouth of 308NB.

8. Eastern Mediterranean trajectories

Transport from North Africa to the EasternMediterranean occurs predominantly during springand is commonly associated with the eastward pas-sage of frontal low-pressure systems. Dust fromsources in the Middle East is more typically trans-

Žported to the Mediterranean in the autumn Dayan,.1986; Kubilay et al., 2000 . Analysis of 23 heavy

dustfalls in Israel over a 20-year-period suggest thatthe North African type is by far the most commonŽ .Ganor et al., 1991 . These storms are usually associ-ated with a cold front with a significant downwardflowing jet stream and are often accompanied by rainŽ .Alpert and Ganor, 1993 . Long-range transport ofSaharan dust to the central Mediterranean is charac-terised by events lasting 2–4 days, compared to anaverage duration of just 1 day for events reaching theEastern Mediterranean from the Arabian DesertŽ .Dayan et al., 1991 .

Central Algeria is the most frequent source areaŽ .for Saharan dust, reaching Israel Ganor et al., 1991 ,

Ž .and Ganor and Foner 1996 distinguish betweenmaterial commonly transported from sources in theHoggar Massif and the Tibesti mountains in North-ern Chad, the latter also picking up material from theWestern and Sinai Deserts.

9. Rates of dust deposition

Estimates of rates of dust deposition exist for anumber of sites at varying distances from the Sahara.These are presented in Table 7.

As might be expected, there is a tendency forrates to be lowest at large distances from potential

Table 7Dust deposition amounts

Source Location Annualdeposition

y2Ž .g m

Ž .Nihlen and Olsson 1995 Aegean Sea 11.2–36.5Ž .Le-Bolloch et al. 1996 Southern Sardinia 6–13

Wagenbach and Swiss Alps 0.4Ž .Geis 1989

De Angelis and French Alps 0.2Ž .Gaudichet 1991Ž .Avila et al. 1996 NE Spain 5.1

Ž .Bergametti et al. 1989 Corsica 12Ž .Loye-Pilot et al. 1986 Corsica 12.5¨Ž .Bucher and Lucas 1984 Central France 1¨

Ž .Pye 1992 Crete 10–100Ž .Herut and Krom 1996 Israeli coast 72Ž .Herut and Krom 1996 SE Mediterranean 36

Measures and Gulf of Guinea 3.4–11.5Ž .Brown 1996Ž .Maley 1982 South Chad 109

McTainsh and Northern Nigeria 137–181Ž .Walker 1982

Ž .Drees et al. 1993 SW Niger 200

Žsources. Thus the values for Western Europe e.g.. y2Central France and the Alps are less than 1 g m .

Further south, in NE Spain, a value of 5.1 g my2 isrecorded, while over Sardinia, Corsica, Crete and theSE Mediterranean, most values are between 10 and40 g my2 . On the south side of the Sahara, values inareas close to Harmattan source regions have valuesaround 100 to 200 g my2 , but they decline to lowvalues over the Gulf of Guinea.

However, given the absence of long-term directmeasurements of dust deposition over large areas,not least over the oceans, estimates of dust deposi-

Žtion have been gained by modelling Prospero,. Ž .1996a , using dust concentration data Fig. 5 . The

model indicates deposition rates for the Mediter-ranean of 3–14 g my2 per year, which are compara-ble to those obtained from direct measurements. Thehighest values in the model are for the 108 box at10–208N and 20–108W, with a value of 30.8 g my2 .

Ž .Schutz et al. 1981 have modelled the annual¨mass budget of dust transported from the Saharaover the Atlantic in the Northeast trade wind zoneŽ . ŽFig. 6 . A high rate of deposition up to 20 cm per

.1000 years occurs over the first 2000 km, whereas


y2 3 Ž .Fig. 5. Annual aerosol deposition rates is gm =10 over the North Atlantic Ocean derived from data in Prospero, 1996a, Table 2B .

Ž . Ž .Fig. 6. Schutz et al.’s 1981 model of dust transport and accumulation over the North Atlantic offshore from the Sahara. A Annual mass¨6 Ž .budget at 15–248N in 10 tonnesryear. B Accumulation rate of dust in deep-sea sediments in centimeters per 1000 years. Source:

modified after Schutz et al., 1981, Figs. 8 and 9.¨


when most of the mass of dust plume has fallen outŽ .at distances greater than 2000 km , a zone of com-

Žparatively low accumulation rates 1–2 cm per 1000.years occurs.

10. Particle size characteristics

The particle size characteristics of Saharan dustare summarised in Table 8. It needs to be noted,however, that nearly all the determinations are fordust storms that are not from major source areas andwhich have travelled outwards into the moister partsof West Africa, to the Atlantic, the Mediterranean orto Europe. It is likely, therefore, that dust stormsfrom near the source will have coarser grain sizecharacteristics than those listed. Mean modal andmedian sizes of the travelled dust tend to be fine siltbetween 5 and 30 mm in diameter, though Harmattan

Ž .dust at Kano Nigeria may have a median diameterŽ .that reaches 74 mm McTainsh and Walker, 1982 ,

Žwhile that from Tanezrouft reaches 72 mm Coude-´.Gaussen, 1981 . Conversely, samples from southern

Ghana, Barbados, Bermuda, the USA and parts ofEurope may be less than 5 mm. Although data aresparse, dust storms may transport substantial amounts

Ž .of clay sized material -2 mm .Although the modal data presented here are use-

ful, they provide little information about the maxi-mum sizes of grain that can be transported in dust

Ž .storms derived from the Sahara. Schroeder 1985found aggregated dust particles up to 150 mm indiameter in samples taken on the coastal belt of

ŽSudan, while samples taken over Sal Island Cape.Verde Islands off West Africa have yielded individ-

ual quartz grains up to 90 mm in diameter and micaŽflakes up to 350 mm in diameter Glaccum and

. Ž .Prospero, 1980 . Prospero et al. 1970 detected indi-Ž .vidual large particles )20 mm in diameter that

were carried more than 4000 km from their SaharanŽ .source, and Arimoto et al. 1997 recorded particles

43–57 mm in diameter at Bermuda. Saharan dustcollected after numerous fallout events over theBritish Isles has shown that large numbers of so-

Ž .called ‘giant’ dust particles )62.5 mm are com-monly carried more than 3000 km to Northern Eu-

Ž .rope Middleton et al., in press .

Table 8Particle size characteristics of dust

Source Location Modal, mean or % clay -2 mmŽ .median size mm

Ž . Ž .McTainsh and Walker 1982 Kano, Nigeria 8.9–74.3 median 2.3–32.0Ž . Ž . Ž .Coude-Gaussen 1981 Tanezrouft central Sahara 72 modal 9.4´Ž . Ž .Coude-Gaussen 1991 Maghreb 5–40 median –´

Ž . Ž .Mattsson and Nihlen 1996 Crete 8–30 modal –´Ž . Ž .Sala et al. 1996 Spain 4–30 mean –

Ž . Ž .Ratmeyer et al. 1991 Sal Island 11.9–18.6 mean –Ž . Ž .Littmann 1991a,b West Germany 2.2–16 median –

Ž . Ž .Pye 1992 Crete 4–16 median 15–45Ž . Ž .Gillies et al. 1996 Mopti, Mali 16.8 modal –

Ž . Ž .Ozer et al. 1998 Genoa, Italy 14.6 median –Ž . Ž .Bucher and Lucas 1984 SW France 4–12.7 median –¨

Ž . Ž .Coude-Gaussen 1991 South of France 8–11 median –´Ž . Ž .Tomadin et al. 1984 Central Mediterranean 2–8 modal –

Ž . Ž . Ž .Coude-Gaussen et al. 1988 Paris Basin France 8 modal –´Ž . Ž .Wagenbach and Geis 1989 Swiss Alps 4.5"1.5 median –

Ž . Ž .Talbot et al. 1986 Barbados 3.2 median –Ž . Ž .Arimoto et al. 1997 Bermuda 2.0–2.3 mean –Ž . Ž .Afeti and Resch 2000 Southern Ghana 1.16 mean –

Ž . Ž .Perry et al. 1997 Continental USA -1.0 mean –

()

A.S.G

oudie,N.J.M

iddletonr

Earth-Science

ReÕiew

s56

2001179

–204

191

Table 9Major element analyses of dust

Ž .Harmattan dust S Sahara Saharan dust over Europe Harmattan European Chinese Arizona World WorldŽ . Ž . Ž . Ž .mean mean mean 10 mean 10 mean 10 rocks 111 2 3 4 5 6 7 8 9

SiO 66.03 57.19 59.05 57.45 65.04 61.33 51.79 54.83 58.0 60.95 56.49 60.26 57.92 59.9 58.932

Al O 11.08 12.11 11.32 10.64 9.97 15.52 12.79 16.33 11.0 11.02 13.91 11.40 12.21 14.13 14.982 3

Fe O 4.45 5.30 4.63 4.34 3.78 8.06 5.32 6.09 6.0 4.50 6.37 2.91 4.72 6.85 6.12 3

FeO – – – – – – – – – – – 1.37 – –MgO 0.82 0.81 0.75 0.81 0.62 2.84 3.86 2.90 2.7 0.76 3.08 – 3.01 2.60 3.81CaO 0.13 3.61 3.01 2.88 1.90 3.47 12.19 10.15 8.6 2.31 8.60 – 2.01 3.94 4.84Na O 0.91 1.46 1.30 2.14 1.12 0.81 1.16 0.98 1.6 1.39 1.14 1.72 1.93 – –2

K O 2.04 2.95 2.87 3.26 2.95 3.26 3.26 2.18 1.8 2.81 2.63 2.13 2.63 2.35 2.992

TiO 0.73 0.83 0.81 0.82 0.92 0.74 1.01 1.22 1.2 0.82 1.04 0.65 0.74 – –2

P O 0.17 0.25 0.22 0.18 0.18 0.18 0.42 0.13 – 0.20 0.24 0.19 – – –2 5

MnO 0.10 0.08 0.08 0.09 0.08 – – 0.05 1.6 0.09 – – – – –SO – – – – – – – – – – – 0.20 – – –3

CO – 4.99 5.47 6.38 4.18 – – – – 5.26 – – – – –2

H O – 9.74 8.94 9.00 7.30 – – – – 8.75 – 0.80 2.14 – –2

LO1 12.79 – – – – – – – – – – – 11.64 – –Total 99.25 99.32 98.45 97.99 98.04 – – – – – – – – – –

Ž .1sKano McTainsh and Walker, 1982 .Ž .2sKano Wilke et al., 1984 .Ž .3sKano Wilke et al., 1984 .Ž .4sZaria Wilke et al., 1984 .Ž .5sZaria Wilke et al., 1984 .

