Top-down archaeology: High resolution satellite images of Rapa Nui on*Google Earth™

Terry L. HuntDepartment of Anthropology, University ofHawai'i

Carl P. LipoDepartment of Anthropology and IIRMES, California State Univer ity Long Beach

lillimul1l

IIll'ush'elle\!1

mull

clIfe

" ARCHAEOLOGY

I'e/milfUI'

.l'orh~ ieal lin l'~

. \('1I\ lIlion

'lIlellile (m;lge

Ped '!>trian ur\'l'~'

]~ill' Photol!.rllph}

~ urfare . lapping

.\erial Pholll~raflh)'

.\uger hole-

Figure 1. A top-down field re earch model for archaeology em­phasizing the conservation ethic.

The organization of the dizzying array of pos ible archaeo­logical observations into a research management strategy iaccomplished by taking a "top-down" approach (Figure I).In thi way, we begin our work with inexpen ive and high-

and overlay as many classes of information as possible,with minimal impacts to the archaeological record, employ­ing iterative sampling designs generating data until reachingredundancy in specified classes of information at levels ofpreci ion required to resolve problems of interest. For sci­entific archaeology, Dunnell (1982:7) has described themethodological dimen ion of empirical ufficiency (i.e.,Are units of meaning measurable in the empirical realm?)and their relation to tolerance limit (How close i clo eenough in our measurements?); see also Hunt et al. 200 I. Inconcert with dynamic sufficiency (i.e., completeness oftheory), our efforts should focu on decreasing the gapbetween explanatory accounts (i.e., models and hypothe es)and our particular case (Dunnell 1982; Hunt et al. 200 I).

"TOP-Do

A Co SERVATIO ETHIC

The archaeological record is, by it very nature, nonre­newable. Each aspect of the record comprises unique

information, the sum of which provides the only testamentfor va t amounts of human hi tory in pace and time. Therecognition of the nonrenewable character of the archaeo­logical record has led to the compelling rationale for a"conservation ethic" (Lipe 1974).

A conservation ethic demands that we minimize dam­age to the archaeological record whenever possible for twodifferent, but complementary reasons (Dunnell and Dancey1978; Dunnell 1984; Moratto and Kelly 1977). First, wemu t endeavor to pre erve cultural heritage for its ownake since it has intrinsic and diver e values to living popu­

lations. uch heritage ha cultural, ymbolic, political, andae thetic value , among other. Second, preservation ofarchaeological phenomena i a cientific imperative, inceuch phenomena comprise the only empirical record of our

pa t that is subject to scientific explanation. Thus as archae­ologists we must seek to preserve the record even in thecour e of our continuing research. As students of archae­ology learn from the outset, archaeological research is oftende tructive - we sometime destroy what we study and seekto preserve - like the metaphor of tearing the pages from abook a we read it. De pite this paradox, archaeologistsmust be fervent advocates for pre ervation.

Tn continuing research of the archaeological record theconservation ethic should guide our choices of data genera­tion techniques. While the' tandard" techniques of excava­tion and survey are just two ets of trategie that poten­tially provide us information, we need to document the ar­chaeological record in ways that best address our researchque tion . Combining research interests and a conservationethic, our field strategies compri e the cheapest (i.e., mostefficient) means for generating ob ervations of archaeologi­cal phenomena, which in tum employs means that are u u­ally the least damaging to the record. Choosing field re-earch technique should include as many kind of observa­

tion a pos ible since the record ha potentially many arti­ficial characteristics expre ed in vi ible light, electromag­netic pectra, as well a variations in chemical and phy icalform re ulting from human behavior and expressed at scalesranging from the ion to the land cape. We should acquire

* This paper has been peer-reviewed. Paper received J2 August 2007; Accepted J3 January 2008; Revised 28 Jan!lGlY 200 .

