Sea-level change and shore-line evolution in Aegean Greece since Upper Paleolithic time.

Sea-level change and shore-line evolution in Aegean Greece since Upper Paleolithic time. 'As the glaciation ended, the ice melted and the sea-levelrose.' Yes - but it has not been as simple as that, as the Earthhas adjusted in several ways to the changing surface-loads it suffersunder ice and under weight of water. The important issues are set out ina simple mathematical treatment, and their varied consequences are shownfor Greece and especially for the Greek coastal plains and the Greekislands, where the impact on human settlement has been large.Nature and consequences of Postglacial post��gla��cial?adj.Relating to or occurring during the time following a glacial period.postglacial?Relating to or occurring during the time following a glacial period.Adj. 1. sea-level changeSea-levels have changed significantly since Late Palaeolithic time,primarily in response to the melting of the great ice-sheets thatcovered northern Europe and North America North America,third largest continent (1990 est. pop. 365,000,000), c.9,400,000 sq mi (24,346,000 sq km), the northern of the two continents of the Western Hemisphere. . These ice-sheets were of asufficiently large In mathematics, the phrase sufficiently large is used in contexts such as: is true for sufficiently large volume that, upon melting, sea-level rose globally byabout 120-130 m. Release of this ice into the oceans was initiated atabout 18,000 years before present (b.p.), although the majority of themelting occurred between about 16,000 and 8000 b.p.(1) Rates of globalsea-level rise reached 15-20 mm per year during this interval. As thesea-levels rose, so did the shorelines migrate with time, at rates thatfor some low-lying regions reached about a kilometre per year. Examplesof where such rapid encroachment of the sea occurred include the PersianGulf and the Gulf of Carpentaria Noun 1. Gulf of Carpentaria - a wide shallow inlet of the Arafura Sea in northern AustraliaCarpentariaAustralia, Commonwealth of Australia - a nation occupying the whole of the Australian continent; Aboriginal tribes are thought to have migrated from of northern Australia The term northern Australia is generally considered to include the States and territories of Australia of Queensland and the Northern Territory. The part of Western Australia (WA) north of latitude 26�� south — a definition widely used in law and State government policy , between about12,000 and 10,000 years b.p. In some areas of the world the sea-levelspeaked at about 6000 years b.p., inundating now low-lying areas beforefalling slowly to their present position. The consequences of thesechanges on human settlement and movement have been recognized in thearchaeological and pre-historic records. Thus it is widely accepted thatlevels during the Last Glacial Maximum The Last Glacial Maximum (LGM) refers to the time of maximum extent of the ice sheets during the last glaciation (the W��rm or Wisconsin glaciation), approximately 20,000 years ago. This extreme persisted for several thousand years. , about 20,000 to 18,000 yearsago, were sufficiently lower than today to leave exposed coastal plainsthat have since flooded. But less attention appears to have been focusedon the timing and rates of change after the onset of melting of thegreat ice-sheets. What discussion there is - with the exception of theimportant paper by van Andel & Shackleton (1982) (see also van Andel1989) - often leaves the distinct impression that this change in leveloccurred early and quickly with rather minimal human impact.This paper sets out to describe, using the Aegean Sea Aegean Sea,Gr. Aigaion Pelagos, Turkish Ege Denizi, arm of the Mediterranean Sea, c.400 mi (640 km) long and 200 mi (320 km) wide, off SE Europe between Greece and Turkey; Crete and Rhodes mark its southern limit. region as anexample, a realistic model of sea-level change and shoreline migrationfor the past 20,000 years, one that can provide a frame-work fordiscussing impacts of such change on human movements and settlement.If, during the decay of the ice sheets, the meltwater melt��wa��ter?n.Water that comes from melting snow or ice.meltwaterNounmelted snow or iceNoun 1. volume isdistributed uniformly over the oceans, then the sea-level change at timet would be[Delta][[Zeta].sub.e](t) = change in ocean volume/ocean surface area(1)This 'eustatic sea-level change' is a function of time. Itrepresents only a zero-order approximation because sea-level does notrespond uniformly to the melting of the ice caps: the rates of rise arespatially variable, and in some localities sea-level may actually befalling relative to the land. This is a consequence of the adjustment ofthe Earth to the changing surface-loads of ice and meltwater. TheEarth's response can be described