Geology

doi: 10.1130/0091-7613(2002)030<0315:SDOOIM>2.0.CO;2 2002;30;315-318Geology

 Gabriel J. Bowen and Bruce Wilkinson 

O in meteoric precipitation18δSpatial distribution of   

Email alerting servicescite this article

to receive free e-mail alerts when new articleswww.gsapubs.org/cgi/alertsclick

Subscribe to subscribe to Geologywww.gsapubs.org/subscriptions/click

Permission request to contact GSAhttp://www.geosociety.org/pubs/copyrt.htm#gsaclick

viewpoint. Opinions presented in this publication do not reflect official positions of the Society.positions by scientists worldwide, regardless of their race, citizenship, gender, religion, or politicalarticle's full citation. GSA provides this and other forums for the presentation of diverse opinions and articles on their own or their organization's Web site providing the posting includes a reference to thescience. This file may not be posted to any Web site, but authors may post the abstracts only of their unlimited copies of items in GSA's journals for noncommercial use in classrooms to further education andto use a single figure, a single table, and/or a brief paragraph of text in subsequent works and to make

GSA,employment. Individual scientists are hereby granted permission, without fees or further requests to Copyright not claimed on content prepared wholly by U.S. government employees within scope of their

Notes

Geological Society of America

on May 25, 2014geology.gsapubs.orgDownloaded from on May 25, 2014geology.gsapubs.orgDownloaded from

q 2002 Geological Society of America. For permission to copy, contact Copyright Permissions, GSA, or editing@geosociety.org.Geology; April 2002; v. 30; no. 4; p. 315–318; 3 figures. 315

Spatial distribution of d18O in meteoric precipitationGabriel J. Bowen Department of Earth Sciences, University of California, Santa Cruz, California 95064, USABruce Wilkinson Department of Geological Sciences, University of Michigan, Ann Arbor, Michigan 48109, USA

ABSTRACTProxy data reflecting the oxygen isotope composition of meteoric precipitation (d18Oppt)

are widely used in reconstructions of continental paleoclimate and paleohydrology. How-ever, actual geographic variation in modern water compositions is difficult to estimatefrom often sparse data. A first step toward understanding the geologic pattern of changein d18Oppt is to describe the modern distribution in terms of principal geographic param-eters. To this end, we empirically model relationships between 18O in modern precipitationand latitude and altitude. We then identify geographic areas where large-scale vaportransport patterns give rise to significant deviations from model d18Oppt compositionsbased on latitude and altitude. Model value and residual grids are combined to derive ahigh-resolution global map of d18Oppt that can serve as a spatial reference against whichproxy data for paleoprecipitation can be compared. Reiteration of the procedure outlinedhere, for paleo-d18Oppt data, may illuminate past changes in the climatic and physiograph-ic parameters controlling the distribution of d18O regimes.

Keywords: isotopes, oxygen, paleoclimates, precipitation.

INTRODUCTIONThe oxygen isotope composition of mete-

oric water as recorded in authigenic mineralsand biogenic hardparts has been used exten-sively in studies of ancient continental climateand hydrology. Important records have beenderived from ice cores (e.g., Dansgaard et al.,1993; Thompson et al., 1995; Petit et al.,1999), groundwater (e.g., Rozanski, 1985;Dutton, 1995), speleothems (e.g., Schwarcz,1986; Dorale et al., 1992; Denniston et al.,1999), meteoric calcite cements (e.g., Haysand Grossman, 1991), freshwater mollusks(e.g., Dettman and Lohmann, 1993), fossilenamel (e.g., Bryant et al., 1996; Fricke et al.,1998; Sharp and Cerling, 1998), soil carbonate(e.g., Koch et al., 1995; Amundson et al.,1996; Cerling and Wang, 1996), iron-oxideminerals (e.g., Bao et al., 1998, 1999), andauthigenic clays (e.g., Chamberlain et al.,1999; Chamberlain and Poage, 2000). Theserecords are frequently used to estimate paleo-temperature based on the strong spatial cor-relation between modern d18Oppt and localmean annual temperature (Dansgaard, 1964).