Rapa Nui Journal 5 Vol. 22 (1) May 2008

resolution satellite imagery to detelmine the distribution ofalterations from human actions across a landscape. We thencould use publicly-available infrared, sub-meter resolutionlow-level aerial photographs available for many years anddifferent seasons to map topography, soil and vegetationdifferences or to examine areas in their historic context.Based on hypotheses about the relationship of the composi­tion of the archaeological record with the landscape gener­ated from the satellite and aerial images, we can then zoomin on area with low-level aerial photographs available fromkite- or balloon-based platforms to document tructures andobjects at the scale of portable artifacts. For example, wecould ask how pattern visible in small-scale imagery mightbe areas that are compositionally distinct as a con equenceof cultural debris deposited on the urface.

Today, using guidance from satellite and aerial im­ages, we might expand our ensory picture of the target areathrough inten ive ground-based surveys and urface map­ping to generate data for the spatial distribution of sub­meter scale artifacts and the details of features. Thi workmight be followed with geophysical studies of image inpredetermined areas in order to map compo itional informa­tion about subsurface and near- urface deposit. Targetedareas could also be characterized using geophysical meth­ods such as resistivity and ground penetrating radar or anyof the other near-surface remote sensing techniques that arecurrently available. Each of these techniques provides amean of resolving structure (i.e., patterns of variability) inthe archaeological record. Subsequent work might layerinformation obtained by inexpensive small-hole excavations(i.e., coring) that can be used to acquire micro-artifact sam­ples as well as smaller sediment and ion-sized artifact infor­mation (e.g., "artificial" chemical concentrations such aspbo phorous). These classes of data at combined scalesform a robust understanding of the archaeological record inthe context of the landscape upon which it is distributed.

RAPA NUl

On Rapa Nui, some of the scales of inquiry have seen moreuse than others. Landscape approaches have tended todominate research practices (e.g., McCoy 1976; Cristino etal. 1981; Stevenson 1984; Vargas 1998; Vargas et al. 1998).Examining the record at the scale of the land cape has pro­vided insight into variability in settlement and sub istencepatterns (e.g., Vargas et al. 2006). Part of the emphasis onland capes is a consequence of the practical and conceptualdifficulty of working at smaller scales, particularly relevantgiven the problematic notion of "site" (Dunnell 1992). It isclear to most researchers that the archaeological landscapeof Rapa Nui exemplifie the impracticality of the "site" as aunit of observation, recording, and analy is. At least insome domains, the "site" has been regarded as a di cretearchaeological unit with ethnographic connotations of"community" (Willey and Philllips 1958), as has been com-

mon in Polynesian archaeology. On Rapa Nui, however,with a nearly continuous distribution of artifacts alongscales ranging from small objects to monumental architec­ture spread over the surface of the island, any identificationof "site' from any scale of observation remains decidedlyarbitrary.

Unfortunately, even though research has tended toemphasize landscape scale patterning and di tributions, theunits of observation have focused on intuitive ethnographicor "ernie" classes (Commendador 2005). These classes(e.g. umu, hare moa, hare paenga), while useful shorthandfor communicating general observations, tend to compre svariability in typological units, idealize form, and rendermeaningful analysis of these entities difficult, particularlyoutside of an ethnographic context. This problem can beavoided through an analytic approach to the functional andstylistic tructure of the archaeological record of Rapa Nui.Such an analytic approach examines patterning at multiplescale - artifacts, their aggregations, and multiple aggrega­tion - without the imposition of the "site," variously con­ceived. "Non-site" analytic approache have been success­fully demonstrated in other contexts (e.g., Dunnell 1983;Dunnell and Campbell 1977; Ebert 1992), and are certainlyapplicable to our examinations of the archaeological phe­nomena of Rapa Nui.

SATELLITE IMAGES

In top-down archaeological research de ign, satellite imageprovide an initial means for generating information on thearchaeological record over the largest areas. Over the pastseveral decade, atellite imagery has become a powerfuland efficient means for documenting the structure of theearth' surface over large areas. The potential of satelliteimages in archaeological research, however, is only begin­ning to be recognized. Part of the slow rate of adoption ofsatellite-based research in archaeology has been the rela­tively high cost, the technological knowledge required, andthe comparatively low-resolution of images available fromearly satellites. Thus, the earliest uses of satellite imageswere limited mostly to remote sensing specialists and thestudy of landscapes or large archaeological features (e.g.,Allan and Richards 1983; Custer 1986; Ebert 1980; Find­low and Confield 1980; Schalk and Lyons 1976).