Ž .6s ItalyrCentral Mediterranean Tomadin et al., 1984 .Ž .7s ItalyrCentral Mediterranean Tomadin et al., 1984 .

Ž .8sPyrenees Bucher and Lucas, 1984 .¨Ž .9sEurope Bucher, 1986 .¨Ž .10sGoudie 1978 .Ž .11sClarke 1916 .


11. Major element characteristics

In Table 9, we present the major element concen-trations for Saharan Dust as sampled in the southernSahararSahel from the Harmattan source and overEurope. For comparison, figures are given for Chi-

Ž .nese and North American Arizona dust and for duststorms on a global basis, together with the meancomposition of the Earth’s surface rocks.

What emerges from these data is that both Har-mattan and European dusts are dominated by SiO2

and Al O , a characteristic they share with North2 3

American and Chinese dusts. The concentrations ofthese two major elements are similar to those foundin world rocks. The dominance of SiO probably2

reflects the importance of quartz in aeolian dust.Saharan dust also appears to contain appreciablequantities of Fe O , MgO and CaO, though Harmat-2 3

tan dust is less rich in MgO and CaO than Saharandust transported northwards to Europe. The CaCO3

content of dust from North African sources has beenrecognised for its influence in increasing the pH of

Ž .rainfall in Corsica Loye-Pilot et al., 1986 and at¨¨Ž .Erdemli in Turkey Ozsoy and Saydam, 2000 , while

Saharan dust has also been identified as an importantsource of atmospheric P, mainly insoluble, to the

Ž .Mediterranean Mignon and Sandroni, 1999 . Over-all, the gross composition of Saharan dust appears tobe comparable with that of other regions and to thegross mean composition of the world’s rocks.

12. Clay mineralogy of Saharan dust

There are now available a large number of studiesof the clay mineralogy of Saharan dust and theseshow major geographical variations in the propor-tions of different clay minerals derived from differ-ent source areas.

Ž .Caquineau et al. 1998 detected variations in theclay minerals present in dust collected at Sal Islandon the basis of the different source areas from whichthe dust was derived. Dust originating from theNorth and West Sahara exhibited the highest amountof illite, whereas kaolinite became predominant whenair mass trajectories indicate a Sahelian origin.Kaolinite was dominant in dust originating from the

south and central Sahara, though the amount of illitecould not be detected.

Such a latitudinal variation in clay mineralogy isconsistent with the observations of Chester et al.Ž .1972 along the coast of Western Africa from 258Nto 308S. Kaolinite concentrations increased towardsthe Equator, whereas illite decreased. Dust samplescollected from the Niger and Northern Nigeria also

Ždisplayed a predominance of kaolinite Drees et al.,1993; Wilke et al., 1984; McTainsh and Walker,

.1982 .Along a transect in the Sahara from 19–358N,

Ž .Paquet et al. 1984 identified four different groupsor sectors. In Northern Algeria, illite and chloriteaccounted for around 70–75% of the clay content,kaolinite about 15% and attapulgite 10–15%. Furthersouth, around Beni Abbes and In Salah, attapulgitereached levels of 20–25%. Even further south, aroundTamanrassett, Tessalit and In Guezzam, illite and

Ž .chlorite were dominant 60–70% , attapulgite wasonly 5–10% and kaolinite 25–30%. South of Hoggarand in the Tanezrouft smectites were dominant, fol-

Ž . Ž .lowed by kaolinite 20–25% , illite 10–25% , atta-Ž . Ž .pulgite 10–15% and chlorite 5% . They attributed

this variability to the nature of the Quaternary sedi-ments and bedrock of the sectors concerned. Forexample, the sediments of the northernmost zonegain some of their characteristics as a result of the

Ž .deflation of the inland basins Chotts , while highkaolinite contents may be derived from ancient lat-eritic weathering profiles.

Ž .Sarnthein et al. 1982 also drew a distinctionbetween northern and southern source areas. Dust

Ž .from the South Sahara and Sahel south of 20–258Nis less rich in carbonate but richer in kaolinite andmontmorillonite, whereas in the North and Central

Ž .Sahara, carbonate contents are higher up to 20–50%and the dominant clay minerals are illite, chlorite,palygorskite and montmorillonite. Palygorskite hasalso been recognised as a characteristic mineral of

Ž .Saharan dust Coude-Gaussen and Blanc, 1985´reaching Sardinia and the Western MediterraneanŽ .Molinaroli, 1996 , in dust falling on Skye, Western

Ž .Scotland Bain and Tait, 1977 , and in dust reachingŽ .the central Mediterranean Tomadin et al., 1984 . At

the eastern end of the Mediterranean, kaolinite is amore significant aeolian clay mineral input, itsAfrican origin indicated by the northward decreasing


Žabundance gradient in marine sediments Foucault.and Melieres, 2000 . Dust reaching Northeastern´ `

Spain from the Northern Sahara had the followingclay minerals: illite and smectite and palygorskite

Ž .and kaolinite Avila et al., 1996 .

13. PeriSaharan loess

Although loess, by definition a wind-depositeddust with a median grain size range of 20–30 mmŽ .Tsoar and Pye, 1987 , has been estimated to cover

Ž .up to 10% of the world’s land area Pesci, 1968 ,and is widespread in many parts of the world, itsoccurrence in Africa is very limited. This appearssurprising given that the Sahara is the world’s largestarea of contemporary dust storm activity and evi-dence from ocean and ice cores suggests that itproduced more dust during the cold phases of the

Ž .Pleistocene see below .The reasons for the relative lack of periSaharan

Žloess deposits are a subject for debate see Wright,.2001 . Some have argued that sufficient silt-sized

material could only be produced in glacial environ-ments and that the Sahara lacks loess because it hasfew mountains and therefore receives insufficient

Žmaterial from mountain glaciers Smalley and Krins-.ley, 1978 . More recent attention has focused on the

mechanics of desert dust formation, transport andŽdeposition Tsoar and Pye, 1987; McTainsh, 1987;

.Yaalon, 1987 . Certainly much Saharan dust hasbeen deposited over the oceans, but on land, onlycertain desert margins appear to have been favourable

Ž .for loess formation. Tsoar and Pye 1987 suggestthat globally the absence of more widespreadperidesert loess is largely due to a lack of availablevegetation traps for dust, an idea also put forward by

Ž .Coude-Gaussen 1990 in comparing loess depositsńorth and south of the Mediterranean. Coude-Gaus-´sen also points to a relative lack of available sedi-ment in North Africa. Another possible reason is the

Žrelative high intensity of rainfall and therefore of.water erosion on the south side of the Sahara. The

mean rainfall per rainy day in the drier parts of WestAfrica averages 9.75 mm, whereas in the drier partsŽ .mean annual rainfall less than 400 mm of the

Ž . Žclassic loess belts it is 4.51 China and 2.56 former.USSR .

Several authors suggest that the current inventoryŽof loess derived from the Sahara is incomplete e.g.

.Coude-Gaussen, 1987; Yaalon, 1987 , but three ar-éas have been studied in some detail: southern TunisiaŽ . ŽCoude-Gaussen et al., 1982 , Northern Nigeria Mc-´

. ŽTainsh, 1987 and the Negev Yaalon and Dan,.1974 . The Matmata plateau loess of southern Tunisia

reaches a thickness of 18 m at Techine and containsúp to five palaeosols typically rich in smectite andpalygorskite. The loess probably derives from thesabkha Chott Djerid and from the Grand Erg Orien-tal.

Ž .Coude-Gaussen et al. 1983 suggest that two´great phases of deposition occurred between 28,000and 10,000 BP and from 6000 to 4000 BP, and

Ž .Coude-Gaussen 1991 provides full details of their´sedimentology. However, while Coude-Gaussen et´

Ž .al. 1983 believed that maximum loess depositionoccurred during humid conditions, this is a view that

Ž .was disputed by Dearing et al. 1996 on the basis oftheir mineral magnetics investigation. They believedthat the period between 15,000 and 20,000 years BPwas a time both of aridity and accelerated loess

Ž .deposition. More recently, Dearing et al. 2001 haveshown that some of the loess is older than this, witha sequence of loess and palaeosols from Techine´being deposited during the period between 100,000and 250,000 years BP.

Material from the Chad basin has provided thesource of the Zaria loess mantle of the Kano plain inNorthern Nigeria, which displays a clear decrease ingrain size with distance from the basin. The domi-nant clay minerals in the Zaria loess are illite andkaolinite. In the Negev, the Netivot loess section isup to 12-m thick and contains distinct palaeosols ofUpper Pleistocene and Holocene age, which indicateclimatic cycles of about 20,000 years duration. Herethe dominant clay mineral is montmorillonite, withsome pedogenic palygorskite.

