From 2d to 3d data: Al-Wu’Ayra’s crusader castle

&3 .{§

1... \t/r t^nr11 \r rvr r r_.,,

§ly

^

n=-t

\J-I

PROCEEDINGS 5th

lnternationa I Congress

on "Science and Technology for the Safeguard of Culturql Heritage in the Mediterraneon Bosin"

VOL. I RESOURCES OF THE TERRITORY

lstanbul, Turkey

:: - 25 November

201i

=.

FRO\I lD TO 3D DATA: AL-WL'A\

R{

S

CRLSADER CASTLE

Gebr-ielli Robertor. Angclini Andrea:. Giìrnta Elisai. Taffi Elenal

:

::::

:rl:r:.1.,:l.s.\pplcd to Cnln,.al H.lra3. iITABC). C\R Ronr..

hat)

-{I]STRACT

-qei ..o

p

". r. _:::::.::..r:b*rtlreeiper;ùen:31ì(,rofr.\sutre\insùerLodoloejesinùrderr.docureùr ::::::::::r..:r.lr..ks!alialirne.aclxnlrn|eNpe.ti!.amùrglLecasrlesol_.4l,qarn..!li/tr!rÌa

::::

: :.:: :::::r:i:,i a.ùal plrorùgraph) iio r braker air baLldnì and [rtttererrjal ajlobal fositìontr C i.ì:::.:r.::::,.rrÌi..oùe.LionphorogransandrheirqorcturerceCloseErscptrorÒgraphltlas . .

: :

. : ..:r.::. i rd .lrifi3rri. drl|r.ice berween rhe lrronqho d ard rhè sur.undine fo.rilìed a,ua :l . r::.rir:r oi i .xflesLaf iìid plad. Òt nLt rhe arca. As resull of rhjs. rhe sotrq,are hare :i:.:i ,:. ...1.1 n)sai. TLereb! Lhe r{oni qroùp has decidcd io erDèinrcnt a : . :::,.:: :r:.i :ilnr:!l t. r.connru.r a lD rrùeric DÒdel of ltrc cante. ::r.a.r1l.dlD ro.el direcred rolhe docùnrenLrrnrn .f drè presènr da),siruaiìon. On ihc :::.r::!li:\.b.cìrÌ3d.nreoi..sfarialaDdvolùneLicjlùllrerinmLharnrodelsehaveÌrc.n : : :: : :::r:::. Ìnrtli .ftL. sl.orgh.lf ard th. sf rlndils lofijtì.d aea. :: : . ::i::.:i.iJl.si.s afd rerprnsile 5otulare have .lioqed us lo rcvisè pr.vLously obrrirìed dat, ) ir rO aaia llr.s. r!neric daia tr,ill frciliÌate lhè ar.haeolosicat conrDrehetui.n and

-..-::

.

\ -ir

)DL

CIIO\

:.:::: ,ìl

: :

"

.lr ijrr.lrr-.r is placÈd

,i

'' b

Ò p...o

rhe ennance

.,...rt.,.rr,

o

ofthe monulllenral atca olperra.

: :- :::l i.riher| lordan. ll ùas olle of several caslles in rhe easrern hìshlrnds l:lì: ueJ. and connr[iDicaion routcs belween Cairo and Dan1ascus. Ar on. :: : .::r!ìl.l:lereirle castle standing tigh on top ollhe steep cliltì on rhrce sides and _ : . ,rr Ìe 1ò!fih. r:: -::.ìr1r.l.i r.fèr to it as "Li l/dtù tttoise

:::::lh

and the Arabs

as.wu'eia . The

irralcurate as ir rcIcrs ro rhe whole ol [/.Jrli Mrl.rr. or Val]ev ofMoses, -: :i :r,: hl.ror ìc;rÌ naLre give lotlecasr)e - _: : i.i rh: kiÌrg of JerusalerÌ Batdovino I al lhe begi|dDs of rlìe l2,r, ceDlluy,