The past several years, however, have seen an explo­sion of new sources of satellite images with increasing reso­lution integrated into user-friendly tools. These new sourcesinclude declassified military imagery and the establishmentof commercial firms that have launched their own satellites.A new generation of satellites such as Corona, SPIN-2,Orbview-3/4, SPOT, EROS, and Ikonos provide high reso­lution images l that are well below 10 m and available at lowcost. Since these images are capable of resolving features ofarchaeological interest (e.g., structures, deposits), their usehas spawned a wide variety of applications in archaeologi-

I In general, "high resolution imagery" is defined as a representation of the Earth's surface of less than 10m2 (Forte 2001: 132).

Rapa Nui Joumal 6 Vol. 22 (1) May 2008

cal research (e.g., Failmezger 2001; Fowler, 1996, 2002;Kennedy 1998, Kouchoul<os 2001; Lipo and Hunt 2005;Mumford and Parcak 2002' Philip et al. 2002).

Significantly, the resolution of recent atellite imagehas proven to be sufficient to re olvc linear prehistoric fea­tures uch as roads (e.g., Ur 2003; Sever and Wagner 1991)and even footpaths (McKee et al. 1994; Sheets and Sever1991; Sheet 2003). A we showed recently (Lipo and Hunt2005), it i possible to detect feature related to Rapa uitatue (maai) roads provided one can acquire image that

resolve features ca. 2 to 5 macro and if views of maairoad related-features are not ob tructed by vegetation orother kind of ground cover. When these conditions are met,it i possible to construct a map of the prehi toric road andu e satellite images as a primary means to study at lea tsome portion of the path across which the maai weremoved (Lipo and Hunt 2005). As the readers of this journalwill appreciate, the par e vegetation of Rapa Nui providesan ideal location for an array of remote sensing applica­tion .

In our initial work u ing atellite data, we acquiredimages from DigitalGlobe' QuickBird atellite. Thi com­mercial satellite wa launched in 2001, orbit the earthevery 93.5 minutes and re i it it path every I to 3.5 daydepending on latitude. Remarkably, the QuickBird satellitei capable of generating panchromatic image with resolu­tion of 61 to 72 cm and multi- pectral images with re olu­tion of 244 to 288 cm. Differences in re olutions dependon the degree to which the atellite i off from the nadirwhen the image is taken. Data in the DigitalGlobe linagesare geo-proces ed 0 that point and feature can be locatedwith an accuracy of23 mat 90% circular error.

The re olution of the QuickBird satellite provides im­age that are comparable with aerial photographs typicallytaken at high elevation. Studies have shown that QuickBirdimagery is of sufficient re olution to provide a ba e formapping between I in = 200 ft and I in = 400 ft scale(Nale 2002). One of the advantages of QuickBird is its abil­ity to acquire data at near nadir and that the correspondingdigital imagery may be con idered true orthophoto . Anadditional advantage that QuickBird has over aerial photo­graph i that images are available in color, panchroma, and4-band multi- pectra ranging from blue to near-infrared(400 nm - 900 nm). Thi multi- pectral information, thoughat a lower resolution (ca. 2.4 m), is of sufficient quality toprovide for a broad range of vegetation and environmentalinformation. atural color imagery can provide crop, fore t,and wetland information. Remarkably, all of the e datasetsare available for areas across mo t of the Earth for aboutUS 30 per square kilometer.

GOOGLE EARTH IMAGERY FOR RAPA VI

One recently available means of di playing data in a spa­tially integrative fashion is Google Earth2

. Google Earth

2 http://earth.google.com/3 Thi imagery is regularly added to and updated by Google.