Two loess deposits in Libya have also been re-cently studied. These have been classified as siltyloess in the Tripoli region in the northwest of thecountry and clayey loess in the Ghat area in the

Ž .southwest Assallay et al., 1996 . Loess has alsoŽbeen identified in the central Sinai Rogner and¨

.Smykatz-Kloss, 1991 . Other sparse deposits are cat-Ž .alogued by Coude-Gaussen 1987 : to the north of´

the Sahara in the Canary Islands, Southern Morocco,


Tripolitania and Southwestern Egypt; and to thesouth in Guinea and Northern Cameroon.

14. Long-term changes in dust supply to the At-lantic

The analysis of deep-sea cores in the Atlanticoffshore from the Sahara provides a picture of long-term changes in dust supply and aeolian activity.Some aeolian activity dates back to the early Creta-

Ž .ceous Lever and McCave, 1983 , and aeolian dust isŽ .present in Neogene sediments Sarnthein et al., 1982 .

However, aeolian activity appears to become moreŽpronounced in the late Tertiary. As Stein 1985, pp.

.312–313 reported: ADistinct maxima of eolian massaccumulation rates and a coarsening of grain size areobserved in the latest Miocene, between 6 and 5 Maand in the Late Pliocene and Quaternary, in the last2.5 million yearsB. They attribute this to both adecrease in precipitation in the Sahara and to anintensified atmospheric circulation. The latter wasprobably caused by an increased temperature gradi-ent between the North Pole and the Equator due toan expansion in the area of Northern Hemisphereglaciation. From about 2.5 Ma, the great tropicalinland lakes of the Sahara began to dry out, and thisis more or less contemporaneous with the time ofonset of mid-latitude glaciation. High dust loadings

Ž .were a feature of the Pleistocene Pokras, 1989 .There is particularly clear evidence for increased

dust inputs at the time of the Last Glacial MaximumŽ . Ž .LGM at around 18 ka . Dust fluxes appear to have

Žbeen 2–4 times higher than at present Kolla et al.,1979; Sarnthein and Koopmann, 1980; Tetzlaff and

.Peters, 1986; Chamley, 1988; Grousset et al., 1998 .By contrast, they appear to have been very lowduring the ‘African Humid Period’, from 14.8 to 5.5ka, when the mass flux off Cape Blanc was reduced

Ž .by 47% De Menocal et al., 2000 .The causes of high dust fluxes during glacial

phases include reductions in precipitation. However,changes in the strength of the northeasterly tradesmay also have been a major contributory factorŽ .Ruddiman, 1997; Grousset et al., 1998 , and variousstudies have been made of wind-transported materi-

Ž .als including diatoms deflated from desiccated lakesto plot wind strength changes over extended periods

Ž .e.g. Hooghiemstra, 1989; Stabell, 1989 . It is possi-ble that increased dust loadings during the LGMwere not only a product of climatic change but also acontributory factor to that change, and this is some-thing that is now being built into climatic modelsŽ .e.g. Mahowald et al., 1999; Overpeck et al., 1996 .

15. Recent changing frequencies of Saharan dustevents

The changing frequencies of Saharan dust eventsover recent decades has been noted by several au-thors, using both data on dust storms observed atmeteorological stations and data on atmospheric dustconcentrations and dustfall deposition rates moni-tored at distance from source areas. Increases in duststorm frequency concurrent with drought periodshave been noted in the Sahelian zone since the

Ž .mid-1960s by Middleton 1985 and by Goudie andŽ .Middleton 1992 using data from Mauritania, Sene-

gal, Nigeria and Sudan. N’Tchayi Mbourou et al.Ž .1997 have also shown increases in both the fre-quency of occurrence and annual duration of dustconditions since the late 1950s, particularly for sta-tions in the Sahel. These trends have been reflectedin rising concentrations of Saharan dust monitored at

Ž .Barbados between 1965 and 1992 Zhu et al., 1997 .The Barbados dust concentrations are inversely re-lated to the previous year’s rainfall in Sahelian AfricaŽ .Prospero, 1996b .

Atmospheric dust loadings are a function of sev-eral climatic parameters that operate on the decadal-scale, including drought as mentioned above. An-other climatic forcing factor that has attracted recent

Žinterest is the North Atlantic Oscillation Moulin et.al., 1997 , while dust storm frequencies may also

Žvary in response to land cover changes Tegen and.Fung, 1995 . Variations in dust event frequencies

could be an indicator of climatic change and thisaspect has attracted the attention of several studies inrecent years. Data from the Mediterranean coast ofSpain, south of Alicante over the period 1949–1994showed a marked increase in the number of dust–rain

Ž .days since the 1970s Sala et al., 1996 . The long-term average there was approximately 2 dust–raindaysryear, but from 1985 to 1994 the annual total


averaged 6.5 dust–rain days, with 9.0 dust–raindaysryear recorded for the period 1989–1994. How-ever, it seems that the frequency of ‘red rain’ eventson the Spanish Mediterranean coast has declined in

Žthe second half of the 1990s Avila et al., 1997;.Avila and Penuelas, 1999 . Eleven years of deposi-˜

Ž .tion records 1984–1994 at Corsica showed thatannual rates peaked in the late 1980s and declined in

Žthe first half of the following decade Loye-Pilot and¨

.Martin, 1996 . This study also noted the high year-to-year variability, with the annual input of Saharandust at Corsica varying between 4.0 and 26.2 g my2

over the study period.Nevertheless, several other authors have remarked

upon the peak in Saharan dustfalls over Europe inŽ .the late 1980s. Dessens and Van Dinh 1990 noted a

marked increase in the frequency of Saharan dustoutbreaks depositing at the Midi–Pyrenees Aerology

Table 10Known Saharan dust falls in the British Isles over the twentieth century

Date Areas affected References

Ž .9 March 1901 Central England Mill 1902Ž .22–23 January 1902 SW England Mill 1902

Ž .21–27 February 1903 Wales, SWrCentral England, E Anglia Mill and Lempfert 1904Ž .28 November 1930 English Channel coast Alexander 1931

Ž . Ž .1 July 1968 England and Wales Pitty 1968 , Stevenson 1969Ž . Ž .6 March 1977 Ireland, W Scotland Tullet 1978 , Bain and Tait 1977Ž .15 May 1979 Ireland Tullet 1980

Ž .28–29 November 1979 Ireland, NW and Central England, Pringle and Bain 1981N Wales, S Scotland

Ž . Ž .28–29 January 1981 NW England, N Ireland Richardson 1981 , George 1981Ž . Ž .11 February 1982 SE England Thomas 1982 , Moon 1982Ž .26–27 January 1983 Southern England Somerset to Kent Thomas 1983

Ž .24 September 1983 Berkshire Pike 1984Ž .29 September 1983 N Ireland Tullett 1984

Ž .22 April 1984 S Wales, Devon Middleton 1986Ž . Ž .9 November 1984 much of England and Wales plus E Scotland Wheeler 1986 , Thomas 1985 ,

Ž . Ž .File 1986 , Cinderey 1987Ž . Ž .4 April 1985 SE England Kent, Cambridge Thomas 1985

Ž .5–6 March 1987 Southern England Burt 1991bŽ .17–18 August 1987 England and Wales Tullet 1988Ž .1 September 1987 N Ireland, Western Scotland Tullet 1988

Ž . Ž .17 September 1987 Southern England Berks Burt 1991bŽ .6 October 1987 England and Wales Tullet 1988Ž .26–27 October 1987 Eastern and Southern England Smith 1988

Ž . Ž .8 May 1988 Southern England Berks Burt 1991bŽ . Ž .18 October 1988 Southern England Berks Burt 1991bŽ . Ž .19 March 1990 Southern England Berks and Hants Burt 1991b

Ž . Ž .7–8 March 1991 Southern England Bucher and Dessens 1992 , Burt 1991a¨Ž .25 March 1991 Kent Thomas 1993Ž .6 September 1991 Kent Thomas 1993

Ž . Ž . Ž .11 October 1991 Southern England Berks and Kent Burt 1992 , Thomas 1993Ž .5 March 1992 Kent Thomas 1993Ž .30 June 1992 Sheffield Thomas 1993Ž . Ž .8 August 1992 Devon, Kent Thomas 1993 , Burt 1995Ž . Ž . Ž .17–18 September 1992 Devon, Berks, Kent Knightley 1993 , Thomas 1993 , Burt 1995

Ž .16–17 March 1993 Berks Burt 1995Ž .21 April 1994 Berks Burt 1995Ž .24 September 1994 Central Southern England Burt 1995

Ž . Ž .14 February 1998 Ireland Co Mayo Sweeney 1998Ž .13 March 2000 Oxfordshire Middleton et al. 2001


Observatory in Lannemezan, France over the period1983–1989. Similarly, a significant increase in thequantities of Saharan dust falling over the FrenchAlps since the early 1970s, with very high inputs

Žoccurring after 1980 De Angelis and Gaudichet,.1991 , was detected from an ice core that yielded

Ždust deposition data over a 30-year period 1955–.1985 .

Contrary to this evidence of increasing frequencyof dust outbreaks across the Western Mediterranean,

Ž .however, Conte et al. 1996 show a decline in thefrequency of strong Siroccos over the period 1951–1990 at Trapani in Sicily. This is probably due to anincrease in anticyclonic activity in the western andcentral parts of the Mediterranean Basin, which tendsto counteract the occurrence of frontal disturbanceswhich generate the strong, dust-laden southerly windsfrom the Sahara.

Nonetheless, the 1980s increase has also beenŽ .noted in the British Isles Burt, 1991b , which de-

rives Saharan dust both from trans-Mediterraneantrajectories and from transport across the Bay of

Ž .Biscay see above . Table 10 shows the Saharandustfalls over British Isles in the twentieth centurydocumented in the literature, which also affirms theimportance of the 1980s and early 1990s, althoughthe increase discernible here may also reflect tosome extent a keener awareness and interest in suchphenomena.