..:i : .: : :: : : :: : . :::..:.:::.-,ì:pl.,noflhePeiravalle!.ll-lI4r7_rra'scasde$,aspaftolauelaborare . : -: .::: : :i.!t:llali.ns $hicll included lol1g \\,aìls, obserYation towers and cislems dùg : .::-. _ .: :.,.k The plìysicaì characleristics of these sandslones are crtremely .. . .: : .: : :r. rranulornehl, of the roclc, and ro the iype and quanljn ofthe cements. :: ::. ::::::r:L-u-i crusadcr casttes ofJ:l?dwhak. Al tl/u,A!tu at\d _ti-H.7tir in the peira ::: .::- : .r.ure ùe region southeast of ihe Dead Sea for tlì. Franks rn Jerusalen. - -:.: :::,::::: ri lr ll;r.,1r., lerne 10 an end when ir lcll ro rhe soldiers ol Salntì ot Din : : i : : . i,:ii_.f: lhe lall of .l/rd,òdi- Fron arclìaeologìcal erca\arr..ns rÌ:d; ir 1000 _ -: : ::::: .; Florcnce has beerì possible lo defiDe nine Fhase! iianinq tìom rhc ::r ,,i:::t.i i,r lxe:ti.. Three phases are ascdbcd tar ile cnLseder eEr rasp.cialh the : ll ..rrlinj. t|e foù r is referr.ed to lhe a1}rbid ar:rdar:e \hjl3 ihe lasl :_ , .. -::::.:::,ì Lir. \iihdras ofùe castle eterr jfarc demonnrat.d su.r.:sire uìe oiihe .: : : r.:: -ill antr,liluntn. : :i:.:: ::::: :ìrrilisaijlns allo§ed io develop considerabie.lem3ri. ri.1r.r\ _3jj: aI: . r-: :.: :. :: ::r::ir .irù.inrer ol setilemeni. lhal comprehensinn i: ::::::.:: ::. -l r:r -

,l

!ù:,.lrgraphic seqlìences such as mmparts pacing. dra§bridges, \varer collection I,slem and :.rbir rrd lopograirhìc dìstribution ofarce and b!ildings.

.llllri'fll.Ttcastlerneasuresl00nofLenglh,6,l-S0moftidrhandrhcruiDsolthetìrÌressca| b. achjeved b) crossing a bridge o\'er rhe Llali ltu'avra and tlro.:,glì a liulc luln1el car\,ed inio lh. rock. Ìhc canle is the largest arrong tlre lorlrcsses near lelra. djstingrìished bl e functionaì dilference bclùcer1 the stronghold and dre suloùnding foftiiied area. J he forrificd sysrem rs conlosed Lr-v a dclcnsi!e line thal exploits the geomorphology ol the area (escarpncùl l:)1' ùe sùr'ourdi|g uadi) aDd \\,here rose 1,1 quadrangular toueÉ. The stuonghold is organizcd il1 Llree areas of dlffereff hcighr, ùe lìrst on rhe level of the cnlrance portal is 10-15 m belo§ thc lcvcl of the intennediate iù1cr ar.a \\,hich in turD is located aLrout 5

llris \rork relales to thc Io]losing

Iù

beloÙ the upper leveì.

archaeological project "l.Ldrer:al Pettd: otchuedÒg.t).1 Ct sader-Aj,j ubid \etl.r1,:r/ in Trdn\iordan" linanced by the \,ljnist{, ofEducariol1, Unr ers q ard Rcscaroh l\4lUR) :rnd by tle \,Iitlistlf of lioreign AlTairs (lvlAE) under the direcdon of Prof. Griido Varnini olrhe Universi§ ofFlorence and in coliaboratìon r'ìth rhe Institule lor lhe leclnologies Applicd ro ùe Culflrral Heritage (T'I^BC) oflhc N.ìrional Councìl ofllesearchcs (CNR). ITABC deaLs §ilh c\periDentatioD and apfììcatjor of imcgrated nethodologies and

lnultidìscipìlnal t:chnìqnes fot thc l(r1o llerilage

1edge. manaqerrent,

fruiron end safcgùerd ofCulfl[al

The e]iperinenlarion js 1ìralized to:

Re\is€ pre\ioùslv obrained dara scl and convert ftorn ll) lo lD dal! rhrough neÉ, menrÒ.1.,1og) rnd rne\pensile soft\\,arel Rcrse ol daia acqun ed iù diflcrent sur\e) s iD order to expcrinerl modern techniques of elaborarior):

chieve a tridnnersìonal n$mÈric model of all r|e castlc rl1 order to obtaiù an high defifrtion orlhophoto. Thc nlain Ì.urpose oftlis ù,or]i has bccr lhe eaperìnentariùn ofne\ sulvel ing methodologres rr order r.r documcnl llrd conre\tualìze the 1l :t.t .4trn siie and ro c|cck in prosped spaiial iirlcricliom anong the o cr canles o1.l/-,ryrlj.§ and.!}arrdk The rcsearch groùp has adopre.t dilièren1 sun evifg mclhùlologies in rhree diffcìcnt oampaigns:

. . .

l00l

1002 1007

ùsc 01 DGIrS: usc oi nedaÌ thoiograph) ùorn hire: use

ofaeriai phdograph) liom bralied ai brlloon.