as that of an elastic layer (thelithosphere lithosphere(lĭth`əsfēr '), brittle uppermost shell of the earth, broken into a number of tectonic plates. The lithosphere consists of the heavy oceanic and lighter continental crusts, and the uppermost portion of the mantle. which includes the crust) overlying overlyingsuffocation of piglets by the sow. The piglets may be weak from illness or malnutrition, the sow may be clumsy or ill, the pen may be inadequate in size or poorly designed so that piglets cannot escape. a viscoelastic Adj. 1. viscoelastic - having viscous as well as elastic propertiesnatural philosophy, physics - the science of matter and energy and their interactions; "his favorite subject was physics" mantle.When changes in the mass distribution occur on the Earth's surface,the lithosphere and mantle respond to the new stress state in differentways: elastic deformation primarily occurs in the lithosphere andviscous flow occurs primarily in the underlying mantle. Thecharacteristic time-scale of this flow is of the order of a few thousandyears. The deformation of the Earth's surface under a changing leadtherefore exhibits both an instantaneous elastic response and a viscousresponse. Such behaviour is well documented by other geophysicalobservations: gravitational grav��i��ta��tion?n.1. Physicsa. The natural phenomenon of attraction between physical objects with mass or energy.b. The act or process of moving under the influence of this attraction.2. attraction of the Sun and Moon raises tidesin the solid Earth; ocean tides lead the sea floor and contribute to thedeformation of the Earth's surface; atmospheric pressure atmospheric pressureor barometric pressureForce per unit area exerted by the air above the surface of the Earth. Standard sea-level pressure, by definition, equals 1 atmosphere (atm), or 29.92 in. (760 mm) of mercury, 14.70 lbs per square in., or 101. fluctuations over the continents induce deformations in the solid Earth.These displacements, when measured with precision scientificinstruments, show both an elastic and a viscous component, with thelatter becoming increasingly important as the duration of the lead orforce increases. Changes in the ice and water loads occurring at timesof major deglaciation de��gla��ci��a��tion?n.The uncovering of glaciated land because of melting or sublimation of the glacier.deglaciation?The uncovering of land that was previously covered by a glacier. are much larger and occur on longer time-scalesthan these small atmospheric and tidal perturbations, but the theoryunderpinning the formulation of the surface rebound problem is similarand has been tested against a range of different observations (e.g.Lambeck 1988).When the ice-sheets melt, the surface lead is changed in two ways.Ice is removed, thereby unloading the formerly glaciated crust, and themeltwater is added into the oceans, loading the oceanic lithosphere. Theresulting change in sea-level relative to the land is a combination oftwo reactions: the increase in ocean volume, and the deflection of theland surface. Depending on which contribution is greater, sea-level isseen to rise or fall relative to the land. In regions near or within theformer centres of glaciation, the crustal crust��al?adj.Of or relating to a crust, especially that of the earth or the moon.Adj. 1. crustal - of or relating to or characteristic of the crust of the earth or moon adjustment is one of uplift;it is usually the more important, and sea-level here is seen to fall.Further away from the ice sheets, the crustal rebound is small andgenerally one of subsidence subsidence,lowering of a portion of the earth's crust. The subsidence of land areas over time has resulted in submergence by shallow seas (see oceans). Land subsidence can occur naturally or through human activity. ; here the change in ocean volume dominates,and sea-level is seen to rise. Because of the viscous nature of theplanet's interior, the crustal deformation continues long after themelting of the Late Pleistocene The Late Pleistocene (also known as Upper Pleistocene or the Tarantian) is a stage of the Pleistocene Epoch. The beginning of the stage is defined by the base of Eemian interglacial phase before final glacial episode of Pleistocene 126,000 �� 5,000 years ago. ice-sheets has ceased; and shorelineshave continued to change up to the present.This combined behaviour of the eustatic change and the crustalrebound describes a first-order solution to the question of what happensto sea-level when ice sheets melt. A more complete description requiresconsideration of two additional effects: * the fact that newly added meltwater loads the sea floor andstresses the mantle, resulting in further mantle flow and surfacedisplacement; * the gravity field of the earth is changing; first, because of thechanging gravitational attraction between