Rapa Nui Journal 7

.Figure 2. Satellite image of Rapa Nui (Image 2007 NASA).

mapping ervice is a free application for browsing andviewing maps, imagery and data from around the world.Utilizing an intuitive graphical interface, images and dataare streamed over the Internet to the application a onezooms and "flies" acro the planet. The application is im­pre ive and sports imagery across the entire Earth. Oneparticularly useful feature of Google Earth is that for manyarea high-re olution color atellite imagery is available3

. InNovember 2006, Google added imagery for Rapa ui withone-meter re olution (Figure 2). Although a small area ofthe northern edge of the island is mis ing and ome mallportions of the island are covered by clouds, this imageryrepresents the most complete and highe t-resolution cover­age for the island that i currently publicly (and freely)available. The images provide a remarkable source of infor­mation on the environment, landscape, land-use, and thearchaeological record, clearly displaying architectural fea­tures, ahu, paths of maai roads and the position of maai,manavai (agricultural enclosures), and even concentrationof agricultural rock mulch (e.g., see Bork et al. 2004;Wozniak 1999), among many other features. Figures 3 to 9provide examples of features vi ible on Rapa Nui in imagesfrom Google Earth. We also note that high-resolution im­ages are available for many Pacific Islands (see Figure 10, aview of the Maeva area of Huahine I land, Society Islands,including stone-constructed fish trap ).

In our continuing field research addressing multiplea pect of Rapa ui prehistory we have integrated commer­cially available satellite images, free images from GoogleEarth, kite- and blimp-based low elevation aerial photogra­phy with pedestrian surveys mapping, geophy ical tech­nique , and small- cale excavations (e.g. Ayala et al. 2005;Lipo and Hunt 2005; Lipo et al. 2005). Thi top-down ap­proach is inspired by the con ervation ethic that we outlinein thi paper. Figure 11 and 12 illu trate the correspon­dence of Google Earth imagery and the higher resolutioncounterpart of ki telbIimp photography of the arne areas of

Vol. 22 (1) May 2008

Rapa Nui Journal 8

Figure 5. Satellite image ofAkahanga, southern coast;the coastal ahu, modemroads, stone mulch areas, andother features are visible (fromGoogle Earth MappingService,http://earth.google.com,Image 2007 DigitalGlobe,lmage 2007 A A).

Vol. 22 (1) May 2008

Rapa Nui Journal 9

Figure 6. Satellite image ofan interior area of Akahangawhere multiple manGvai(agricultural enclo ure) c1u tersare visible (from GoogleEarth Mapping ervice,http://earth.google.com,Image 2007 DigitalGlobe,Image 2007 NASA).

Figure 7. Satellite image ofan interior area of Akahangawhere maai road feature,historic pirca (walls), andmodem agricultural fields arevi ible (from Google EarthMapping Service,http://earth.google.com,Image 2007 DigitalGlobe,Image 2007 A A).

Vol. 22 (I) May 2008

Rapa Nui Journal 10

Figure 9. Satellite image ofan interior area near MaungaTe Kahu Rere, outh coa t,where land-use change indeveloping parcela foragriculture are vi ible(from Google Earth Mapping

ervice,http://earth.google.comImage 2007 DigitalGlobe,Image 2007 ASA).

Figure 10. SateUite image ofMaeva area ofHuahine I land,Society Islands FrenchPolynesia; the //larGe com­plex and traditional fi h trapare visible (from GoogleEarth Mapping Service,http://earth.google.com,Image 2007 DigitalGlobe,Image 2007 A A).

Vol. 22 (1) May 2008

Figure 11. Satellite image ofan interior area of Oroi,outh coa t; the moai road,

hare paenga, manavai, andother features are vi ible, butnote the greater detai I in thekite photograph show in Figure12 (from GoogJe Earth Map­ping Service,http://earth.google.com,Image 2007 DigitalGlobe,Image 2007 NASA).

12. A kite-ba ed aerial pho­tograph of archaeologicalfeatures in the interior area ofOroi, outh coa t; the higherre olution clearly shows themoai road, hare paenga,manavai, and other features,but in greater detail than in thesatellite image (Figure 11).Low elevation photographyfrom a kite- or blimp-platformprovides details useful in docu­menting the archaeologicalrecord.