It has become increasingly apparent thatpast changes in atmospheric circulation andother climatic parameters may render simpletemperature-based interpretation of paleo-d18Oppt data incorrect (Amundson et al., 1996;Edwards et al., 1996; Boyle, 1997; Fricke andO’Neil, 1999; Pierrehumbert, 1999). Theoret-ical models for d18Oppt, combining Rayleighdistillation with a more or less comprehensivetreatment of the meteorological evolution ofindividual air parcels, have been developedand applied to individual precipitation events(Covey and Haagenson, 1984) and locations(Dansgaard, 1954; Boyle, 1997; Pierrehum-bert, 1999). Similar treatments, coupled toglobal circulation models (GCMs), have pro-

duced reasonably accurate, low-resolutionmodel d18Oppt maps for the modern Earth andestimated values for the Last Glacial Maxi-mum (Joussaume et al., 1984; Jouzel et al.,1994; Hoffmann and Heimann, 1997). Appli-cation of theoretical models to paleo-d18Oppt

proxy data, however, is limited by inadequateknowledge of relevant meteorological param-eters and uncertainties regarding GCM repro-duction of past climates.

Despite the abundance of work that has ad-dressed relationships between climate variablesand d18Oppt, relatively little research has fo-cused on the geographic distribution of d18Oppt

and its variation through time. To address theseissues, a modern spatial reference framework isneeded against which ancient values can becompared. Here we develop the necessaryframework as a high-resolution map of d18Oppt

based on an empirically derived model incor-porating basic geographic parameters.

DATAOur effort begins with data from the third

release of the International Atomic EnergyAgency–World Meteorological OrganizationGlobal Network for Isotopes in Precipitation(GNIP) database (IAEA/WMO, 1998). The da-tabase includes data from 583 stations, althoughfewer than half of these have oxygen isotoperecords representing one or more years (see Roz-anski et al., 1993, for a complete review of anearlier release of this data set). Model equationswere derived from station geographic data andfrom average amount-weighted annual d18Oppt

values that were calculated from the month-by-month GNIP isotopic and precipitation measure-ments. To calculate annual averages, amount-weighted mean values were calculated for eachmonth for all years represented at each station.These monthly d18Oppt values were used to de-

rive an amount-weighted mean annual d18Oppt

for each of 232 stations. Other routines for cal-culating average annual d18Oppt were exploredand do not produce results significantly differentfrom those presented here. The high-resolutionmap of d18Oppt was created by using topographicdata from the U.S. Geological Survey ETOPO5digital elevation model (DEM), with 5 min lat-itudinal and longitudinal resolution (U.S. Na-tional Geophysical Data Center, 1998).

APPROACHPreviously, estimation of d18Oppt at a given

geographic location has been accomplished byinterpolation between GNIP stations or by ref-erence to the nearest station. However, the ox-ygen isotope composition of precipitation iscontrolled by Rayleigh distillation of atmo-spheric vapor, driven primarily by changes inair-mass temperature (e.g., Yurtsever, 1975;Rozanski et al., 1993), and interpolated esti-mates ignore the geographic parameters thatcontrol temperature (latitude and altitude). Weincorporate these variables by using the globalmean relationships between latitude and alti-tude and d18Oppt to predict local d18Oppt.Large-scale patterns of atmospheric vaportransport also affect d18Oppt by changing thelength and origin of vapor-transport pathways(Amundson et al., 1996; Edwards et al., 1996;Boyle, 1997). These effects appear as regionaldeviations of GNIP station observations fromthe predictions based on latitude and altitude.We incorporate vapor-transport effects in ourbest estimate d18Oppt map by spatially inter-polating the residuals from our altitude andlatitude model and adding those to the mapgrid generated from the model.

MODEL DERIVATIONWe used a two-step regression technique to

deconvolve the effects of latitude and altitude ond18Oppt. The negative correlation betweend18Oppt and the absolute value of station latitude(zLATz, Fig. 1) results from the cooling and dis-tillation of water vapor during transport fromlow-latitude regions toward the poles. Data fromstations at ,200 m elevation (n 5 155) wereisolated to evaluate latitudinal variation ind18Oppt values while minimizing the complicat-ing effects of topography. The relationship be-tween low altitude station zLATz (independentvariable) and d18Oppt (dependent variable) is bestdescribed by the second-order polynomial:

18 2d O 5 20.0051(zLATz)ppt

1 0.1805(zLATz) 2 5.247 (1)

on May 25, 2014geology.gsapubs.orgDownloaded from

316 GEOLOGY, April 2002

Figure 1. d18O of meteoric precipitation plotted against station latitude for 232 GNIP stations,with best-fit polynomial for low-altitude stations (<200 m; solid) and model lines for 1–5 kmelevations. Model equation is shown.

