Downloaded from http://sp.lyellcollection.org/ at Pennsylvania State University on March 5, 2016

Luminescence dating of sand from the Kelso Dunes, California STEPHEN

R. E D W A R D S

Institute of Earth Studies, University College of Wales, Aberystwyth, Dyfed SY23 3DB, UK

Abstract: A relatively new luminescence dating technique, infra-red stimulated luminescence (IRSL), has been used to date dune sand from the Kelso Dunes, eastern Mojave Desert, California. This is the first time the technique has been applied to desert aeolian sands. The IRSL technique measures the time since sediment grains were last exposed to light, by sampling a highly light-sensitive signal which reduces rapidly to a negligible level when so exposed. The technique can be used to date very young samples. Of the ten samples dated, the surface sample gave an age of 40 + 17 years, while subsurface samples gave high-precision middle and late Holocene dates. These results, obtained from the upper 8 m of the sand mass, indicate significant aeolian activity during the later part of the Holocene. Clustering of IRSL ages from the data so far available suggests that the aeolian activity may have been episodic.

The principal aim of this study is to evaluate the relatively new luminescence dating technique of infra-red stimulated luminescence (IRSL) when applied to dune sand from a desert environment. The technique, in common with other luminescence techniques, measures the time since samples were last exposed to light. It is, therefore, likely to be suitable for determining the timing of periods of aeolian activity and dune formation. As with almost all desert dune areas, no direct, independent age control is available for the Kelso Dunes in the Mojave Desert; however, the dunefield was considered a good test area because there were areas with different dune orientation, indicating different periods of aeolian activity.

Geological background The Kelso Dunes are located at latitude 34~ longitude 115~ in the eastern Mojave Desert, San Bernardino County, California, and comprise, in terms of sand volume and height, the largest dunefield in the Mojave Desert (Figs 1 & 2). The Mojave River, draining north from the San Bernardino Mountains, is the only through-flowing river of any consequence in the Mojave; today floodwaters from heavy precipitation in the river's source area only fill the river's terminal basin, Silver Lake playa, in exceptionally wet years. On emerging from Alton Canyon (Fig. 1), the Mojave River forms a plain of active alluviation from which a narrow aeolian sand sheet - - the Devil's Playground - extends for about 55 km to the ESE. At its eastern end, the sand sheet merges into the Kelso Dunes. The source of sand for the dunes is thought to be the Mojave River alluvial plain

(Sharp 1966); however, unconsolidated sands on the floors of the Soda and Silver Lake playas, exposed during former low lake level stands, may also have formed an important source (Fig. 1; Lancaster 1990). The Kelso Dunes lie in the Kelso Valley, surrounded by mountains on their southern, eastern and northern sides. The dune mass is elliptically-shaped, covers roughly 175 km 2 and contains c. 4 km 3 of sand. The sand is we|l rounded, well sorted and rather coarse relative to typical dune sand, with 90% of grains between 0.25 and 0.50 mm in diameter. Active dunes are mostly confined to the SE portion of the dunefield; within the active area, three nearly parallel, complex linear ridges, bearing N 65~ constitute the largest dunes (Fig. 3). The southernmost of the linear ridges is the largest, being around 7 km long and rising nearly 170 m above the alluvial fan to the south (Fig. 2). Surrounding and partly superimposed upon the linear ridges is an irregular complex of medium-sized crescentic dunes. Areas of sand stabilized by vegetation occur primarily on the lower slopes of the dune mass. In these areas, relatively stable and subdued crescentic dunes form strongly linear patterns of differing orientations (Fig. 3). A 25-35% vegetation cover, including large creosote bushes, is typical of the relatively stable areas. Prevailing winds in the Kelso Dunes are from the WNW, but the work performed by these pervasive sand-transporting winds is locally counterbalanced by infrequent but strong orographically-controlled winds from other directions (Sharp 1966). Despite a considerable part of the dunefield being currently active, the Kelso Dunes are at the present time undergoing stabilization.

From Pye, K. (ed.), 1993, The Dynamics and EnvironmentalContextof Aeolian SedimentarySystems. Geological Society Special Publication No. 72, pp. 59-68.