the ice, water and land as theice sheets melt, and second, because of the redistribution of masswithin the solid Earth in response to the mantle flow and surfacedeflection.The total effect of these various processes is called'glacio-hydro-isostasy' alter the (regional) isostatic i��sos��ta��sy?n.Equilibrium in the earth's crust such that the forces tending to elevate landmasses balance the forces tending to depress landmasses. response of the Earth's surface to the ice and water loads. Itleads to a complex spatial and temporal pattern of sea-level change asthe ice sheets wax and wane, making the concept of a uniform globaleustatic change of only very limited value. This complexity is more thana scientific curiosity, as its understanding has broad scientificimpacts. In geophysics, for example, the spatial changes in sea-levelprovide a valuable insight into physical properties of the Earth thatcannot otherwise be measured. In environmental studies they provide aframework for separating natural from man-caused changes in sea-level.In prehistory prehistory,period of human evolution before writing was invented and records kept. The term was coined by Daniel Wilson in 1851. It is followed by protohistory, the period for which we have some records but must still rely largely on archaeological evidence to and archaeology, they provide a basis for detailedgeographic reconstructions of coastal regions and for assessing theimpact of changing sea-levels on past societies.Models for sea-level changeThe evaluation of the evolution of shorelines during times ofglaciation and deglaciation requires knowledge of several quantities: * the change in sea-level relative to the land for the region inquestion, a change that is usually regionally variable; * the present topography of the low-lying land areas and shallowsea-floor bathymetry ba��thym��e��try?n.The measurement of the depth of bodies of water.bathy��met for the region; * any information on tectonics, sedimentation or erosion that mayhave changed the geometry of the coastal zone.Because of the spatial variability Spatial variability is characterized by different values for an observed attribute or property that are measured at different geographic locations in an area. The geographic locations are recorded using GPS (global positioning systems) while the attribute's spatial variability is of the sea-level change, the bestway to establish a record of the past positions at a particular locationis to infer it from the geological, geomorphological ge��o��mor��phol��o��gy?n.The study of the evolution and configuration of landforms.geo��mor and archaeologicalrecord The archaeological record is a term used in archaeology to denote all archaeological evidence, including the physical remains of past human activities which archaeologists seek out and record in an attempt to analyze and reconstruct the past. . But observations that permit a quantitative estimate of thechanging position of the sea surface seldom exist for the area ofinterest and - at best - it probably consists of a partial time series.Also, such records become increasingly rare and imprecise the furtherone goes back in time. What is required is a physical model forinterpolation interpolationIn mathematics, estimation of a value between two known data points. A simple example is calculating the mean (see mean, median, and mode) of two population counts made 10 years apart to estimate the population in the fifth year. between the available fragments of information so as to beable to predict sea-level at any place and time. For tectonically stableregions the glacio-hydro-isostatic theory provides a relatively smoothlyvarying and predictable function of such change in time and space thatcan be used to interpolate between isolated observations. But whenactive vertical tectonic processes are important, resulting in surfacefaulting or warping, the interpolation model becomes much more difficultbecause the crustal displacements can change abruptly over comparativelyshort distances and need not be uniform in time when viewed over periodsof the order of millennia.The approach used here is to develop models of eustasy andglacio-hydro isostasy isostasy(īsŏs`təsē): see continent. isostasyTheory describing the mass balance in the Earth's crust, which treats all large portions of the crust as though they were floating on a denser underlying layer, about 70 for tectonically stable areas so that physicalparameters describing the Earth response and the ice sheets can beevaluated. The result is a comprehensive portrayal results of globalsea-level change in response to the growth and decay of the large icesheets. When these models are applied to tectonically active areas anydepartures of observed sea-levels from the model-predicted values can beattributed to tectonic factors. The success of this