(solid line, Fig. 1) for which r2 5 0.80. Theshape of this trend is largely the result of thenonlinear relationship between latitude andtemperature and is exaggerated by the nega-tive correlation between annual precipitationamount and d18Oppt in the tropics (e.g., Roz-anski et al., 1993).

There is a tendency for the d18Oppt valuesof stations above 200 m to be below the low-altitude station trend (Fig. 1), and deviationsof measured values from the low-altitudetrend systematically increase with increasingstation altitude (ALT, meters). The associationbetween these parameters is a function of thecooling of air masses as they rise to higherelevations and is roughly linear. To derive therelationship between altitude and d18Oppt, wefirst used equation 1 to estimate d18Oppt valuesfor all stations based on their latitudes. Thenegative correlation between d18Oppt and al-titude (oxygen isotope lapse rate, OILR) wasquantified by least-squares regression betweenstation altitudes (independent variable) andthe residuals from equation 1 (DEV, depen-dent variable), where the Y-intercept was fixedat 0. This approach allowed us to examine thelatitude-independent effects of altitude ond18Oppt. The best-fit equation is

DEV 5 20.002(ALT), (2)

where r2 5 0.17. This yields an oxygen iso-tope lapse rate of 2‰/km, a value almostidentical to that calculated by Chamberlainand Poage (2000). The slope of this relation-ship has a standard error of 60.0002 and ishighly significant at the 95% confidence level(p , 0.0001).

Our model equation, describing d18Oppt forthe continents as a function of latitude and al-titude, is the sum of equations 1 and 2:

18 2d O 5 20.0051(zLATz)ppt

1 0.1805(zLATz)

2 0.002(ALT)

2 5.247. (3)

This formulation assumes that the oxygen iso-tope lapse rate is similar at all geographic lo-cations, a necessary simplification given thecurrent lack of data from high altitudes. Theaverage difference between d18Oppt values es-timated from equation 3 and the measured val-ues is 20.21‰ (s 5 2.49‰). The relationshipbetween estimated and measured values is lin-ear with a slope of 1.04 and R2 5 0.76.

SPATIAL DISTRIBUTION OF MODELRESIDUALS

Equation 3 was applied to each GNIP sta-tion, and a second set of residuals was cal-culated that describes the difference betweenobserved values and those predicted by thecombined model. To investigate the spatialdistribution of these residuals, a 0.58 3 0.58grid was interpolated on a spherical surface.Residuals were weighted according to:

(2d8/48)weight 5 e , (4)

where d is the angular distance between a gridpoint and data point. This function allows rel-atively smooth interpolation of unevenlyspaced data and highlights variation over dis-

tances of 102 to 103 km, a scale at which theeffects of large-scale vapor-transport patternsshould be evident.

The spatial distribution of interpolated re-siduals (Fig. 2) shows significant variation atthe 102 to 103 km scale. Low-magnitude in-terpolated residuals cannot be consideredmeaningful across regions with sparse data(e.g., Africa and Siberia). In other regions(e.g., South America, southern North Ameri-ca, the Mediterranean, and eastern Europe),however, low-magnitude residuals are inter-polated from abundant data and confirm thatd18Oppt is primarily dependent on latitudinaland altitudinal temperature variations. In ad-dition, there are large oceanic areas (e.g., thenorthern Atlantic and the western Pacific)where the model is in good agreement withobservations made at numerous, but widelyseparated, island stations.

Regions of high-magnitude residuals occurat northern middle to high latitudes and likelyreflect the zonal heterogeneity of vapor trans-port. In particular, the latitude and altitudemodel significantly overestimates the d18O ofprecipitation over Canada, while underesti-mating d18Oppt in the Norwegian and BarentsSeas. Comparison of the residual map withvertically averaged patterns of vapor transport(Peixoto and Oort, 1992) suggests that resid-uals in the North Atlantic may relate to north-ward and eastward transport of moistureacross the warm North Atlantic, a process thatwould provide relatively 18O enriched waterto northern coastal Europe. Conversely, over-estimation of d18Oppt in Canada may be as-sociated with a component of 18O-depleted,Arctic-derived moisture and with eastwardtransport of moisture across the cool Pacificeastern boundary currents before reaching thecontinent.