Downloaded from http://sp.lyellcollection.org/ at Pennsylvania State University on March 5, 2016

60

S. R. EDWARDS

N T 0

10

~xSilver Lake ~Playa

20 km

to

o

Q~'~'

J

.Baker

~ S o d a Lake

West Cronese

Lake Playa~ ~=.EastCronese ~ ~ Lake Playa ~ ~ , / ~ Afton

I

, . - - ~ ' - ~, ~~/,_, / _

i!i!i!iiiiliiii!iii!: ii!i;iiii i

.... i

iil iiiiiiliiiii Los 9 ( :::::::::::::::::::::::::::::::::::: geles KELSO -.. :.:-:-:. :!:!:!:!~~~

i~i~::i::iii::i::i::i::iii::i:/:ilililiiii~iii~iiii::iii::i~:~ ....

DUNES

!iiiiiiii!i!!iii!i!iiiiiiiiiiiiiii!i!iiiiiii!i!iiiiiiiiiiii.. !i!i!i!i!i!i!i!!!i!!!!!!!!!!!!!!!!!!!!!!!!!!!!!i "0 !!!i !i!~i!i!i!i!i!i!:. '

'

'

~iiiiiX ..........

/

/ I~.

)

~

/

Playa ~

~

///,4/

~Tf'///J'/~

/

/f

b ,#

//J

~o~7 q~ K E ,_ s o

~

/ ,/~//XX///////x/" ~

MOUNTAINS

~o,r

#~f~"

~"

~ GRANITE MOUNTAINS

i"

Fig. 1. Location map for the Kelso Dunes, eastern Mojave Desert, California.

By studying Cottonwood Wash, a major wash which cuts through the Kelso Dunes, Smith (1967) concluded that the active dunes represent a relatively recent renewal of m o v e m e n t through rejuvenation of a once entirely stabilized surface; the Wash was seen to post-date a main dune-building phase in the Kelso Dunes and predate the more recent time of reactivation.

Sampling strategy The objective of the sampling strategy was to obtain samples of representative dunes from areas with different dune orientation within the Kelso Dunes so as to maximize the possibility of collecting samples last exposed to light during

Fig. 2. The southernmost complex linear ridge in the Kelso Dunes, viewed from the NW. The Granite Mountains and an alluvial fan surface covered with regularly spaced creosote bushes can be seen in the background.

Downloaded from http://sp.lyellcollection.org/ at Pennsylvania State University on March 5, 2016

LUMINESCENCE DATING OF SAND FROM THE KELSO DUNES different dune-forming episodes. A Landsat image of the Kelso Dunes (Fig. 3) was used to guide the selection of sampling sites. All but one set of samples were selected from areas of stabilized sand, the samples being obtained by augering to a depth of up to 8 m; one set of samples was taken from one of the currently active linear ridges. The use of an auger ensured that the likelihood of many samples being affected by recent reworking of dune crests would be minimized. Care was taken not to expose samples to any light on collection. A brief description of the sites from which analysed samples were taken is given in Table 1.

Luminescence dating Both the thermoluminescence (TL) and the IRSL techniques measure the time since

61

sediments were last exposed to light, as such exposure removes the geologically acquired luminescence signal. Light, however, does not remove the entire TL signal, so that, in TL sediment dating, an assessment of the T L remaining after deposition, known as the 'residual', is required. Grains can generally be assumed to have had a long exposure to sunlight prior to burial in an environment such as that of the study area, thereby reducing the uncertainty in the residual value determined in the laboratory. In IRSL dating, residuals need not be assessed, as the signal sampled is highly sensitive to light, rapidly reducing to a negligible level on exposure. IRSL dates are here obtained on potassium feldspars; no IRSL signal has been observed from quartz (Spooner et al. 1990). For comparison, two samples were also dated using the

Fig. 3. Landsat image of the Kelso Dunes. Light-coloured areas indicate the presence of mobile sand and dark coloured areas of stabilized sand. Within the main active area, three parallel, complex linear ridges are clearly seen; elsewhere, stabilized crescentic dunes form strongly linear patterns of differing orientations. The northernmost Granite Mountains and a number of washes are visible; the Union Pacific railroad cuts across the image. Width of field of view is approximately 20 km.