•

Rapa Nui Journal 11 Vol. 22 (1) May 2008

the south coast (near Ahu Oroi) of Rapa Nui. For complex,fragile, and sensitive architectural features such as ahu, theintegration of remote imagery from satellites and kite pro­vides unmatched data in producing detailed maps. Suchmap could be produced with little, if any direct contactwith the tructure (e.g., walking over it). As pre ervationconcerns increa e with continuing impact, particularly inthe remarkable archaeological record of Rapa ui, the enon-invasive mapping techniques will likely prove es en­tial.

We sugge t that field data can be mapped and pro­vided in its spatial context using Google Earth and di play­ing data, such as for moai, ahu moai road, and other ar­chaeological features for haring u ing Internet venues, aswe have developed with the Mu eo Antropol6gico P. Se­ba tian Englert (e.g., Torre Hochstetter et al. in pres ).Public acces and haring of data - for the Rapanui andresearch communitie in particular - have become particu­larly urgent with rapid land-use changes on the island.

The recently and freely available high-re olution satel­lite image from Google Earth hould prove u eful to re­earchers working on Rapa Nui. Integration of the eources no longer ju t economical, but now free, facilitates

a top-down research trategy and a con ervation ethic. Itshould also prove to be a vital tool in integrating and har­ing archaeological data for purpo es of research and pre er­vation ( ee Kintigh 2006).

REFERE CES

Allan, J. A., J. S. Latham and R. W. Smith. 1980. Monitoring landand water use in the Gefara region by satellite imagery. An­nual Reportfor Libyan Studies 12:79-80.

Allan, J. A. and T. S. Richard. 1983. U e of ateLlite imagery inarchaeological surveys. Annual Report for Libyan Studies14:4-8.

Ayala, I., C. Lipo, and T. Hunt. 2005. The use of satellite imageryto study prehistoric agricultural features (manGvai) on RapaNui. VIIntemational Conference on Rapa Nui and the Paci­fic, C. M. Stevenson, 1. M. Ramirez, F. J. Morin, and N. Bar­bacci, eds. :113-123. Los 0 0 : Ea ter I land Foundation.

Bork, H-R., A. Mieth & B. T chochner. 2004. othing but tones?A review of the extent and technical efforts of prehistorictone mulching on Rapa ui. Rapa Nui Journal 18: 10-14.

Commendador, A. 2005. Measuring variability in prehistoricstone constnlction on Rapa Nui, Chile. Unpublished M.A.the i . Department of Anthropology, University of Hawai'i,Manoa.

Cri tino, c., P. Varga Ca anova and R. Izaurieta. 1981. AtlasArqueol6gico de Isla de Pa cua. Centro de Estudios Isla dePascua, Universidad de Chile. Santiago, Chile: Corporaci6nToe ca.

Custer, J. F. 1986. Application of LA DSAT data and synopticremote sensing to predictive model for prehi toric archaeo­logical ites: an example from the Delaware coa tal plain.American Antiquity 51 (3):572-588.

Dunnell, R. C. 1982. Science, ocial science, and common en e:The agonizing dilemma of modem archaeology. Joumal ofAnthropological Research 38: 1-25.

Dunnell, R. C. 1983. Aspect of the patial structure of the Mayo

Rapa Nui Journal 12

ite (l5-JO-14), John on County, Kentucky. AnthropologicalPapers-Museum ofAnthropology, University ofMichigan(72): 109-165.

Dunnell, R. C. 1984. The Ethic of Archaeological SignificanceDecisions. Ethics and Values in Archaeology. E. Green, ed.:61-74. ewYork:FreePre .

Dunnell, R. C. 1992. The otion ite. Space. Time and Archaeo­logical Landscapes, J. Ro signol and L. Wandsnider, ed .:21-41. ew York: Plenum Pre s.

Dunnell, R. C. & S. K. Campbell. 1977. Aboriginal Occupation ofHamilton I land, Washington. eattle: Univer ity ofWa h­ington, Dept. of Anthropology, Reports in Archaeology, 4.

Dunnell, R. and W. Dancey. 1978. Assessment of Significance;cultural re ource management plan. ASCA Bulletin 5:2-7.