In some cases, adequate explanation of theresiduals will require additional data. Theanomaly across East Africa, for example, re-flects the fact that water compositions fromtwo high-altitude stations (Addis Ababa at2.36 km and Entebbe at 1.16 km) are close tothose anticipated for sea-level stations at thatlatitude. Recently published d18Oppt measure-ments from a transect (10–4050 m altitude, n5 20) in equatorial East Africa (Gonfiantiniet al., 2001) are heavier than the values pre-dicted by equation 3 according to:

18 18d O 2 d Omeasured model

5 0.0004(ALT) 1 1.371, (5)

where the r2 5 0.57 and the slope is signifi-cant at 95% confidence with p , 0.001. Thissuggests that the oxygen isotope lapse rate inthis region is 20% lower than the global av-erage and, although the East African anomalymay characterize precipitation at all altitudes,

on May 25, 2014geology.gsapubs.orgDownloaded from

GEOLOGY, April 2002 317

Figure 2. GNIP stations (shaded diamonds) with model residuals contoured at 1‰ intervals. Note significant residual valuesover East Africa and at high latitudes over northern North America and North Atlantic.

Figure 3. Map of d18Oppt calculated for continents at 59 3 59 geographic resolution. Map values are sum of model-derived d18Opptestimates from latitude and altitude data (Fig. 1) and spatially interpolated model residuals (Fig. 2). Color bands represent 1‰intervals between 230‰ and 0‰.

its magnitude in Figure 2 is exaggerated be-cause data from this region are from high-alti-tude stations.

SPATIAL DISTRIBUTION OF d18O INMETEORIC PRECIPITATION

A 59 3 59 grid of equation 3 residuals wasinterpolated from the 0.58 3 0.58 grid just dis-cussed. This grid was added to a 59 3 59 grid

generated from application of equation 3 tothe ETOPO5 DEM grid points, producing amap grid representing our best estimation ofd18Oppt (Fig. 3). The pattern of d18Oppt distri-bution is similar to that previously document-ed (Yurtsever, 1975), but with considerable re-finement. The map reproduces the d18Oppt ofcontinental GNIP stations closely, with an av-erage error of 0.08‰ (s 5 0.85‰, n 5 168).

The relationship between map d18Oppt valuesand measurements is linear with a slope of1.00 and R2 5 0.97. To test the predictivepower of the map, we compared d18Oppt datafrom Goldstone, California (35.358N,116.898W, 920 m; Friedman et al., 1992), andAmes, Iowa (42.038N, 93.688W, 287 m; Simp-kins, 1995), to Figure 3 values and to esti-mates from linear interpolation between the

on May 25, 2014geology.gsapubs.orgDownloaded from

318 GEOLOGY, April 2002

nearest GNIP stations. Map and linearly inter-polated estimated values at Goldstone were27.6‰ and 27.1‰, respectively, comparedto the measured value of 210.4‰. At Ames,map and interpolated predictions were 28.9‰and 29.9‰, respectively, and the measuredvalue was 28.0‰. In each case, the map val-ues more closely approximate measurements.The estimation of confidence limits for mappredictions will be the focus of future work.

The map shown as Figure 3 affords severalpotential uses. Because the map accounts forthe effect of local topography on d18Oppt, tem-poral change in d18Oppt at a single location maybe better evaluated with reference to map val-ues than interpolated estimates. Second, themap depicts modern, globally averaged rela-tionships between d18Oppt and a number of ex-plicit (latitude, altitude) and more obscure (va-por sources, storm tracks) geographic variables.Reiteration of the process described here forhigh-quality paleo-d18Oppt proxy data repre-senting a previous climate state will be a firststep toward understanding the stability ofthese relationships through time. Finally, be-cause regions where d18Oppt deviates signifi-cantly from the global mean geographic andphysiographic trends may reflect large-scalevapor-transport patterns, recognition of suchareas in a data set for paleo-d18Oppt couldhighlight temporal variations in atmosphericor oceanic circulation.

ACKNOWLEDGMENTSWe thank the University of California, Santa Cruz, pa-

leoclimate modeling group and Justin Revenaugh for assis-tance with data processing and display. Early drafts of themanuscript benefited from reviews by C. Page Chamber-lain, Andrea Dutton, David Fox, Linda Ivany, Paul Koch,and Michal Kowalewski. Bowen was supported by the Na-tional Science Foundation Graduate Research FellowshipProgram.

REFERENCES CITEDAmundson, R.G., Chadwick, O.A., Kendall, C., Wang, Y.,

and DeNiro, M.J., 1996, Isotopic evidence for shiftsin atmospheric circulation patterns during the lateQuaternary in mid-North America: Geology, v. 24,p. 23–26.

Bao, H., Koch, P.L., and Hepple, R.P., 1998, Hematite andcalcite coatings on fossil vertebrates: Journal of Sed-imentary Research, v. 68, p. 727–738.