Downloaded from http://sp.lyellcollection.org/ at Pennsylvania State University on March 5, 2016

62

s . R . EDWARDS

Table 1. Location of analysed samples within the Kelso Dunes* Sample

Location

Kt

Sample collectedfrom top 1 mm of surface within Unit II, an area of stabilized sand Sample located at 0.5 m depth within Unit IX, an area of stabilized sand in the NW part of the dunefield exhibiting a surface rhombic waffle pattern Sample located at 5 m depth within Unit IX, an area of stabilized sand in the NW part of the dunefield exhibiting a surface rhombic waffle pattern Sample located at 1.1 m depth within Unit XIII, an area of stabilized sand exhibiting crescentic dunes aligned NW-SE Sample located at 8 m depth from within a crestal hollow in the southernmost active linear ridge; the sample is located within Unit XI, an area of active sand Sample located at 3 m depth within Unit VII, an area of stabilized sand forming a hollow north of the southernmost linear ridge and mostly surrounded by active sand Sample located at 8 m depth within Unit VII, an area of stabilized sand forming a hollow north of the southernmost linear ridge and mostly surrounded by active sand Sample located at 8 m depth within Unit II, an area of stabilized sand in the NE part of the dunefield exhibiting crescentic dunes aligned NNE-SSW Sample located at 5 m depth within Unit XIV, an area of stabilized sand exhibiting a surface rhombic waffle pattern Sample located at 8 m depth within Unit XIV, an area of stabilized sand exhibiting a surface rhombic waffle pattern

K7

K10

K12

K20

K21

K24

K29

K35

K36

Sample preparation and luminescence measurements

* Table to be used in conjunction with Figs 3 & 5.

TL technique, t h e TL dates being likewise obtained on potassium feldspars; in TL dating, potassium feldspars hold a n u m b e r of advantages over quartz as a dosimeter (see Wintle, this volume). The age of a sample can be calculated from the age equation: Age = Equivalent dose

natural TL; and (2) the annual dose is the radiation dose delivered annually since burial from the material surrounding the sample (the external dose) and from the sample itself (the internal dose). The equivalent dose (ED) was determined using the Additive Dose method: a luminescence growth curve, establishing the sensitivity of a sample's luminescence response to dose, is obtained by adding known laboratory doses in addition to the sample's natural dose; the E D is then found by extrapolating the growth curve to its intercept with either: (i) the residual (i> 1 day's sunlight exposure or equivalent) TL level for TL dating; or (ii) the dose axis for IRSL dating. IRSL signals were obtained by 'short shines' - - s h o r t (0.1 s - l . 0 s) exposures of an aliquot to IR stimulation. As the latent TL signal is not significantly reduced by a short shine, for the two samples on which TL dating was performed, multiple disc short shine IRSL dating was undertaken first on the prepared aliquots. A new IRSL dating technique, the Single Aliquot technique, was used to determine sample ages for all the analysed samples. The technique, which is fast to apply and requires less sample than multiple disc techniques as it allows (subject to the qualification below) E D determinations to be obtained from single aliquots of a sample, has been described by Duller (1991). The technique uses the Additive Dose method, growth curves being constructed for single aliquots by the addition to such aliquots of dose increments. A pre-heat correction is required to raise the luminescence signal corresponding to each cumulative dose level to its true value; this correction is determined on separate undosed aliquots of each sample.

(1)

Annual dose where: (1) the equivalent dose is the laboratoryadministered dose necessary to replicate the

Sample preparation was undertaken in subdued orange light. Samples were sieved to obtain the 180/.tm to 211/.tm grain size fraction; this fraction was then treated with 10% HC1 to remove carbonates, and 30% H202 to remove organics. The alkali feldspar-rich fraction of each sample was obtained by separating out subsamples of no more than 5 g using sodium polytungstate solution as the heavy liquid, set at a specific gravity of 2.62. The separated light fraction was itself separated using a sodium polytungstate solution set at a specific gravity of 2.58 so as to obtain the high potassium alkali feldspar-rich fraction. This light fraction was

Downloaded from http://sp.lyellcollection.org/ at Pennsylvania State University on March 5, 2016

LUMINESCENCE DATING OF SAND FROM THE KELSO DUNES

+1 +1 ~1 ~1 +1 +1 ~1 ~ ~1 WI ~1 r

"5

n~

+1 +[ +1 +1 +1 +1 +1 +1 +1 -H +1

~b +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1

• t ~, • ~-q oo

i~

~.1

+1 -I-I +1 -H + 1 ~ +1 -H +1 +1 +1

~.~

i~

~o~ o

~

~

o

-H-

~

+1 +1 +1 +1 +1 +1 +1 +1 +1 +1

~

~ ~.g

~ ~

=.~.~

§

~ ~ . ~ J

~g

0

;n 0 _ 0 0 o O ~ O

,~ ~ 0 ~

o

63

Downloaded from http://sp.lyellcollection.org/ at Pennsylvania State University on March 5, 2016