Ebert, J. I. 1977. The role of remote sensing. Settlement and Sub­sistence along the Lower Chaco River, C.A. Reher, ed.: 191­215. Albuquerque: Univer ity of New Mexico Pre .

Ebert, J. I. 1992. Distributional Archaeology. Albuquerque: Uni­versity of New Mexico Press.

Ebert, 1. I. and T. R. Lyon . 1980. Prehistoric irrigation canalsidentified from Skylab III and Landsat imagery in Phoenix,Arizona. Cultural Re ources Remote Sen ing. T. R. Lyonsand Frances J. Mathien eds.: 209-228. Washington, D.C.:Cultural Resources Management Division, ational ParkService.

Failmezger, V. 200 I. High re olution aerial color IR, multi-pectral, hyper- pectral, and AR imagery over the Oatland

Platation archaeological site near Leesburg - a pilot study ofthe new high resolution satellite imagery. Remote Sen ing inArchaeology. S. Campana and M. Forte, eds. :143-148. Flor­ence: Al1'lnsegna Del Giglio.

Findlow, F. 1. and L. Confeld. 1980. LANDSAT imagery and theanaly is of archaeological catchment territorie : a te t of themethod. Anthropology 10 (I ):31-52.

Fowler M. J. F. 1996. High-re olution atellite imagery in ar­chaeological application: a Ru ian satellite photograph ofthe Stonehenge region. Antiquity 70 (269):667-671.

Fowler, M. J. F. 2002 Satellite remote sensing and archaeology: acomparative study of satellite imagery of the environs ofFigsbury Ring, Wilt hire. Archaeological Pro pection 9(2):55-69.

Hunt, T. L., C. P. Lipo and . terling. 2001. Posing que tion fora scientific archaeology. Posing Questionsfor a ScientificArchaeology. T. L. Hunt, C. P. Lipo and S. Sterling, ed . :1­21. London: Bergin and Garvey.

Kennedy, D. 1998. Declassified satellite photographs and archae­ology in the Middle Ea t: Ca e tudie from Turkey. Antiq­uity 72(277):553-561.

Kintigh, K. 2006. The promise and challenge of archaeologicaldata integration. American Alltiquity 71 :567-578.

Kouchouko, . 200 I. atellite images and ear Ea tern land­capes. Near Eastern Archaeology 64 (1):80-91.

Lipe, W. D. 1974. A Con ervation Model for American Archae­ology. The Kiva 39:213-248.

Lipo, C. P. and T. L. Hunt. 2005. Mapping prelli toric tatue roadon Easter Island. Antiquity 79: 15 -168.

Lipo, C. P., T. L. Hunt, and S. Rapu. 2005. Recent geophysicaland archaeological studie at Anakena. VllnternationalConference on Rapa Nui and the Pacific. C. M. Stevenson, 1.M. Ramirez, F. J. Morin, and . Barbacci, eds. : 97-105. Loo os: Ea ter I land Foundation.

McCoy, P. 1976. Easter Island Settlement Patterns in the LatePrehistoric and Proto-hi toric Period Bulletin 5. New York:Easter Island Committee, International Fund for Monuments.

Vol. 22 (1) May 2008

McKee B. R., Sever T. L., and Sheets P. 1994. Prehistoric foot­paths in Costa Rica: Remote sensing and field verification.Archaeology. volcanism and remote sen ing in the Arenalregion, Costa Rica. P. Sheets and B. McKee, ed . :142-157.Austin: University ofTexa .

Moratto, M. and M. S. Kelley. 1977. Significance in Archaeology.The Kiva 42: 193-202.

Mumford, G. and S. Parcak. 2002. Satellite image analysis andarchaeological fieldwork in EI-Markha plain (south Sinai).Antiquity 76 (294):353-4.

Nale, D. K. 2002. QuickBird aerial produci comparison. Preparedby eMap International for Digital Globe, Inc.<http://turfimage.net/geospatial/Sci_documentslAeriaLvs_DG.pdf.>

Philip, G., D. Donoghue, A. Beck and N. Galiatsatos. 2002. CO­RONA satellite photography: an archaeological applicationfrom the Middle East. Antiquity 76(291): I08-118.