Bao, H., Koch, P.L., and Rumble, D., III, 1999, Paleocene-Eocene climatic variation in western North America;evidence from the d18O of pedogenic hematite: Geo-logical Society of America Bulletin, v. 111,p. 1405–1415.

Boyle, E.A., 1997, Cool tropical temperatures shift theglobal d18O-T relationship: An explanation for the icecore d18O-borehole thermometry conflict?: Geophys-ical Research Letters, v. 24, p. 273–276.

Bryant, J.D., Koch, P.L., Froelich, P.N., Showers, W.J., and

Genna, B.J., 1996, Oxygen isotope partitioning be-tween phosphate and carbonate in mammalian apa-tite: Geochimica et Cosmochimica Acta, v. 60,p. 5145–5148.

Cerling, T.E., and Wang, Y., 1996, Stable carbon and ox-ygen isotopes in soil CO2 and soil carbonate: Theory,practice, and application to some prairie soils of up-per midwestern North America, in Boutton, T.W., andYamasaki, S., eds., Mass spectrometry of soils: NewYork, Marcel Dekker, Inc., p. 217–231.

Chamberlain, C.P., and Poage, M.A., 2000, Reconstructingthe paleotopography of mountain belts from the iso-topic composition of authigenic minerals: Geology,v. 28, p. 115–118.

Chamberlain, C.P., Poage, M.A., Craw, D., and Reynolds,R.C., 1999, Topographic development of the South-ern Alps recorded by the isotopic composition of au-thigenic clay minerals, South Island, New Zealand:Chemical Geology, v. 155, p. 279–294.

Covey, C., and Haagenson, P.L., 1984, A model of oxygenisotope composition of precipitation: Implications forpaleoclimate data: Journal of Geophysical Research,v. 89, p. 4647–4655.

Dansgaard, W., 1954, The O18-abundance in fresh water:Geochimica et Cosmochimica Acta, v. 6,p. 241–260.

Dansgaard, W., 1964, Stable isotopes in precipitation: Tel-lus, v. 16, p. 436–468.

Dansgaard, W., Johnsen, S.J., Clausen, H.B., Dahl-Jensen,D., Gundestrup, N.S., Hammer, C.U., Hvidberg, C.S.,Steffensen, J.P., Sveinbjornsdottir, A.E., Jouzel, J.,and Bond, G., 1993, Evidence for general instabilityof past climate from a 250-kyr ice-core record: Na-ture, v. 364, p. 218–220.

Denniston, R.F., Gonzalez, L.A., Asmerom, Y., Baker, R.G.,Reagan, M.K., and Bettis, E.A., III, 1999, Evidencefor increased cool season moisture during the middleHolocene: Geology, v. 27, p. 815–818.

Dettman, D.L., and Lohmann, K.C, 1993, Seasonal changein Paleogene surface water d18O; fresh-water bi-valves of western North America, in Swart, P.K., etal., eds., Climate change in continental isotopic re-cords: American Geophysical Union GeophysicalMonograph 78, p. 153–164.

Dorale, J.A., Gonzalez, L.A., Reagan, M.K., Pickett, D.A.,Murrell, M.T., and Baker, R.G., 1992, A high-reso-lution record of Holocene climate change in speleo-them calcite from Cold Water Cave, northeast Iowa:Science, v. 258, p. 1626–1630.

Dutton, A.R., 1995, Groundwater isotopic evidence forpaleorecharge in U.S. High Plains aquifers: Quater-nary Research, v. 43, p. 221–231.

Edwards, T.W.D., Wolfe, B.B., and MacDonald, G.M.,1996, Influence of changing atmospheric circulationon precipitation d18O-temperature relations in Cana-da during the Holocene: Quaternary Research, v. 46,p. 211–218.

Fricke, H.C., and O’Neil, J.R., 1999, The correlation be-tween 18O/16O ratios of meteoric water and surfacetemperature: Its use in investigating terrestrial climatechange over geologic time: Earth and Planetary Sci-ence Letters, v. 170, p. 181–196.

Fricke, H.C., Clyde, W.C., and O’Neil, J.R., 1998, Intra-tooth variations in d18O (PO4) of mammalian toothenamel as a record of seasonal variations in conti-nental climate variables: Geochimica et Cosmochim-ica Acta, v. 62, p. 1839–1850.

Friedman, I., Smith, G.I., Gleason, J.D., Warden, A., andHarris, J.M., 1992, Stable isotope composition of wa-ters in southeastern California: 1. Modern precipita-tion: Journal of Geophysical Research, v. 97,p. 5795–5812.