64

S. R. EDWARDS

then etched for 40 minutes in 10% HF to remove the outer alpha-irradiated layer of the grains, and then washed in 10% HC1 to remove any fluorides. Finally, the etched sample was sieved through a 100/~m diameter mesh to remove fines created by the break-up of feldspars during etching. Monolayers of grains from each etched sample were mounted on clean 10 mm diameter aluminium discs. Both the TL and the IRSL signals were measured on an automatic Rise Reader System (B0tter-Jensen et al. 1991); the photomultiplier tube used was a type EMI 9635B, the luminescence from the grains being observed through Corning 7-59 and Schott BG-39 filters. A supply of argon provided an inert atmosphere within the Ris0's glow oven/sample changer unit during measurement. Around 60 discs for each of two samples (K10 and K24) were prepared for multiple disc IRSL and TL ED determinations. Laboratory irradiations were performed using a Daybreak 9~ 9Oy beta source; laboratory bleaching was performed by a H6nle SOL2 solar simulator, with a spectrum closely matching that of sunlight. Each aliquot was given a 0.1 s short shine so as to obtain a multiple disc IRSL E D to compare with the TL ED. TL glow curves (plots of TL versus temperature) were then obtained for each disc, up to a temperature of 450~ using a ram~ rate of 3~ s -1 for sample K10 and 5~ s- for sample K24. Second-glow normalization was used to normalize both the IRSL and the TL signals, TL EDs being calculated over the temperature range c. 250-420~ Eleven samples were analysed using the Single Aliquot technique: for each sample, ten discs were prepared, six for ED determinations, four for ascertaining the pre-heat correction.

Laboratory irradiations were performed using the Ris0's 9~176 beta source. The IRSL was measured each time at 50~ for 1 s, an elevated temperature being used both to create a stable temperature for measurement and, together with the 'long' short shine chosen, to increase the IRSL signal (Duller & Wintle 1991) as the Kelso samples were expected to have low light levels. Six determinations of the E D were obtained for each sample.

Dosimetry The external dose-rates were determined (i) by counting alpha, beta and gamma emissions and (ii) by determining the bulk sample abundances of U, Th and K. In this way a comparison was made between dose-rates obtained by direct (emission counting) and indirect (element abundance) techniques; the direct techniques were considered preferable (for details, see Edwards 1991). The internal dose-rates were also obtained by the determination of the abundances of U, Th and K in the separated mineral fractions. Results are given in Table 2. The internal dose-rates form, on average, around 20% of the total dose-rates, the K contents being high (c. 10%) for all mineral separates; the total dose-rates are themselves very high (Table 2). Table 3 shows that the decay of 4~ accounts for about 90% of the external beta dose-rate; the dominance of K in contributing to the external beta dose-rate may be explained by the richly feldspathic nature of the bulk material: by weight, potassium feldspars account, on average, for one-fifth of the bulk samples. Conversely, only a very small proportion of each bulk sample is composed of silts and very fine sands.

Table 3. Contributions f r o m U, Th and K to the external beta dose-rate Sample

Kt K7 K10 K12 K20 K21 K24 K29 K35 K36a K36k

From U and Th * (pGy a -1)

fl Dose-rate (%)

From K? (ItGy a -1)

fl Dose-rate (%)

206 270 260 182 151 200 235 271 314 359 373

8 11 11 7 6 8 9 11 11 16 16

2255 2285 2190 2431 2431 2285 2285 2263 2532 1948 2007

92 89 89 93 94 92 91 89 89 84 84

* U and Th concentrations determined from a-counts by using the 'pairs' technique. t K concentrations determined using an atomic absorption technique.

Downloaded from http://sp.lyellcollection.org/ at Pennsylvania State University on March 5, 2016

LUMINESCENCE DATING OF SAND FROM THE KELSO DUNES

~x

t~

+1

+1

O

R O t-,l

~

+1

+1

'5

9,-..i I +1+1+1+1+1+1+1+1+1+1

+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1

.,..4

-t-I

+1

+1 ~

+1 tr

ea

-bO

+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1

NoN

+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1

[,,.