TOPO(]RAPIJIC SURVEY 20(T ln 2001 occrù-red the irst rrethodoloeical slÌr!e), ofrhe ITABC

in rle research progran oftfe Uni!ersil\' of Florencc related to the crusader casrles ol the Trans-lordan. Thìs arohaeoìogical rssioù al Pelra invoh,cd lhe ITABC in a topographical surlc,r o1'rhe 1l fitl! A,-ro casrte. Stafrng ftom r cafiography pre\'iously car ed ou1, the airr ofrhal sff\'e\ \\as rhe seneration of a llTN4 (l)igrlal Terrain Modcl) ofùe area ìn order ro acquire jmpoflanr jrfonnadon aboul altjlnetr]'. rnclinadons and exposùre of the castle ìo relatìon 10 nro othc6 ir11p!fiant casllcs: ,.j/ H./r1!. insidc ùe Petra Vallcv near the ancierrr roman settierrent. ard .\'i./rì,rdl I8 kln north east au,ay froù lllTz'.trì-r. llased on dlese informalior- ir \vas ìmportaùl lo eshblìsh ifrhere \\,ere Iisual colmeclions anong rhese ùree caslìes dumrg a de1ì|ire historical mo eDt (ctusader à]t,ubid perìod). The rescarch group decided lo use a DGPS (Differenrial Global posirromn-g Svsrem) in these piaccs, kirìenatic way and §,ilh the sarrpling rale of 1 pomr e\'ery' 2 secords h the )i ll;1r'Aro cestle the DGPS jlìvestigation ùa-! focused on the stronghold, considered by dre archaeoloqisl Lhe lrair atea ol ihc caslle. The measlLtemcrts !\,ere can.ied out also Ior rhe archeological ruins presert on thc stonghold for n\'o leasorls: georelerencing an existrng archseological map of the casde and al the same tilne rectijl,iig i1 i.om occurred ùaDslete errors. Irom lhs DTN4 l\,as possiblc lo elidence ho$, tÌre qualiB oflhe surve), save us ì portanl r.\ulls in arcleologv and l(nowledse ofthe conte-lr oftlis castlc (Fic.1).

ri\l ::à|urJ rlrdr DCIS u r,i. Bi.:: ..1ù È\idÈr.es rhe ::i: :r r:r: ihr tu:'tr.

FÌg I

ar h.

r..ilii.d

ae.

a |hotier.Ih oi riId

Cartle ! nrùrSÌrold ii.nl tlre klie D!e to lhe iorriii.d òea r15 nlL acqriEd

r]]j rhe

::: .. ::i i.r: :laborared \\ìth sophìsticàred irlrcryolalion aìgorithms drat tfanslonrcd ùem .\ ::-rd.l llriargulated hegulal Nettork) that refreserrs srìrlacc as a se1 of contiguor6 :: .:!ing iriangles ir order to crear. qualjiy ir.ps. Throlgh rhese maps. ì1 \\,as possible . ij :... 3:rùm:li.j oD th€ Lllodel and verìf,v them or the gr ourd.

LI'(lGR{PIIICT -st RvE} 2002 ::::::.:.n.ar.rreologicalsùr\e)lheiear]luseda[rodifiedldlc\\ithananalogical]ight ::::::i. :rrrì:J ior a\oldirg dargers duc to the flight. This experìmÈnlal device alìowed ro ' : :: ariLnriion lìarnes tiom abour 2j nr grcùnd cLljleiion lresolurion oi I 2 cllr fionr 20 ì : :r: --i {rd.l.\a1ion) despite the shots were condjtiored by rhe sready §ind -:: ::::::i:_--t.he srLIle! was focLÌsed only on the stronglold area both archaeolosical r::,::: .:-:r..,:l.r:ical e(jdcrccs corceDrrtron) and §,eathù rcasons that cordjtioncd the I

:: :; .:::ir :i: ir: \ ind rhat rras a cr,DnaDl anrospheric fàctol oI dre zone, above all on the ::.:::: -::::.\::o:lrend.?i. lhe expcimentrlìon \\,as conducred iD t§,o steps: ìnrhe1ìr.llnc : ::::i:_: 'r: ur'rnan seoreferetcing. in the second o|e photographic processine aild :::,:::. :. :.:::r.ìng Tbe oblained dele were elaborated x,ith a saiellite georeferÈroairon :: ---= :111S ER apper 6.0) Irom u,hich rvas got an hlgh resolutìon orlhophoio rnLìJrr.