approach does dependon whether the isostatic predictions are representative of the regionunder consideration. If this can be demonstrated, then the model can beused to predict the positions of former shorelines and palaeo-waterdepths.The sea-level equationIn the presence of vertical tectonic motions of the Earth'ssurface, the relative sea-level change [Delta][Zeta]([Phi], t) at a site[Phi] and time t can be written schematically as:[Delta][Zeta]([Phi], t) =[Delta][[Zeta].sub.e](t)+[Delta][[Zeta].sub.I]([Phi],t)+[Delta][[Zeta].sub.T]([Phi], t) (2)where [Delta][[Zeta].sub.e](t) is the previously defined eustaticsea-level (1), [Delta][[Zeta].sub.I]([Phi], t) is the combinedglacio-hydro-isostatic contribution, and [Delta][[Zeta].sub.T]([Phi], t)is any additional tectonic contribution. This last contribution isdiscussed further below and the emphasis here is on the isostaticfunction, which can be written as the sum of three parts:[Delta][[Zeta].sub.I]([Phi], t) = [Delta][[Zeta].sub.r]([Phi], t) +[Delta][[Zeta].sub.i]([Phi], t) + [Delta][[Zeta].sub.w]([Phi], t) (3)The total sea-level change is expressed as:[Delta][Zeta]([Phi], t) = [Delta][[Zeta].sub.e](t) +[Delta][[Zeta].sub.r]([Phi], t) + [Delta][[Zeta].sub.i]([Phi], t) +[Delta][[Zeta].sub.w]([Phi], t) + [Delta][[Zeta].sub.T]([Phi], t) (4)The eustatic term [Delta][[Zeta].sub.e](t) in (4) is a function ofthe rate of change in ocean volume and its evaluation requires aknowledge of the rate of melting of the totality of the ice caps. Thesecond term [Delta][[Zeta].sub.r]([Phi], t) describes the departures ofsea-level change from eustasy on a rigid planet by allowing for thechange in gravitational attraction between the ice and water as the icesheets melt, and between the water and land as the ocean volumesincrease. During times of ice-sheet growth, the increasing gravitationalattraction of the ice pulls the water towards the expanding ice dome; inthe absence of the other contributions, sea-level rises in theneighbourhood of the ice while further away it falls.[Delta][[Zeta].sub.r] is therefore a function of the ice sheet geometrythrough time and of the geometry of the ocean basin from which the wateris extracted. This term, most significant in areas close to the icesheets, remains non-negligible out to considerable distances beyond theice margins. The attraction by the northern ice is still sufficient inGreece, for example, to 'pull up' the Mediterranean water, andthis correction is positive at times of major glaciation [ILLUSTRATIONFOR FIGURE 2A OMITTED]; sea-levels at these early epochs would have beenhigher than at the present if the other contributions had been ignored.The major contribution is from the Fennoscandian ice, but contributionsfrom the North American North Americannamed after North America.North American blastomycosissee North American blastomycosis.North American cattle ticksee boophilusannulatus. and Antarctic ice sheets are not negligible ifsea-level models accurate to a metre or less are sought.The term [Delta][[Zeta].sub.i]([Phi], t) in (4) describes theconsequence on sea-level of the deformation of the Earth under thechanging ice load; it includes a contribution from changes in thegravity field associated with the redistribution of mass within thedeforming planet. This 'glacio-isostatic correction' is afunction of the elastic and viscous properties of the Earth, as well asof the temporal and spatial distribution of the ice sheets. This term isalso most important for the former areas of glaciation where the crusthas been depressed by the ice load. As the ice melts, the crust reboundsat a rate determined by the mantle viscosity and by the ice thickness.For large ice loads, crustal rebound exceeds the eustatic change so thatsea-level appears to fall relative to the land. This is seen in the Gulfof Bothnia Noun 1. Gulf of Bothnia - a northern arm of the Baltic Sea; between Sweden and FinlandAaland islands, Ahvenanmaa, Aland islands - an archipelago of some 6,000 islands in the Gulf of Bothnia under Finnish control , for example, where there are raised shoreline features whoseages increase with height [ILLUSTRATION FOR FIGURE 1A OMITTED]. Rates ofrelative sea-level fall in excess of 20 mm per year have been inferredfor some localities. Because of the viscous nature of the mantle flow,the rebound and relative sea-level change continues long afterdeglaciation has ceased. For small ice sheets such as that formerly overBritain, or regions near the margins of the ice sheets, as in southernNorway or Denmark, the crustal rebound and the change in relativesea-level was much smaller, the latter remaining within a few tens ofmetres of the present level since Late Glacial time [ILLUSTRATION FORFIGURE 1B OMITTED]. Here, the crustal rebound is less than for the sitescloser to the former centres of ice loading; it is of comparablemagnitude