65

Downloaded from http://sp.lyellcollection.org/ at Pennsylvania State University on March 5, 2016

66

S. R. EDWARDS

Results The IRSL signals from Kelso samples were seen to be rapidly zeroed on exposure to light (Edwards 1991). The dose-rates, EDs and resultant ages for the analysed samples are given in Table 4, each sample's Single Aliquot ED being the arithmetic mean of six ED determinations. A representative Single Aliquot growth curve for one of the analysed samples is given in Fig. 4; each ED determination for this sample is also listed in the figure. Table 4 and Fig. 4 illustrate the high precision that is obtained for Kelso sample EDs by the Single Aliquot technique. The table also shows that the multiple disc IRSL, TL and Single Aliquot ages for sample K24 are in excellent agreement; however, there is at best only reasonable agreement between the three techniques for sample K10. The difference between the multiple disc IRSL and TL ages for K10 may have been reduced by the application of a more prolonged laboratory bleach time; however, it is not clear why the Single Aliquot technique results in an underestimation of the K10 age relative to the other techniques, given that no similar underestimation occurs for sample K24. The issue is dealt with further by Edwards (1991), where it is shown not to be a significant problem and, therefore, does not invalidate the discussion which follows. Figure 5 gives the Single Aliquot ages of

Single Aliquot EDs on Sample K36a Gy 6.86 -+0.44 7.30 -+0.08 6.59 -+0.09 8.35 t o.22 7.08_+0.11 6.90+0.16 Mean ED: 7.18-+0,62 Gy

S a m p l e K36a ~" 20000

._

J

10000-

9

rr

o

ED

9

'

;,

'

+

'

l'a

'

1'6

Dose (Gy)

Using the jackknifing technique, the ED equals 7,08 _,0.11 Gy

Fig. 4. A representative Single Aliquot growth curve and Single Aliquot ED determinations on a sample.

samples in relation to their location and depth; accepting these ages as valid (see above) it can be seen from the figure that sample ages increase with depth, there being no 'stratigraphical inversions' present, that the 'age' nf the surface sample is very low, that the samples are comparatively young (all dates indicate middle and late Holocene ages) and that within this time range, despite there being a considerable spread of ages, some clustering of ages is nevertheless discernible.

Discussion The very low 'age' of 40 _ 17 years for surface sediment shows that, in the field, a negligible residual level of luminescence is obtained for surface material, thus implying that the dating signal is reduced to negligible levels before burial - - one of the key attractions of IRSL dating. The luminescence dates so far obtained from subsurface samples are not inconsistent with the notion of episodic aeolian activity within the Kelso Dunes; the clustering of IRSL ages suggests possibly three main periods of late Holocene aeolian activity and deposition at c. 4000, c. 1500 and 450-800 years BP. If this is the case, the crescentic dunes of Unit II may have formed around 4000 years BP and the crescentic dunes of Unit XIV around 1500 years BP; extensive reworking of sands in Units VII, IX and XIII may have occurred in the period 450--800 years BP. The currently active complex linear ridges are being shaped by transverse winds at the present day (Sharp 1966); the age of the K20 sample (64 _+ 22 years), taken from the southernmost of these ridges, indicates that 8 m net thickness of sand has accumulated at this locality, by migration of superimposed dunes, in a very short period of time. As luminescence dates provide an estimate of the time since sediments were last buried, and as dune areas may be subject to periods of remobilization, it may prove impossible using luminescence dating techniques to estimate the age of the Kelso Dunes as a whole. The absence of dates older than 4000 years BP from the upper 8 m of the Kelso sand mass may be due to dune areas having been extensively reworked prior to this date. Work carried out elsewhere in the eastern Mojave Desert is consistent with this interpretation: in the Cronese Basin, 55 km WNW of the Kelso Dunes, desert varnish developed on wind-grooved bedrock surfaces after 5000 years BP indicates increased aeolian activity prior to this date (Dorn et al. 1989)

Downloaded from http://sp.lyellcollection.org/ at Pennsylvania State University on March 5, 2016

LUMINESCENCE DATING OF SAND FROM THE KELSO DUNES

0

All agesin years

5

7

~] K 7 ~ 102+45

K'~

[ I +~o [\~) ji"" t K3s~,498_227~.~ XIII ",\ //' L K36

1562+11:/ I

,"

,

67

"\\

~

V

~

I

VIII

//

~..

c~

,"~

I

~ S

)

II ; /

,,

~

v

t

1 \ ''''~ K29L"-~4074•

XI

Fig. 5. Location, depths and Single Aliquot ages of samples.