:::

-:::-::::i: \i.r.o\ er due ro \eather oonditìons the orthophoro : : ::ì .:._-:Ì ..\3r.d b\ frànles (Fig. 2). I']PO GR{PHI C SLRVEY 2OO7 : :: : i:-: ::. ::ol]gical sunet rhe researcl

group i.leqraled

mosaic sho§.ed solle iìr.as

ùe usr oi

aeri..1

.l

:r:. ::.1ìrl-\

:: :::r: :::::._ B-.tl.r 3ll,r!1.d ro rake ligh a..rn'acl phoros il cm 1ìrÌ.- n ::i :: -::. : : r . .! .re .rìrnpacr digiial carn.rarll\lPì. \ sri!o conir!,]rJr_.::::::r::: .:r .,,, ''. o- o. l,

:utp,rn had gol a rnechanical devjce holdjrg carrrcra levcìed. This approach had ihree kind of ad\.anraces rhrn kirer làcilitv ol Ieading it lvith three operarors, slabiliry end thc possibilis to reach a morc hcighr.

Furlhc rorc, ir $'as pcrlomed à dual ecquisitìon oI coordinates \À,ith the DGPS: on the stronghold aÌ1d on ùc sùroundirg lìrlified area lìr an amount ol i,l2 GCP Garqets and remarkablc ìroi s) . h the firsl case targeis \\ere posiiìoned on the stronghold's ground and u,crc located abour l0 meiers aìrarl so thal a160 clers liom the ground they could be seen on ùe ùaùe. ln thc sccond casc \\,c acLluircd coordiDares o f geomoryhologicà1ly re arkable points recognizable on rhc frames. Thc GCP sc 'c to gcncrale a coù1rol poirlls nel lirking the fiarne previously acqujred to the \lor'ld coordirrates sysrem. Dela set acquìrcd il] these campaigns ba\'e beeD thc srbjecl experirnentation ìn order to reuse 2D dara ior the lD lnodel.

of lhc plcsenl

methodological

DATA PROCESSIN(; In 2011 the research group lave decided 1.] re-organìze all ihe daia set 1o reach dìe aiùìs abolc described. The inieraotion berween lopographiral drla and dìgital close range fiames hàve let us 1o cnablish mcllodological crìlcrià 1ìr !.ilirg arld dc\ elopng a nlIrleric lnodel ofallrhe castle: . lDpul ol sprlLel dara collc.r3d ilLr rg d1e three topographic survev 1r a \\'orld coordirrates sysle (D(ìPS ì .1 \\ùh arr e.cùrrrc) of= 2 clnl: . Colle.troD and aràr!.rìrrf r ol lìerr.i iionl ihe lait illl.\ e\ ; . Tcsting oI lù\i .ùit s,rlltnarc rrÌch !i PhLìro\.ieùic. Tl.I Sharc, and PhotoN4odclcl in order io comfare darà serl l'horoMetric rs a fhotu r..ir1ìir1in:r r.iì\àrL'lhrÌ h!! àllù\ed us ro achieve an oÌ1holrroto o1 Lhe \h.rle ar.a L.ùrrn-hLìld rnd rLr:1'r.un.ling fofified area: fig.3) The fiame seleciion has be.f .!ì:Jjilor].i L.) rh. fr.s.rce oial lea5l i]\e GCI located unjfoùriy o\'er lhe ùlir. frèrÌ1es. \\e Ì1!\. rri .n.oir.rered problems obrairli g er or'lhopholo plane of the srronghold becarìs: rhe rìr:d flarl: ùas horrfscnlous. Due to rh: ahitude dilfereùcc relaled t! rh! ùc!r .jr,rlrrirh] :lìor3 all ne\r ro rairpani Fà.ingrlh. loiin,are is able ro deierni e t|e In.an squared ernr L\ijl ìnn lh! rliìr!n!Lr planc. L11os:ric

fis I

Ofiho Lectìned aeLi.l lnoroemfl

ior

b.lloon

.htìooied ùnL rloromerl.. Ox the nosaìr can

)ni.rd irar lri.s nfsl ichii!

are

l.x,ofa..ù:r.r

be

fig ,1

OfihÒjertìled aÈrìalthologGph liorr balloon elab.nred *iù lN-SIarc with lbe ;ane probieùs olaccùracy.