Additional evidence for increasing Saharan dust-raising activity in recent decades comes from the

Ž .eastern Mediterranean. Yaalon and Ganor 1979estimated that some 25 million tonnes of Saharandust reached the East Mediterranean Basin annually,most settling into the Mediterranean Sea. This figurehas subsequently been revised upward, to 70 million

Ž .tonnesryear Ganor and Mamane, 1982 and moreŽrecently to 100 million tonnesryear Ganor and

.Foner, 1996 . The increase reflects the steady rise infrequency of Saharan dust episodes over Tel Avivfrom 10 per year in 1958 to 19 per year in 1991Ž .Ganor, 1994 .

16. Natural and human factors in recent changingdust event frequencies

Decadal-scale periods of increased frequency ofdust events, as reviewed above, reflect some form of

environmental change. However, deciding whethersuch change is driven primarily by natural environ-

Ž .mental factors e.g. droughts, changes in wind fieldsŽ .or by human action e.g. cultivation is no easy task.

ŽOur understanding of the wind erosion system Chepil.and Woodruff, 1963 makes clear the links between

Ženhanced deflation and factors such as drought dueto decreased protective vegetation cover and less

. Žcohesive soil particles and greater wind speeds due.to increased erosivity . Further, changes in dust sup-

ply to the North Atlantic over the long-term havebeen used as indicators of past aridity and as ameasure of the intensity of the atmospheric circula-

Ž .tion see above , although the physical basis forinterpretation of the sedimentary record has not beenfirmly established by measurements of dustrtrans-

Žport relationships in the present day Rea, 1994;.Duce, 1995; Prospero, 1996b .

In the contemporary era, human activities thatdestabilise soil surfaces andror remove vegetation,

Žso exposing soils to enhanced wind erosion Middle-.ton, 1990 , are a further complicating factor. High

dust emission rates have been generally associatedwith semi-arid regions where desert–marginal landsare used for agriculture and herding. Such areas arealso periodically affected by drought, when denudedand disturbed soils are particularly prone to en-

Ž .hanced deflation. Indeed, Goudie’s 1983 findingthat dust storm frequency worldwide is at a peak inareas where mean annual rainfall is between 100 and

Ž .200 mm has been suggested by some Pye, 1987 tobe primarily explained by the cultivation of desert–marginal soils. Other possible explanations for therelatively low frequency of dust storms in areas withless than 100-mm mean annual rainfall may includeinfrequent stream runoff, which limits the amount ofalluvium available as a dust supply, and the relativerarity of strong winds associated with fronts and

Ž .cyclonic disturbances in such areas Goudie, 1983 .However, the Sahara does not fit this global pic-

ture of a peak in dust storm activity in the 100–200-mm mean annual rainfall zone. The two major Saha-ran dust sources identified from remote sensing

Žshown in Fig. 1 the Bodele Region and the area in´ ´.the West Sahara are both located in areas receiving

less than 100 mm of rainfall each year. Goudie’sŽ .1983 relationship between dust-raising and rainfallwas based on terrestrially observed meteorological


data, and meteorological stations are relatively sparsein many of the driest desert regions. The fact thatthis relationship does not hold for the Sahara’s,indeed the world’s, two largest contemporary dustsources could reflect the Sahara as an anomaly to theglobal pattern. Conversely, Goudie’s relationshipmay be more apparent than real, given the relativelack of meteorological stations in the driest zones.This factor certainly applies in Bodele and the West´ ´Sahara. Further, both of the Sahara’s two main sourceareas are little affected by any anthropogenic activi-ties, since both areas have very few settlements andare too dry to support settled agriculture.

The conclusion that the Sahara’s two major dustsources are primarily driven by natural climatic andgeomorphological factors, being little affected byhuman activities, has several implications. ProsperoŽ .1996b has raised the question of ASahara vs. Sa-helB as the primary source for transatlantic dusttransport, and the fact that Barbados dust concentra-tions are inversely related to the previous year’srainfall in Sahelian Africa suggests a link withsources in the Sahel. Indeed, several studies haveshown the importance of areas in Sahelian latitudesas source areas that have increased their dust outputduring recent periods characterised both by pro-longed drought and intensified land use, in places,

Žleading to desertification Middleton, 1985; Goudieand Middleton, 1992; N’Tchayi Mbourou et al.,

.1997 .However, this relationship becomes less certain in

the light of the character of the Sahara’s two maindust source areas. While Harmattan dust blown fromthe Bodele depression tends not to travel far over the´ ´Gulf of Guinea, dust entrained in the West Saharansource area does make a significant contribution totransatlantic flows. However, while the effect ofdrought on dust-raising in Sahelian latitudes can beestablished, reference to AdroughtB in an area receiv-ing less than 100 mm in mean annual rainfall is lesssound. Two explanations for the established increasein transatlantic dust transport in recent decades canbe proposed. Firstly, that flows from a constant WestSaharan source area have been augmented by mate-rial from more southerly, drought-affected sources.Secondly, it is possible that the relationship betweenBarbados dust concentrations and Sahelian rainfallreflect other changes in atmospheric circulation asso-

Ž .ciated with drought in the Sahel. As Prospero 1996bhas pointed out, both the Hadley circulation and themid-tropospheric easterly jet are more intense during

ŽSahelian dry spells Nicholson, 1986; Newell and.Kidson, 1984 . The second hypothesis is not neces-

sarily incompatible with the first. Stronger winds, bethey low-level entraining airflows, upper tropo-spheric transporting flows, or both, could mean morematerial being transported from the West Saharansource area andror from Sahelian source areas. Onlyfurther research on tracking specific transatlantic dustevents can establish the validity of these proposals.

17. Conclusions

The Sahara is the world’s largest source of desertdust, indicating the importance of aeolian geomor-phology in this major world desert. Saharan dust hasimportant influences on nutrient dynamics and bio-geochemical cycling in both oceanic and terrestrialecosystems in North Africa and far beyond, due tofrequent long-range transport across the AtlanticOcean, the Mediterranean Sea and the Red Sea, tothe Americas, Europe and the Middle East. Atmo-spheric dust concentrations may also have consider-able climatic significance through a range of possiblemechanisms, and the frequency of dust events canchange substantially in response to climatic changesover several time scales.

Although the precise locations of Saharan dustsource areas are not well known, data from the Total

Ž .Ozone Mapping Spectrometer TOMS suggest twomajor source areas: the Bodele depression and an´ área covering eastern Mauritania, Western Mali andsouthern Algeria. Trajectories of long-distance trans-port are relatively well documented, and confirmedby the TOMS data, but the links between sourceareas and seasonal Saharan dust pathways are not.One of the few definitive links commonly made inthe literature in this respect is between Harmattandust from the Bodele depression and the prominent´ ´winter plume over the tropical North Atlantic thatreaches South America. However, a seasonal com-parison between data from the only period whenregular monitoring of atmospheric mineral dust wascarried out at Cayenne and data on thick dust haze atMaiduguri in Northern Nigeria, a station directly


within the Harmattan trajectory, suggests that thislink is erroneous.

The Sahara’s two major dust sources are locatedin areas that receive very low rainfall totals andhence the world’s largest dust source does not fit theotherwise global picture of a peak in dust stormactivity in the 100–200-mm mean annual rainfallzone. Further, the Sahara’s two major dust sourcesare largely driven by natural factors of geomorphol-ogy and climate, being little affected by humanactivities. This finding contrasts with those of otherstudies that have found the frequency of Saheliandust events to have varied markedly in recent decadesdue to climatic factors such as drought and anthro-pogenic disturbance of desert marginal surfaces. Ref-erence to AdroughtB in source areas receiving lessthan 100 mm in mean annual rainfall is not sound.These findings have implications for the documentedestablished increase in transatlantic dust transport inrecent decades, but only further research on trackingspecific transatlantic dust events can establish thevalidity of the suggested explanations.

Acknowledgements

We would like to acknowledge the benefit wehave gained from collaboration with our colleagues,Richard Washington and Martin Todd, and the greatstimulus that we and many others have gained fromthe work of Dr. Joseph Prospero. We are grateful toAilsa Allen for drawing the figures, and to thejournal referees.

References

Afeti, G.M., Resch, F.J., 2000. Physical characteristics of Saharandust near the Gulf of Guinea. Atmospheric Environment 34,1273–1279.

Alexander, J., 1931. The fatal Belgian fog. Science 73, 96.Alpert, P., Ganor, E., 1993. A jet stream associated heavy dust

storm in the eastern Mediterranean. Journal of GeophysicalŽ .Research 98 D4 , 7339–7349.

Alpert, P., Kaufman, Y.J., Shay-El, Y., Tanre, D., da Silva, A.,Schubert, S., Joseph, J.H., 1998. Quantification of dust-forcedheating of the lower troposphere. Nature 395, 367–370.

Andreae, M.O., 1995. Climatic effects of changing atmosphericŽ .aerosol levels. In: Henderson-Sellers, A. Ed. , World Survey

of Climatology. Future Climates of the World, vol. 16. Else-vier, Amsterdam.

Arimoto, R., Ray, B.J., Lewis, N.F., Tomza, U., Duce, R.A.,1997. Mass particle size distribution of atmospheric dust andthe dry deposition of dust to the remote ocean. Journal ofGeophysical Research D 102, 15867–15874.

Assallay, A.M., Rogers, C.D.F., Smalley, I.J., 1996. Engineeringproperties of loess in Libya. Journal of Arid Environments 32,373–386.

Avila, A., Penuelas, J., 1999. Increasing frequency of Saharan˜rains over northeastern Spain and its ecological consequences.Science of the Total Environment 228, 153–156.

Avila, A., Queralt, I., Gallart, F., Martin-Vide, J., 1996. Africandust over northeastern Spain: mineralogy and source regions.