Conclusions

the precision offered by the technique correspondingly high. A very low 'age' of 40 • 17 years was obtained for surface sediment. Luminescence dates obtained from the upper 8 m of the sand mass indicate significant aeolian activity during the later part of the Holocene, with clustering of IRSL ages from the data so far available suggesting that this aeolian activity may have been episodic. The IRSL dating technique appears to work well for desert dune sands and holds much promise in unravelling geological problems in the future.

The good agreement obtained between the multiple disc TL and IRSL techniques for two analysed samples demonstrates the validity and applicability of the latter technique in dating dune sand from a desert environment; agreement between the multiple disc IRSL and the Single Aliquot techniques was also good for one of the two samples. Eleven subsamples were dated using the Single Aliquot technique; the scatter between ED determinations on individual aliquots of each sample was low,

The author wishes to thank Dr N. Lancaster of the Desert Research Institute, Reno, Nevada for extensive assistance in the field and for stimulating discussions; Dr H. M. Rendell of Sussex University for beta counting measurements; Dr A. G. Wintle for supervision of the project and G. A. T. Duller for generously providing software and practical help in the laboratory. The automated TL/IRSL reader was obtained with NERC grant GR3/8190. Fieldwork expenses were provided by NATO collaborative research grant 900151. This is publication number 236 of the Institute of Earth Studies, UCW, Aberystwyth.

while increased aridity is indicated for the period 5060--6800 years BP by palaeobotanical data obtained from packrat middens in the McCullough Range 100km NE of Kelso (Spaulding 1991). Evidence is available for dune formation and aridity elsewhere in the southwest United States during the middle Holocene, e.g. the Great Plains (Gaylord 1990) and the southern High Plains (Holliday 1989).

Downloaded from http://sp.lyellcollection.org/ at Pennsylvania State University on March 5, 2016

68

S. R. EDWARDS

References B~TTER-JENSEN, L., DITLEFSEN, C. • MEJDAHL, V. 1991. Combined OSL (Infrared) and TL Studies of Feldspars. Nuclear Tracks and Radiation Measurements, 18, 257-263. DORN, R.I., JULL, A.J.T., DONAHUE, D. J., LINICK, T. W. & TOOLIN, L. J. 1989. Accelerator Mass Spectrometry Radiacarbon Dating of Rock Varnish. Bulletin of the Geological Society of America, 101, 1363-1372. DULLER, G. 1991. Equivalent Dose Determination using Single Aliquots. Nuclear Tracks and Radiation Measurements, 18, 371-378. & WINTLE, A. G. 1991. On Infrared Stimulated Luminescence at Elevated Temperatures. Nuclear Tracks and Radiation Measurements, 18,379-384. EDWARDS, S. R. 1991. Luminescence Dating of Sand from the Kelso Dunes, California. Unpublished MSc Dissertation, University of Wales. GAYLORD, D. R. 1990. Holocene Palaeoclimatic Fluctuations revealed from Dune and Interdune Strata in Wyoming. Journal of Arid Environments, 18, 123-138. HOLLIDAY, V, T. 1989. Middle Holocene Drought on the Southern High Plains. Quaternary Research, 31, 74-82. LANCASTER, N. 1990. Dune Morphology and Chronology, Kelso Dunes, Mojave Desert,

California. Geological Society of America Abstracts with Programs 22, 1990 Annual Meeting, A86. PRESCOTr, J. R. & Hu'rroN, J. T. 1988. Cosmic Ray and Gamma Ray Dosimetry for TL and ESR. Nuclear Tracks and Radiation Measurements, 14, 223-227. SHARP, R. P. 1966. Kelso Dunes, Mojave Desert, California. Bulletin of the Geological Society of America, 77, 1045-1074. SMITH, H. T. U. 1967. Past Versus Present WindAction in the Mo]ave Desert Region, California. Air Force Cambridge Research Laboratories (USAF Bedford, Massachusetts) Publication AFCRL67-0683. SPAULDING, W. G. 1991. A Middle Holocene Vegetation Record from the Mojave Desert of North America and its Paleoclimatic Significance. Quaternary Research, 35,427-437. SPOONER, N. A., AITKEN,M. J., SMITH,B. W., FRANKS, M. & McELROV, C. 1990. Archaeological Dating by Infrared Stimulated Luminescence using a Diode Array. Radiation Protection Dosimetry, 34, 83-86. WINTLE, A. G. 1993. Luminescence dating of aeolian sands - - an overview. In. PYE, K. (ed.) The Dynamics and Environmental Context of A eolian Sedimentary Systems. Geological Society, London, special publication, 72, 49-58.