TemNova Sharc (IN Sharc) is a geosraphl.al inro.malion slslem and a sparral dclabase manàgemeni solflvare. ALso in ùis case lhe iiames ha1: been selected according 1lj a srarrdard ovcrlap \!here §,ere colImatcd fi\e GCP oblaining an onhophoro-plare looked lìke io pre\ iors rcsulr (Fig. 4). PholoN{odeler is a photograiÌìclric sollware rhat proyides inage-based nlodeling, Ior accurale measùremenl and 3D models, lhis is rhe mosi cosr ellecijve and affordable \\'ay ro do accuraie

3D

sca

liùg, measulement and

sune,vi,g.

The grorip has decided to experirìcnt

a

photogrannneiric software that alìo§,ed to reco slruct a lD numeric model ofthe castìe directed !o thc docunenlelion olthe presenr dav sjtuation. Picurcs has been chosen so that: each frane had ro overlap by 60% upoD ùe other; ihe arìglc bet$,een the carnera and the refèrence pìane had not to be less than 15' dcgrees. On the nodel have been nlade the ò11owing analyses: metric, spatial and volùmetric; turther from that modeì 1le have obtajned accuate aeial mosaio both of the stÌonghold and the surrounding lortified

The stud) and rhe anaìysis

ol that progranìs

have allowed ùs

to do

comparisons and

considemtions about d1e elaboration ànd the methodologl, applied Durirg the elaboratjon stcp lhe geomor?hology ollhe cestle has detemined the mid-pìane protrlenl. lhe sbonghoìd area can be corsidered as an homogeneolìs plane while the rampafts cannot be considered at ihe sarne way due to lhe slrong altitude differences. The soltware choice has been defined ìn order to 1ìnd the proper soIÌlvare lhal could solve this problem. From the anal_vsjs of the outputs, rhc group has evidenced horv photogrannletic soflware soìves better this matier. Ir TNsharc and I'hotometric soltw e lhe use of CCP is 1ìnalized to define the georelerentiation of single taùrcs, io stiich the liames in a orlhopholo mosaic aDd to evaluate the mjd-plane errols. Insread Ihotomodeler elaborales GCP only 1or the globaì georelerentialion \ùhile the evaluatjon of mìdplare is obtained arìtomarjcally by thc software liom the 3D model jn a sccorld program stcp. Despite the soft\i,arc does noi require so many GCP (142) Ior tlle georelerenriarjor howeler all control points have been compuled tu order to improve a d verif) the lnetic accurac\ oflhe lD

model(Fis.5)

:: :

:Lr rì:.Li.,,rnh Ìb.CaP lo.aÌ.d

ùr,f.rùl\ onfie qn,lrd

'Ihe l\,ork processing in pholomodeler has been sùtrdivided inio two phases: insenior ofmanual and homologoùs poinls on dre various lìames and generation ofsurfaces. The insertìon ofpoiÌrts has been can-ìed out coDsidering a honlogeneous distribrìrjon ol lhuln oD the fi-arnes aùd conterrporarily poinriDg out ùe geo]nolphology ofthe ground in order to a\,oid qeo]neùic er.ors in the restirution of rhe casilc's shapc. Thc colÌeciìon of vedices, edges and faces rhat defines the shape ofth€ olìect has bccn organizcd widl lhe same method ofthe poiÌrts too The neshes have been maDuaily refined iù tiaDgular sùfaces \\'i1h a bisger concentratio| close ro ihe ar.eas \\,ith remarkable ahitud€ dilfercnccs ordcr 1o describe corectìy the geo otphoÌogy of rhe

it

ca§le (Fig.6).

ris s

6

Tle picrur. Épr3srnrs nle ìoorlrins .lgorirh:r L!. ùe rÌ!]]r)

lD

nlodcì

ofd,. .,§le nnh ùiailula, D.r,.s (.i,

dì? Lehl ind

*ith ih.