Ž .In: Guerzoni, S., Chester, R. Eds. , The Impact of DesertDust Across the Mediterranean. Kluwer Academic Publishing,Dordrecht, pp. 201–205.

Avila, A., Queralt-Mitjans, I., Alarcon, M., 1997. Mineralogicalćomposition of African dust delivered by red rains over north-

Ž .eastern Spain. Journal of Geophysical Research 102 D18 ,21977–21996.

Bain, D.C., Tait, N.M., 1977. The mineralogy and origin ofdust-fall on Skye. Clay Minerals 12, 353–355.

Bergametti, G., Gomes, L., Remoudaki, E., Desbois, M., Martin,D., Buat-Menard, P., 1989. Present transport and deposition´patterns of African dusts to the north-western Mediterranean.

Ž .In: Leinen, M., Sarnthein, M. Eds. , Palaeoclimatology andPalaeometeorology: Modern and Past Patterns of Global At-mospheric Transport. NATO ASI Series, C, vol. 282, pp.227–252.

Bertrand, J., Cerf, A., Domergue, J.L., 1979. Repartition in spaceand time of dust haze south of the Sahara. WMO 538,409–415.

Borushko, I.S., 1972. Dust storm distribution in the Tropics.ŽGlavnaya Geofizicheskaya Observatoriya 284, 76–83 In Rus-

.sian .Brooks, C.E.P., 1920. The meteorology of British Somaliland.

Quarterly Journal of the Royal Meteorological Society 46,434–438.

Brooks, N., Legrand, M., 2000. Dust variability over northernAfrica and rainfall in the Sahel. In: McLaren, S.J., Kniverton,

Ž .D. Eds. , Linking Land Surface Change to Climate Change.Kluwer Academic Publishing, Dordrecht, pp. 1–25.

Bryson, R.A., Barreis, I.A., 1967. Possibilities of major climaticmodifications and their implications: northwest India, a casefor study. Bulletin of the American Meteorological Society 48,136–142.

Bucher, A., 1986. Recherches sur les poussieres minerals d’origine¨ `saharienne. Universite de Reims-Champagne, Ardenne, PhD´Thesis.

Bucher, A., Dessens, J., 1992. Saharan dust over France andËngland, 6–9 March 1991. Journal of Meteorology 17, 226–233.

Bucher, A., Lucas, G., 1984. Sedimentation eolienne interconti-¨ ´ ńentale, poussieres sahariennes et geologie. Bulletin des Cen-` ´tres de Recherches Exploration-Production Elf-Aquitaine 8,151–165.


Burt, S., 1991a. Falls of dust rain in Berkshire, March 1991.Weather 46, 248.

Burt, S., 1991b. Falls of dust rain within the British Isles. Weather46, 347–353.

Burt, S., 1992. Fall of dust rain, 11 October 1991. Weather 47,142.

Burt, S., 1995. Fall of dust rain in Berkshire, 24 September 1994.Weather 50, 30–31.

Callot, Y., Marticorena, B., Bergametti, G., 2000. Geomorpho-logic approach for modelling the surface features of aridenvironments in a model of dust emissions: application to theSahara desert. Geodinamica Acta 13, 245–270.

Caquineau, S., Gaudichet, A., Gomes, L., Magonthier, M.-C.,Chatenet, B., 1998. Saharan dust: clay ratio as a relevant tracerto assess the origin of soil-derived aerosols. GeophysicalResearch Letters 25, 983–986.

Carlson, T.N., 1979. Atmospheric turbidity in Saharan dust out-breaks as determined by analyses of satellite brightness data.Monthly Weather Review 107, 322–335.

Carlson, T.N., Prospero, J.M., 1972. The large-scale movement ofSaharan air outbreaks over the northern equatorial Atlantic.Journal of Applied Meteorology 11, 283–297.

Chamley, H., 1988. Contribution eolienne a la sedimentation´ `marine au large du Sahara. Bulletin de la Societe Geologique´ ´ ´de France IV 6, 1091–1100.

Chepil, W.S., Woodruff, N.P., 1963. The physics of wind erosionand its control. Advances in Agronomy 15, 211–302.

Chester, R., Elderfield, J.J., Griffin, J.J., Johnson, L.R., Padgham,R.C., 1972. Eolian dust along the eastern margins of theAtlantic Ocean. Marine Geology 13, 91–106.

Chiapello, I., Prospero, J.M., Herman, J.R., Hsu, N.C., 1999.Detection of mineral dust over the North Atlantic Ocean andAfrica with the Nimbus 7 TOMS. Journal of Geophysical

Ž .Research 104 D8 , 9277–9291.Cinderey, M.B., 1987. Dust deposits in England on 9 November

1984. Weather 42, 28.Clarke, F.W., 1916. The data of Geochemistry, 3rd edn. U.S.

Geological Survey Bulletin, 821 pp.Conte, M., Colacino, M., Piervitali, E., 1996. Atlantic distur-

bances deeply penetrating the African continent: effects overSaharan regions and the Mediterranean Basin. In: Guerzoni,

Ž .S., Chester, R. Eds. , The Impact of Desert Dust Across theMediterranean. Kluwer Academic Publishing, Dordrecht, pp.93–102.

Coude-Gaussen, G., 1981. Etude detaillee d’un echantillon de´ ´ ´ ´poussieres eoliennes preleve au Tanezrouft, le 10 decembre` ´ ´ ´ ´1980. Recherches Geographiques a Strasbourg 16r17, 121–´ `130.

Coude-Gaussen, G., 1987. The peri-Saharan loess: sedimentologi-ćal characterisation and palaeoclimatic significance. GeoJour-nal 15, 177–183.

Coude-Gaussen, G., 1990. The loess and loess-like deposits along´the sides of the western Mediterranean Sea: genetic andpalaeoclimatic significance. Quaternary International 5, 1–8.

Coude-Gaussen, G., 1991. Les poussieres Sahariennes. John´ `Libbey Eurotext, Montrouge.

Coude-Gaussen, G., Blanc, P., 1985. Presence de grains eolises de´ ´ ´ ´palygorskite dans les poussieres actuelles et les sediments` ´recents d’origine desertique. Bulletin de la Societe Geologique´ ´ ´ ´ ´

Ž .de France 8 4 , 571–579.Coude-Gaussen, G., Mosser, C., Rognon, P., Torenq, J., 1982.´

Une accumlation de loess Pleistocene superieur dans le sud-´Tunisien: coupe de Techine. Bulletin de la Societe Geologique´ ´ ´ ´de France 24, 283–292.

Coude-Gaussen, G., Olive, L., Rognon, P., 1983. Datation de´depots loessiques et varations climatiques a la bordure nord du´ ˆ `Sahara algero–tunisien. Revue de Geologie Dynamique et de´ ´Geographie Physique 24, 61–73.´

Coude-Gaussen, G., Desire, E., Regrain, R., 1988. Perticularite´ ´ ´ ´des poussieres sahariennes distales tombees sur la Picardie et` ´l’Ile de France le 7 mai 1988. Hommes et Terres du Nord 4,246–251.

D’Almeida, G.A., 1986. A model for Saharan dust transport.Journal of Climate and Applied Meteorology 25, 903–916.

D’Almeida, G.A., 1987. Desert aerosol characteristics and effectsŽ .on climate. In: Leinen, M., Sarnthein, M. Eds. , Palaeoclima-

tology and Palaeometeorology: Modern and Past Patterns ofGlobal Atmospheric Transport. NATO ASI Series, C, vol.282, pp. 311–338.

Dayan, U., 1986. Climatology of back trajectories from Israelbased on synoptic analysis. Journal of Climate and AppliedMeteorology 25, 591–595.

Dayan, U., Heffter, J., Miller, J., Gutman, G., 1991. Dust intru-sion into the Mediterranean basin. Journal of Applied Meteo-rology 30, 1185–1199.

De Angelis, M., Gaudichet, A., 1991. Saharan dust depositionŽ .over Mont Blanc French Alps during the last 30 years. Tellus

43B, 61–75.Dearing, J.A., Livingstone, I.P., Zhou, L.P., 1996. A late Quater-

nary magnetic record of Tunisian loess and its climatic signifi-cance. Geophysical Research Letters 23, 189–192.

Dearing, J.A., Livingstone, I.P., Bateman, M.D., White, K., 2001.Palaeoclimate records from OIS 8.0–5.4 recorded in loess inpalaeosol sequences on the Matmata Plateau, southern Tunisia,based on mineral magnetism and new luminescence dating.Quaternary International 76r77, 43–56.

De Menocal, P., Ortiz, J., Guilderson, T., Sarnthein, M., 2000.Coherent high- and low-latitude climate variability during theholocene warm period. Science 288, 2198–2202.

Dessens, J., Van Dinh, P., 1990. Frequent Saharan dust outbreaksnorth of the Pyrenees: a sign of climatic change? Weather 45,327–333.

Drees, L.R., Manu, A., Wilding, L.P., 1993. Characteristics ofaeolian dusts in Niger, West Africa. Geoderma 59, 213–233.

Dubief, J., 1953. Les vents de sable dans le Sahara Francais.Colloques Internationaux du CNRS 35, 45–70.

Duce, R.A., 1995. Sources, distributions, and fluxes of mineralaerosols and their relationship to climate. In: Charlson, R.J.,

Ž .Heintzenberg, J. Eds. , Aerosol Forcing of Climate. Wiley,Chichester, pp. 43–72.

File, R.F., 1986. Dust deposits in England on 9 November 1984.Weather 41, 191–195.


Fisher, W.B., 1978. The Middle East, 7th edn. Methuen, London,615 pp.

Foucault, A., Melieres, F., 2000. Palaeoclimatic cyclicity in cen-´ `tral Mediterranean Pliocene sediments: the mineralogical sig-nal. Palaeogeography, Palaeoclimatology, Palaeoecology 158,311–323.