Gonfiantini, R., Roche, M.A., Olivry, J.C., Fontes, J.C., andZuppi, G.M., 2001, The altitude effect on the isotopic

composition of tropical rains: Chemical Geology,v. 181, p. 147–167.

Hays, P.D., and Grossman, E.L., 1991, Oxygen isotopes inmeteoric calcite cements as indicators of continentalpaleoclimate: Geology, v. 19, p. 441–444.

Hoffmann, G., and Heimann, M., 1997, Water isotope mod-eling in the Asian monsoon region: Quaternary In-ternational, v. 37, p. 115–128.

IAEA/WMO, 1998, Global network for isotopes in pre-cipitation. The GNIP database. Release 3, October1999: URL: http://www.iaea.org/programs/ri/gnip/gnipmain.htm (11/2000).

Joussaume, S., Sadourny, R., and Jouzel, J., 1984, A gen-eral circulation model of water isotope cycles in theatmosphere: Nature, v. 311, p. 24–29.

Jouzel, J., Koster, R.D., Suozzo, R.J., and Russell, G.L.,1994, Stable water isotope behavior during the LastGlacial Maximum: A general circulation model anal-ysis: Journal of Geophysical Research, v. 99,p. 25 791–25 801.

Koch, P.L., Zachos, J.C., and Dettman, D.L., 1995, Stableisotope stratigraphy and paleoclimatology of the Pa-leogene Bighorn Basin (Wyoming, USA): Palaeo-geography, Palaeoclimatology, Palaeoecology,v. 115, p. 61–89.

Peixoto, J.P., and Oort, A.H., 1992, Physics of climate: NewYork, American Institute of Physics, 520 p.

Petit, J.R., Jouzel, J., Raynaud, D., Barkov, N.I., Barnola,J.M., Basile, I., Bender, M., Chappellaz, J., Davis,M., Delaygue, G., Delmotte, M., Kotlyakov, V.M.,Legrand, M., Lipenkov, V.Y., Lorius, C., Pepin, L.,Ritz, C., Saltzmann, E., and Stievenard, M., 1999,Climate and atmospheric history of the past 420 000years from the Vostok ice core, Antarctica: Nature,v. 399, p. 429–436.

Pierrehumbert, R.T., 1999, Huascaran d18O as an indicatorof tropical climate during the Last Glacial Maximum:Geophysical Research Letters, v. 26, p. 1345–1348.

Rozanski, K., 1985, Deuterium and oxygen-18 in Europeangroundwaters; links to atmospheric circulation in thepast: Chemical Geology, v. 52, p. 349–363.

Rozanski, K., Araguas-Araguas, L., and Gonfiantini, R.,1993, Isotopic patterns in modern global precipita-tion, in Swart, P.K., et al., eds., Climate change incontinental isotopic records: American GeophysicalUnion Geophysical Monograph 78, p. 1–36.

Schwarcz, H.P., 1986, Geochronology and isotopic geo-chemistry of speleothems, in Fritz, P., and Fontes,J.C., eds., Handbook of environmental isotope geo-chemistry: Amsterdam, Elsevier, p. 271–304.

Sharp, Z.D., and Cerling, T.E., 1998, Fossil isotope recordsof seasonal climate and ecology—Straight from thehorse’s mouth: Geology, v. 26, p. 219–222.

Simpkins, W.W., 1995, Isotopic composition of precipita-tion in central Iowa: Journal of Hydrology, v. 172,p. 185–207.

Thompson, L.G., Mosley-Thompson, E., Davis, M.E., Lin,P.N., Henderson, K.A., Cole-Dai, J., Bolzan, J.F., andLiu, K.B., 1995, Late glacial stage and Holocenetropical ice core records from Huscaran, Peru: Sci-ence, v. 269, p. 46–50.

U.S. National Geophysical Data Center (NGDC), 1998,ETOPO-5 five minute gridded world elevation: Boul-der, Colorado, USA, NGDC, URL: http://www.grid.unep.ch/datasets/earth.html (11/2000).

Yurtsever, Y., 1975, Worldwide survey of stable isotopesin precipitation: Vienna, International Atomic En-ergy Agency, Report of the Isotope Hydrology Sec-tion, 40 p.

Manuscript received September 20, 2001Revised manuscript received December 5, 2001Manuscript accepted December 6, 2001

Printed in USA

on May 25, 2014geology.gsapubs.orgDownloaded from