.ON(]I,I,SIO\S a high definìlioù o lophoto-mosajc \1jt| aresolutionofI pircl eve[,2cm. Froff rhe rrrase and ùe rnodel has been possiblc carl ins orìi ffeù.jc and volurìelric chcclG end aDait zlllg spatial djiribulion of archaeo logjcal evjdcnces (fig. l). Ahcr 1e\r lelrs ne\ meihodologies and inexpensjle soflware have allo\\,.d us 1.) rerjse prellousl) obrened iìala ser and coDvert ùom 2D to lD dara These numeric data will làcìlirate dre archaeological .omprehension and koo§,ledge ofthe casrle: morcover.jn perspecti!e the) §,ill condbute to iri preservarìon. sa1èguard, ùuirioD and vìl1ual .ecoDsfuclion. Tìrough this expedmenrador. based on simplc ard evident crite.ia, we have reached rhe airrs defincd ù,ithout a ftìfihcr.ìata acquisition su'e). The lD rrodcl is cullentl] the onl] restituiion ofrhe e\isting sratus of the casde. For flrture devebpnents will be re-organiTed rhe dala, rhrough oper source ard automatic phologrammerrlc soli\\,are. for lìiÌing the poinl clouds. This expeimentatìon $,ill let us to flll areas not jnvestigated such as the nddl àround the casde. lD this way it will be possible to contextÌalize all dre archaeolosical aree ollhe crusade. castlc of {1-Irlr 7rrd.

Al lasl \e ha\. loi

REFERENCES l1l Ahsì T. M., Dcl Bnìbo A., D.ap P., Gàbrìelli R, Nucciotri M., "Archaeologicat ilvestisatjon

l2l

ar

Shawbak: conbmed use ofmeasuremenr and archaeological recording", tu Vannnri C , Nucciolij M., DÒ Pealt a Shd$bolt ..lrcheolagid diuna Frcaliera. Giunli Editore, Firenze, (2009) 166 t?: Bini N4, "Surveyìng and fomlal delinjtion oI ihe castles of Wu'Ayra, Hàbis and Shawbak", ìn Varnuli G. Nuccìotil M., Dd Petru .t Shawbdk. Archeologia lli una Frontierd, Giunrl Edilore.

!irerze, (2009)

102-10s

:6

':.::li::: a \ri:rr.r Lì. r,.ir.r ::: "'::: :,i ir g.r/iore u€l -::: : . i' :'i :: i: : :.-:::::..:l::i:l:..!.r.:ll.Eiui.::l . :ì:::::i:l G::,:.. ir::.2r. illoll19 :: : : ,1. r-::.::: r. \:.:r:ìrrr !1. \1.:ri::r \1. \1-:r.l L '.: ', :.:::i.. : t):!gtdrrìì.n.ì.1 :.::.r:.:irr.,,i\llrjl.1ll

.:iì:r D.rLdÈri.ì. "ll.air.llÒ

-11 r:olerl1r!..i.i r..ri: R. _::. I ì lli--.1-: dr \\:ua\ra. l.mr. ar.ì-,eol.!irh. ::: a.ij:. " .. ..-r r:r ir.:i,,r ! rr

..:,:.r,'ri,i...i.r,rr:r..(l..i,go.Pro.e.din!;rrinre\o-,ii:ri:.1 ,1: ,r.1,:i::i l l-rr..7e , ..d ..:'. .ì,. \ drmi G . Nu..ionj 11., Tonghnn C Vanni Deslderi 1 . ll:,ìi:..:-l f::::. i..|:::,ì L-r:! .r rie

Crusader-A)tubid ior(ilìcd insrallitio.s ir Trans-Jordan..\..haiol.::.: \il;;rL.Ìr.::ìrLxr:rsrr\ ol r..lor.ncc', in ( i|illò Lì.1 PdsrttÒ iidlq. del pte\e"i.: -]tur;.,i: :j .:-- :.r ..:.-,i, (;r.rd-.r,-.r

iAnmaù.

L_ ottobre 10021 Anrnren.

il00:l l8l-lt)0

I Finll o ho|hoto mosaìc ofrreal Wu'Ayra's .:srl3 oblarned ùiih thc photoglanmetrìc solì\lire

l:l-s