Schalk, R. F. and T. R. Lyons. 1976. Ecological applications ofLANDSAT imagery in archeology: an example from the SanJuan Basin, New Mexico. Reporls ofthe Chaco Center(I): 173-186.

Sever T. L. and Wagner D. W. 1991. Analysis of prehistoric road­way in Chaco Canyon using remotely sensed digital data.Ancient road networks and selliemeni hierarchies in the NewWorld. C. Trombold ed. :42-52. Cambridge: CambridgeUniversity Pre s.

Sheets, P. D. 2003. Prehistoric human footpath in Co ta Ricaindicate intimate ties with villages, cemeteries. ScienceDaily,October 8, 2003.<http:www.sciencedaily.comreleases/2003/101031008064657.htm>.

Sheet, P. D. and T. L. Sever. 1991. Prehistoric footpaths in CostaRica: Transportation and communication in a tropical rainfor­est. Ancienl Road Networks and Seltlemenl Hierarchies in theNew World, C. D. Trimbold, ed. :53-65. Cambridge: Cam­bridge Univer ity Press.

Stevenson, C. M. 1984. Corporate Descenl Group Struclure inEaster Island PrehistOlY. Ph.D. Di ertation, The Pennsyl­vania State University.

Torres Hochstetter, F., S. Rapu Haoa, C. Lipo, and T. Hunt. Inpress. A public database of archaeological resources onEaster Island (Rapa Nui) using Google Earth. Latin AmericanAntiquity.

Ur, J. 2003. CORONA satellite photography and ancient roadnetworks: a northern Mesopotamian case study. Antiquity 77(295):96-115.

Vargas C., P. 1998. Rapa Nui Settlement Patterns: Types, func­tion, and spatial di tribution of households structural compo­nents. Easter Island and Easl Polynesian PrehistOly, P. Var­gas Casanova, ed. : J 11-130. Santiago: Instituto de EstudiosIsla de Pascua, Facultad de Arquitectura y Urbanismo, Uni­versidad de Chile.

Vargas C., P., C. Cristino and R. Izaurieta. 2006. 1000 anos enRapa Nui: Arqueologia del asenlamiento. Santiago: Institutode Estudios Isla de Pascua, Universidad de Chile.

Vargas c., P., R. Izaurieta, C. Cristino and C. Arias. 1998. RapaNui ettlement and ub i tence pattern studies on Easter Is­land. A GIS based analysis of the Archaeological SurveyData. EaSier Island and East Polynesian Prehi tory. Pro­ceedings ofthe Second International Congress on EasterIsland and East Polynesia, P. Vargas, ed. :147-152. Santiago:Universidad de Chile, Instituto de Estudios I la de Pascua.

Willey, G. R. and P. Phillips. 1958. Method and Theory in Ameri­call Archaeology. Chicago: Univer ity of Chicago Pre .

Wozniak J. A. 1999. Prehistoric horticultural practices on EasterIsland - lithic mulched gardens and field systems. Rapa NuiJournal 13:95-99.

PREkisTORic RApA Nui. LANdSCAPEANd SETTLEMENT ARCkAEOloGY

AT HANGA Ho'ONU

By CJmis OpJtER S EVE SON & So iA HAOwith contributions b}' Joan Wozniak

Helene Marrinsson- aIlin, & Paw Wallin

As the authors of this b k show conrrJry to past perceptions,thc Eastcr I land landscape \ a highly transformed andmanaged agriculrural rcrrnin that emergcd in response to

deforestation by the PoJynesian \ ho ettlcd there. Thi olumeadds a nC\ dimension to scholarly in esrigation. about \ hy the

island' prehistoric odery evol cd the \ r it did.

ofr cover, 297 pagel & fold-our map' an Ell- publicarion, 2008 [:# H08)I BN 978·1-880636-26~8 • PRICE: $30.00 + shipping

Rapa Nui Journal 13 Vol. 22 (1) May 2008