Fouquart, Y.B., Bonnell, G., Brogniez, G., Buriez, J.C., Smith, L.,Morcrette, J.J., Cerf, A., 1987. Observations of Saharanaerosols: results of ECLATS field experiment: II. Broadbandradiative characteristics of the aerosols and vertical radiativeflux divergence. Journal of Applied Meteorology 26, 38–52.

Franzen, L.G., Hjelmroos, M., Kallberg, P., Brorstrom-Lunden,E., Juntto, S., Savolainen, A.-L., 1994. The ‘yellow snow’episode of northern Fennoscandia, March 1991—a case studyof long-distance transport of soil, pollen and stable organic

Ž .compounds. Atmospheric Environment 28 22 , 3587–3604.Franzen, L.G., Hjelmroos, M., Kallberg, P., Rapp, A., Mattsson,

J.O., Brorstrom-Lunden, E., 1995. The Saharan dust episode¨ óf south and central Europe, and northern Scandinavia, March1991. Weather 50, 313–318.

Freeman, M.H., 1952. Duststorms of the Anglo–Egyptian Sudan.Meteorological Office, Meteorological Reports No. 11, 22 pp.

Ganor, E., 1994. The frequency of Saharan dust episodes over TelAviv, Israel. Atmospheric Environment 28, 2867–2871.

Ganor, E., Foner, H.A., 1996. The mineralogical and chemicalproperties and the behaviour of aeolian Saharan dust over

Ž .Israel. In: Guerzoni, S., Chester, R. Eds. , The Impact ofDesert Dust Across the Mediterranean. Kluwer AcademicPublishing, Dordrecht, pp. 163–172.

Ganor, E., Mamane, M., 1982. Transport of Saharan dust acrossthe eastern Mediterranean. Atmospheric Environment 16,581–587.

Ganor, E., Foner, H.A., Brenner, S., Neeman, E., Lavi, N., 1991.The chemical composition of aerosols settling in Israel follow-ing dust storms. Atmospheric Environment 25, 2665–2670.

George, D.J., 1981. Dustfall and instability rain over NorthernIreland on the night of 28–29 January 1981. Weather 36,216–217.

Gillies, J.A., Nickling, W.G., McTainsh, G.H., 1996. Dust con-centrations and particle-size characteristics of an intense dusthaze event: inland delta region, Mali, West Africa. Atmo-spheric Environment 30, 1081–1090.

Glaccum, R.A., Prospero, J.M., 1980. Saharan aerosols over thetropical north Atlantic. Mineralogy, Marine Geology 37, 295–321.

Goudie, A.S., 1978. Dust storms and their geomorphologicalimplications. Journal of Arid Environments 1, 291–310.

Goudie, A.S., 1983. Dust storms in space and time. Progress inPhysical Geography 7, 502–530.

Goudie, A.S., Middleton, N.J., 1992. The changing frequency ofdust storms through time. Climatic Change 20, 197–225.

Grousset, F.E., Parra, M., Bory, A., Martinez, P., Bertrand, P.,Shimmield, G., Ellam, R.M., 1998. Saharan wind regimestraced by the Sr–Nd isotopic composition of subtropical At-lantic sediments: last glacial maximum vs. today. QuaternaryScience Reviews 17, 395–409.

Gruber, N., Sarmiento, J.L., 1997. Global patterns of marine

nitrogen fixation and denitrification. Global BiogeochemicalCycles 11, 235–266.

Guerzoni, S., Chester, R., Dulac, F., Herut, B., Loye-Pilot, M.D.,Measures, C., Mignon, C., Molinari, E., Moulin, C., Rossini,P., Saydam, C., Soudine, A., Ziveri, P., 1999. The role ofatmospheric deposition in the biogeochemistry of the Mediter-ranean Sea. Progress in Oceanography 44, 147–190.

Herrmann, L., Stahr, K., Jahn, R., 1999. The importance of sourceregion identification and their properties for soil-derived dust:the case of Harmattan dust sources for eastern West Africa.Contributions to Atmospheric Physics 72, 141–150.

Herut, B., Krom, M., 1996. Atmospheric input of nutrients anddust to the SE Mediterranean. In: Guerzoni, S., Chester, R.Ž .Eds. , The Impact of Desert Dust Across the Mediterranean.Kluwer Academic Publishing, Dordrecht, pp. 349–358.

Hooghiemstra, H., 1989. Variations of the N.W. African tradewind regime during the past 140,000 years: changes in pollenflux evidenced by marine sediment records. In: Leinen, M.,

Ž .Sarnthein, M. Eds. , Palaeoclimatology and Palaeometeorol-ogy: Modern and Past Patterns of Global Atmospheric Trans-port. Kluwer Academic Publishing, Dordrecht, pp. 733–770.

Hsu, N.C., Herman, J.R., Torres, O., Holben, B., Tanre, D., Eck,T.F., Smirnov, A., Chatenet, B., Lavenu, F., 1999. Compar-isons of the TOMS aerosol index with sun photometer aerosoloptical thickness: results and applications. Journal of Geophys-ical Research 104, 6269–6279.

Husar, R.B., Prospero, J.M., Martin, L., 1997. Characterisation oftropospheric aerosols over the oceans with the NOAA veryhigh resolution radiometer optical thickness operational prod-uct. Journal of Geophysical Research 102, 16889–16909.

Jaenicke, R., 1979. Monitoring and critical review of the esti-mated source strength of mineral dust from the Sahara. In:

Ž .Morales, C. Ed. , Saharan Dust. Wiley, Chichester, pp. 233–242.

Jickells, T.D., Dorling, S., Deuser, W.G., Church, T.M., Arimoto,R., Prospero, J.M., 1998. Air-borne dust fluxes to a deep watersediment trap in the Sargasso Sea. Global BiogeochemicalCycles 12, 311–320.

Kalu, A.E., 1979. The African dust plume: its characteristics andpropagation across West Africa in winter. In: Morales, C.Ž .Ed. , Saharan Dust: Mobilisation, Transport and Deposition.Wiley, Chichester, pp. 95–118.

Knightley, P., 1993. Dust rain—17 September 1992. Weather 48,58–59.

Kolla, V.P., Biscaye, P.E., Hanley, A.F., 1979. Distribution ofquartz in Late Quaternary Atlantic sediments in relation toclimate. Quaternary Research 11, 261–277.

Kubilay, N., Nickovic, S., Moulin, C., Dulac, F., 2000. Anillustration of the transport and deposition of mineral dust ontothe eastern Mediterranean. Atmospheric Environment 34,1293–1303.

Le-Bolloch, O., Guerzoni, S., Molinaroli, E., 1996. Atmosphere–ocean mass fluxes at two coastal sites in Sardinia 39–41degrees N, 8–10 degrees E. In: Guerzoni, S., Chester, R.Ž .Eds. , The Impact of Desert Dust Across the Mediterranean.Kluwer Academic Publishing, Dordrecht, pp. 217–222.

Lever, A., McCave, I.N., 1983. Eolian components in Cretaceous


and Tertiary North Atlantic sediments. Journal of SedimentaryPetrology 53, 811–832.

Levin, Z., Ganor, E., Gladstein, V., 1996. The effects of desertparticles with sulphate on rain formation in the EasternMediterranean. Journal of Applied Meteorology 35, 1511–1523.

Li, X., Maring, H., Savoie, D., Voss, K., Prospero, J.M., 1996.Dominance of mineral dust in aerosol light-scattering in theNorth Atlantic trade winds. Nature 380, 416–419.

Littmann, T., 1991a. Rainfall, temperature and dust storm anoma-lies in the African Sahel. Geographical Journal 157, 136–160.

Littmann, T., 1991b. Recent African dust deposition in WestGermany—sediment characteristics and climatological as-pects. Catena Supplement 20, 57–73.

Loye-Pilot, M.D., Martin, J.M., 1996. Saharan dust input to the¨western Mediterranean: an eleven year record in Corsica. In:

Ž .Guerzoni, S., Chester, R. Eds. , The Impact of Desert DustAcross the Mediterranean. Kluwer Academic Publishing, pp.191–199.

Loye-Pilot, M.D., Martin, J.M., Morelli, J., 1986. Influence of¨Saharan dust on the rainfall acidity and atmospheric input tothe Mediterranean. Nature 321, 427–428.

Macleod, D.A., 1980. The origin of the red Mediterranean soils inEpirus, Greece. Journal of Soil Science 31, 125–136.

Mahowald, N., Kohfeld, K., Hansson, M., Balkanski, Y., Harri-son, S.P., Prentice, I.C., Schulz, M., Rodhe, H., 1999. Dustsources and deposition during the last glacial maximum andcurrent climate: a comparison of model results with paleodatafrom ice cores and marine sediments. Journal of GeophysicalResearch 104D, 15895–15916.

Mainguet, M., 1980. Aeolian interdependencies of arid Saharanareas on the borders of the Sahel and the consequences for thepropagation of desertification. Stuttgarter Geographische Stu-dien 95, 107–123.

Mainguet, M., Chemin, M.-C., 1990. Le Massif du Tibesti dans lesysteme eolien du Sahara: reflexion sur la genese du Lac` ´ ´ `Tchad. Berliner Geographische Studien 30, 261–276.

Maley, J., 1982. Dust, clouds, rain types and climatic variations intropical north Atlantic. Quaternary Research 18, 1–16.

Marticorena, B., Bergametti, G., 1996. Two-years simulations ofseasonal and interannual changes of the Saharan dust emis-sions. Geophysical Research Letters 23, 1921–1924.

Mattsson, J.O., Nihlen, T., 1996. The transport of Saharan dust to´southern Europe: a scenario. Journal of Arid Environments 32,110–111.

McDonald, W.F., 1938. Atlas of climatic charts of the oceans.Department of Agriculture, Weather Bureau, Washington DC.

McTainsh, G.H., 1987. Desert loess in northern Nigeria. Zeitschriftfuer Geomorphologie 26, 417–435, NF.

McTainsh, G.H., Walker, P.H., 1982. Nature and distribution ofHarmattan dust. Zeitschrift f ur Geomorphologie 26, 417–435.¨

Measures, C.I., Brown, E.T., 1996. Estimating dust input to theAtlantic Ocean using surface water aluminium concentrations.

Ž .In: Guerzoni, S., Chester, R. Eds. , The Impact of DesertDust Across the Mediterranean. Kluwer Academic Publishing,Dordrecht, pp. 301–311.

Middleton, N.J., 1985. Effect of drought on dust production in theSahel. Nature 316, 431–434.

Middleton, N.J., 1986. The geography of dust storms. Unpub-lished PhD thesis, University of Oxford, UK.

Middleton, N.J., 1990. Wind erosion and dust storm prevention.Ž .In: Goudie, A.S. Ed. , Desert Reclamation. Wiley, Chich-

ester, pp. 87–108.Middleton, N.J., Goudie, A.S., 2001. Saharan Dust: sources and

trajectories. Transactions of the Institute of British Geogra-phers 26, in press.

Middleton, N.J., Betzer, P.R., Bull, P.A., 2001. Long-range trans-port of ‘giant’ aeolian quartz grains: linkage with discretesedimentary sources and implications for protective particletransfer. Marine Geology, in press.

Mignon, C., Sandroni, V., 1999. Phosphorus in rainwater: parti-tioning inputs and impact on the surface coastal ocean. Lim-nology and Oceanography 44, 1160–1165.

Mill, H.R., 1902. Dust showers in the south-west of England.Symons’s Monthly Meteorological Magazine 37, 1–4.

Mill, H.R., Lempfert, R.G.K., 1904. The great dust-fall of Febru-ary 1903, and its origin. Quarterly Journal of the RoyalMeteorological Society 30, 57–91.

Miller, R.L., Tegen, I., 1998. Climate response to soil dustaerosols. Journal of Climate 11, 3247–3267.

Molinaroli, E., 1996. Mineralogical characterisation of Saharandust with a view to its final destination in Mediterranean

Ž .sediments. In: Guerzoni, S., Chester, R. Eds. , The Impact ofDesert Dust Across the Mediterranean. Kluwer AcademicPublishing, pp. 153–162.

Moon, A.E., 1982. Dust-fall at Hastings, Sussex. Journal ofMeteorology 7, 92.

Morales, A.F., 1946. El Sahara Espanol. Alta Comisaria de Espana˜ ˜en Marruecos, Madrid, 167 pp.

Moulin, C., Lambert, C.E., Dulac, F., Dayan, U., 1997. Control ofatmospheric export of dust by the North Atlantic Oscillation.Nature 387, 691–694.

Muhs, D.R., Bush, C.A., Stewart, K.C., 1990. Geochemical evi-dence of Saharan dust parent material for soils developed onQuaternary limestones of Caribbean and western Atlantic is-lands. Quaternary Research 33, 157–177.

Nalivkin, D.V., 1983. Hurricanes, Storms and Tornadoes.Amerind, New Delhi, 597 pp.

Naval Intelligence Division, 1943. Algeria. Geographical Hand-book Series, vol. 1, OUP, 313 pp.

Newell, R.E., Kidson, J.W., 1984. African mean wind changesbetween Sahelian wet and dry periods. Journal of Climatology4, 27–33.

Nicholson, S.E., 1986. The spatial coherence of African rainfallanomalies: interhemispheric teleconnections. Journal of Cli-mate and Applied Meteorology 25, 1365–1381.

Nicholson, S.E., 2000. Land surface processes and Sahel climate.Reviews of Geophysics 38, 117–139.

Nihlen, T., Olsson, S., 1995. Influence of eolian dust on soilformation in the Aegean area. Zeitschrift f ur Geomorphologie¨39, 341–361.

N’Tchayi Mbourou, G., Berrand, J.J., Nicholson, S.E., 1997. Thediurnal and seasonal cycles of wind-borne dust over Africanorth of the Equator. Journal of Applied Meteorology 36,868–882.

Overpeck, J., Rind, D., Lacis, A., Healy, R., 1996. Possible role


of dust-induced regional warming in abrupt climate changeduring the last glacial period. Nature 384, 447–449.

Ozer, P., Erpicum, M., Cortemiglia, G.C., Luchettti, G., 1998. Adustfall event in November 1996 in Genoa, Italy. Weather 53,140–145.

Ozsoy, T., Saydam, A.C., 2000. Acidic and alkaline precipitationin the Cilician Basin, north-eastern Mediterranen Sea. Scienceof the Total Environment 253, 93–109.

Paquet, H., Coude-Gaussen, G., Rognon, P., 1984. Etude´mineralogique de poussieres sahariennes le long d’un itineraire´ ` éntre 198 et 358 de latitude nord. Revue de Geologie Dy-´ ńamique et de Geographie Physique 25, 257–265.´

Perry, K.D., Cahill, T.A., Eldred, R.A., Dutcher, D.D., 1997.Long-range transport of North African dust to the easternUnited States. Journal of Geophysical Research 102, 11225–11238.

Ž .Pesci, M., 1968. Loess. In: Fairbridge, R.W. Ed. , The Ency-clopaedia of Geomorphology. Reinhold, New York, pp. 674–678.

Peterson, J.T., Junge, C.E., 1971. Source of particulate matter inthe atmosphere. In: Matthews, W., Kellog, W., Robinson,

Ž .G.D. Eds. , Man’s Impact on Climate. MIT Press, Cam-bridge, MA, pp. 310–320.

Ž .Petit-Maire, N. Ed. , 1991. Paleoenvironments du Sahara. Edi-´tions du Centre National de la Recherche Scientific, Paris, 237pp.

Pike, W.S., 1984. Dust-fall, west Berkshire, 24 September 1983.Journal of Meteorology 9, 21–22.

Pitty, A.F., 1968. Particle size of the Saharan dust which fell inBritain in July 1968. Nature 220, 364–365.

Pokras, E.M., 1989. Pliocene history of South SaharanrSahelianŽ .aridity: record of freshwater diatoms Genus Melosira and

opal phytoliths, ODP sites 662 and 664. In: Leinen, M.,Ž .Sarnthein, M. Eds. , Palaeoclimatology and Palaeometeorol-

ogy: Modern and Past Patterns of Global Atmospheric Trans-port. Kluwer Academic Publishing, Dordrecht, pp. 795–804.

Pringle, A.W., Bain, D.C., 1981. Saharan dustfalls on north-westEngland. Geographical Magazine 53, 729–732.

Prospero, J.M., 1996a. Saharan dust transport over the NorthAtlantic Ocean and Mediterranean: an overview. In: Guerzoni,

Ž .S., Chester, R. Eds. , The Impact of Desert Dust Across theMediterranean. Kluwer Academic Publishing, Dordrecht, pp.133–151.

Prospero, J.M., 1996b. The atmospheric transport of particles tothe ocean. In: Ittekkot, V., Schafer, P., Honjo, S., Depetris,

Ž .P.J. Eds. , Particle Flux in the Ocean. Wiley, Chichester, pp.19–52.

Prospero, J.M., 1999. Long-term measurements of the transport ofAfrican mineral dust to the south-eastern United States: impli-cations for regional air quality. Journal of Geophysical Re-

Ž .search D 104 13 , 15917–15927.Prospero, J.M., Carlson, T.N., 1981. Saharan air outbreaks over

the tropical North Atlantic. PAGEOPH 119, 677–691.Prospero, J.M., Bonatti, E., Schuber, C., Carlson, T.N., 1970.

Dust in the Caribbean traced to an African dust storm. Earthand Planetary Science Letters 9, 287–293.

Prospero, J.M., Glaccum, R.A., Nees, R.T., 1981. Atmospherictransport of soil dust from Africa to South America. Nature289, 570–572.

Prospero, J.M., Nees, R.T., Uematsu, M., 1987. Deposition rate ofparticulate and dissolved aluminium derived from Saharandust in precipitation at Miami, Florida. Journal of GeophysicalResearch 92, 14723–14731.

Pye, K., 1987. Aeolian Dust and Dust Deposits. Academic Press,London.

Pye, K., 1992. Aeolian dust transport and deposition over Creteand adjacent parts of the Mediterranean Sea. Earth SurfaceProcesses and Landforms 17, 271–288.

Quijano, A.L., Sokolik, I.N., Toon, B.O., 2000. Radiative heatingrates and direct radiative forcing by mineral dust in cloudyatmospheric conditions. Journal of Geophysical Research 105Ž .D10 , 12207–12219.

Rajkumar, W.S., Chang, A.S., 2000. Suspended particulate matterconcentrations along the east-west corridor, Trinidad, westIndies. Atmospheric Environment 34, 1181–1187.

Rapp, A., 1984. Are terra rosa soils in Europe eolian depositsfrom Africa? Geologiska Foereningens i Stockholm Foerhan-dlingar 105, 161–168.

Ratmeyer, V., Balzer, W., Bergametti, G., Chiapello, I., Fischer,G., Wyputta, U., 1991. Seasonal impact of mineral dust ondeep-ocean particle flux in the eastern Atlantic Ocean. MarineGeology 159, 241–252.

Rea, D.K., 1994. The paleoclimatic record provided by eoliandeposition in the deep sea: the geologic history of wind.Reviews of Geophysics 32, 159–195.

Reiff, J., Forbes, G.S., Spieksma, F.T.M., Reynders, J.J., 1986.African dust reaching north-western Europe: a case study toverify trajectory calculations. Journal of Climate and AppliedMeteorology 25, 1543–1567.

Richardson, J.S.W., 1981. Pink dustfall. Journal of Meteorology6, 89–90.

Rodriguez, S., Querol, X., Alastney, A., Kallos, G., Kalaliagou,O., 2001. Saharan dust contributions to PM10 and TSP levelsis southern and eastern Spain. Atmospheric Environment 35,2433–2447.

Rogner, K., Smykatz-Kloss, W., 1991. The deposition of eolian¨sediments in lacustrine and fluvial environments of Central

Ž .Sinai Egypt . Catena Supplement 20, 75–91.Ruddiman, W.F., 1997. Tropical Atlantic terrigenous fluxes since

25,000 years B.P. Marine Geology 136, 189–207.Sala, J.Q., Cantos, J.O., Chiva, E.M., 1996. Red dust within the

Spanish Mediterranean area. Climatic Change 32, 215–228.Sarnthein, M., Koopmann, B., 1980. Late Quaternary deep-sea

record on northwest African dust supply and wind circulation.Palaeoecology of Africa 12, 239–253.

Sarnthein, D., Thiede, J., Pflaumann, U., Erlenkeuser, H., Fut-terer, D., Koopman, B., Lange, H., Seibold, E., 1982. Atmo-spheric and oceanic circulation patterns off northwest Africaduring the past 25 million years. In: von Rad, U., Hinz, K.,

Ž .Sarnthein, M., Seibold, E. Eds. , Geology of the NorthwestAfrican Continental Margin. Springer, Berlin, pp. 545–604.

Schlatter, T., 1995. Long distance dust. Weatherwise 48, 38–39.


Schollaert, S.E., Merrill, J.T., 1998. Cooler sea surface west of theSahara Desert correlated to dust events. Geophysical ResearchLetters 25, 3529–3532.

Schroeder, J.H., 1985. Eolian dust in the coastal desert of theSudan: aggregates cemented by evaporites. Journal of AfricanEarth Sciences 3, 370–380.

Schutz, L., 1979. Saharan dust transport over the North AtlanticÖcean—model calculations and measurements. In: Morales,

Ž .C. Ed. , Saharan Dust. Wiley, Chichester, pp. 267–278.Schutz, L., 1980. Long range transport of desert dust with special¨

emphasis on the Sahara. Annals of the New York Academy ofSciences 338, 515–532.

Schutz, L., Jaenicke, R., Pietrek, H., 1981. Saharan dust transport¨Ž .over the North Atlantic Ocean. In: Pewe, T.L. Ed. , Desert´ ´

Dust. Geological Society of America, Special Paper, vol. 186,pp. 87–100.

Shinn, E.A., Smith, G.W., Prospero, J.M., Betzer, P., Hayes,M.L., Garrison, V., Barber, R.T., 2000. African dust and thedemise of Caribbean coral reefs. Geophysical Research Letters27, 3029.

Sivall, T., 1957. Sirocco in the Levant. Geografiska Annaler 39,114–142.

Smalley, I.J., Krinsley, D.H., 1978. Loess deposits associated withdeserts. Catena 5, 53–66.

Smith, J.E., 1988. The dust fall of 27 October 1987 over Britain.Journal of Meteorology 13, 115–120.

Stabell, B., 1989. Deflation and humidity during the past 700,000years in NW Africa from the marine record. In: Leinen, M.,

Ž .Sarnthein, M. Eds. , Palaeoclimatology and Palaeometeorol-ogy: Modern and Past Patterns of Global Atmospheric Trans-port. Kluwer Academic Publishing, Dordrecht, pp. 771–777.

Stein, R., 1985. The post-Eocene sediment record of DSDP site366: implications for African climate and plate tectonic drift.Geological Society of America, Memoir 163, 305–315.

Stevenson, C.M., 1969. The dust fall and severe storms of 1 July1968. Weather 24, 126–132.

Swap, R., Garstang, M., Greco, S., Talbot, R., Kallberg, P., 1992.Saharan dust in the Amazon Basin. Tellus 44B, 133–149.

Swap, R., Ulanski, S., Cobbett, M., Garstang, M., 1996. Temporaland spatial characteristics of Saharan dust outbreaks. Journal

Ž .of Geophysical Research 101 D2 , 4205–4220.Sweeney, M., 1998. Saharan dust fall in western Ireland. Journal

of Meteorology 23, 140.Talbot, R.W., Harriss, R.C., Browell, E.V., Gregory, G.L., Se-

bacher, D.I., Beck, S.M., 1986. Distribution and geochemistryof aerosols in the tropical North Atlantic troposhere: relation-ship to Saharan dust. Journal of Geophysical Research 91Ž .D4 , 5173–5182.

Tegen, I., Fung, I., 1994. Modelling of mineral dust in theatmosphere: sources, transport and optical thickness. Journal

Ž .of Geophysical Research 99 D11 , 22879–22914.Tegen, I., Fung, I., 1995. Contribution to the atmospheric mineral

aerosol load from land surface modification. Journal of Geo-Ž .physical Research 100 D9 , 18707–18726.

Tetzlaff, G., Peters, M., 1986. Deep-sea sediments in the easternequatorial Atlantic off the African coast and meteorologicalflow patterns over the Sahel. Geologische Rundschau 75,71–79.

Thomas, F.G., 1982. Saharan dust-fall in Dover, Kent. Journal ofMeteorology 7, 92–93.

Thomas, F.G., 1983. Dust fall-out over Kent and Sussex. Journalof Meteorology 8, 126–127.

Thomas, F.G., 1985. Weather of 8 November 1984 and theSaharan dust-fall. Journal of Meteorology 10, 147–148.

Thomas, F.G., 1993. Dust rain. Weather 48, 193.Tomadin, L., Lenaz, R., Landuzzi, V., Mazzucotelli, A., Van-

nucci, R., 1984. Wind-blown dusts over the central Mediter-ranean. Oceanologica Acta 7, 13–23.

Tout, D.G., Kemp, V., 1985. The named winds of Spain. Weather40, 322–329.

Tsoar, H., Pye, K., 1987. Dust transport and the question of desertloess formation. Sedimentology 34, 139–153.

Tullet, M.T., 1978. A dust-fall on 6 March 1977. Weather 33,48–52.

Tullet, M.T., 1980. A dust-fall of Saharan origin on 15 May 1979.Journal of Earth Science Royal Dublin Society 3, 35–39.

Tullett, M.T., 1984. Saharan dust-fall in Northern Ireland. Weather39, 151–152.

Tullet, M.T., 1988. A dust fall on 1 September 1987. Weather 43,340.

Urvoy, Y., 1942. Les bassins du Niger. Memoire de L’InstitutFondemental d’Afrique Noire 4.

Vine, H., 1987. Wind-blown materials and W. African soils: anexplanation of the ‘ferralitic soil over loose sandy sediments’profile. Geological Society Special Publication 35, 171–183.

Wagenbach, D., Geis, K., 1989. The mineral dust record in a highŽ .alpine glacier Colle Gnifett, Swiss Alps . In: Leinen, M.,Ž .Sarnthein, M. Eds. , Paleoclimatology and Paleometeorology:

Modern and Past Patterns of Global Atmospheric Transport.Kluwer Academic Publishing, Dordrecht, pp. 543–564.

Wheeler, D.A., 1986. The meteorological background to the fallof Saharan dust, November 1984. Meteorological Magazine115, 1–9.

Wilke, B.M., Duke, B.J., Jimoh, W.L.O., 1984. Mineralogy andchemistry of Harmattan dust in northern Nigeria. Catena 11,91–96.

Wright, J., 2001. Making loess-sized silt: data from laboratorysimulations and implications for sediment transport pathwaysand the formation of ‘desert’ loess deposits associated with theSahara. Quaternary International 76r77, 7–19.

Yaalon, D.H., 1987. Saharan dust and desert loess: effect ofsurrounding soils. Journal of African Earth Sciences 6, 569–571.

Yaalon, D.H., Dan, J., 1974. Accumulation and distribution ofloess-derived deposits in the semi-desert and desert fringe ofIsrael. Zeitschrift f ur Geomorphologie, Supplementband 20,¨91–105.

Yaalon, D.H., Ganor, E., 1973. The influence of dust on soilsduring the Quaternary. Soil Science 116, 146–155.

Yaalon, D.H., Ganor, E., 1979. East Mediterranean trajectories ofdust-carrying storms from the Sahara and Sinai. In: Morales,

Ž .C. Ed. , Saharan Dust. Wiley, Chichester, pp. 187–193.Zhang, X.Y., Arimot, R., An, Z.S., 1997. Dust emission from

Chinese desert sources linked to variations in atmosphericŽ .circulation. Journal of Geophysical Research 102 D23 ,

28041–28047.


Zhu, X.R., Prospero, J.M., Millero, F.J., 1997. Daily variability ofŽ .soluble Fe II and soluble total Fe in North African dust in the

trade winds at Barbados. Journal of Geophysical Research 102Ž .D17 , 21297–21305.

Andrew Goudie was born in Cheltenham, England in 1945 andwas educated at Cambridge University before being appointed toOxford in 1970. He is Professor of Geography and a formerPro-Vice-Chancellor. He has worked extensively on deserts inAfrica, South Asia and the Middle East.

Nick Middleton was born in London, England in 1960. Havingreceived his PhD in 1986 from Oxford University, he remainsthere in the School of Geography and the Environment as aFellow at St. Anne’s College. His interest in dryland geomorphol-ogy continues to be a strong theme in his work and his researcheshave taken him to the Sahara, Gobi, Thar, Namib and Kalahari.The human use, as well as the nature of deserts and their marginsis also a major research theme and Nick has edited the twoeditions of the UN Environment Programme’s World Atlas ofDesertification.