but of opposite sign to the eustatic contribution. Furtheraway again, beyond the areas of former glaciation, the response of thecrust to the removal of the ice is one of subsidence in reaction to theflow of the underlying mantle material towards the formerly loadedareas. Here sea-level appears to be rising to the present, even thoughall melting ceased much earlier. This is the case, for example, insouthern England, or along the Atlantic margin of France [ILLUSTRATIONFOR FIGURE 1C OMITTED]. In the Aegean, sea-level also appears to berising throughout the Holocene because of the crustal response to themelting of, primarily, the Scandinavian ice sheet.The hydro-isostatic term [Delta][[Zeta].sub.w]([Phi], t) describesthe contribution to the sea-level change from the meltwater loading ofthe crust and the associated change in gravitational potential. The termis a function of the Earth's elastic and viscous parameters (itsrheology) and of the relative sea-level change itself, as well as of theshifting coastlines. Small compared with the maximum glacio-isostaticand eustatic terms, it becomes dominant in areas away from the icesheets at postglacial times. At continental margin sites, thecharacteristic hydro-isostatic signal is a falling sea-level as the seafloor slowly subsides under the new water load [ILLUSTRATION FOR FIGURE1D OMITTED], but only for about the past 6000 years does it dominateover the other contributions; relative sea-levels will have been higherthan today from about 6000 years b.p. to the present. The amplitudes ofthe highstands are small, up to about 3 m depending on the coastalgeometry, but in flat and low-lying areas such as lower Mesopotamia orthe southern shore of the Gulf of Carpentaria, this can have led tosubstantial inundation INUNDATION. The overflow of waters by coming out of their bed. 2. Inundations may arise from three causes; from public necessity, as in defence of a place it may be necessary to dam the current of a stream, which will cause an inundation to the upper lands; in mid-Holocene time.For the Aegean region all three contributions to the isostatic term(3) are important, as FIGURE 2 illustrates, for a model discussed indetail elsewhere (Lambeck 1995a). FIGURE 2a illustrates the variation inthe rigid term [Delta][[Zeta].sub.r] over the region at 18,000 and10,000 years b.p. This corrective term is positive, indicating that thegravitational pull on the water by the ice over Fennoscandia isimportant out to distances of about 2000-2500 km from the centre of theice load. FIGURE 2b illustrates the ice load term [Delta][[Zeta].sub.i]for the same two epochs. This term is negative in this region of crustalsubsidence upon ice unloading [ILLUSTRATION FOR FIGURE 1C OMITTED], andpartly cancels out the rigid term. The water-load term is illustrated inFIGURE 2c. This correction is mainly positive over the land and negativeat sea, and the contours tend to follow the coastline, their relativelysmooth character illustrating the filtering effect of the elasticlithosphere on the crustal deformation.Together, the three corrective terms in the Aegean region result insea-level at the time of the Last Glacial Maximum varying between -115and -135 m, compared to the eustatic value of about -125m for thisepoch. At 10,000 years ago the predicted sea-levels over the region varyfrom -43 to -55 m, compared to a eustatic level of about -43 m. Theactual amplitudes and details of the spatial patterns depend on theparameters used to describe the ice load and the Earth's rheologybut the overall trends and magnitudes are representative of thedepartures of the sea-level change from the simple eustaticapproximation.Other than predicting temporal and spatial changes in sea-level, theglacio-hydro-isostatic model also predicts the location of pastshorelines and palaeo water depths if the present water depth[H.sub.o]([Phi]) at location [Phi] is known. That is, the water depth ata time t and position [Phi] are related to [Delta][Zeta]([Phi], t) byH([Phi], t) = [H.sub.o]([Phi]) - [Delta][Zeta]([Phi], t) (5)To evaluate the spatial and temporal distribution of sea-level andshoreline change during a glacial cycle several requirements have to bemet. These include: * A model of the eustatic sea-level function[Delta][[Zeta].sub.e](t) - usually inferred from the sea-levelobservations themselves because independent estimates of the volumes ofthe former ice sheets do not exist. These estimates are derived fromlocalities where the isostatic corrections are believed to be small suchthat they can be estimated in an iterative it��er��a��tive?adj.1. Characterized by or involving repetition, recurrence, reiteration, or repetitiousness.2. Grammar Frequentative.Noun 1. manner (Nakada & Lambeck1988). Many published estimates of the eustatic sea-level curve haveignored the isostatic contributions: the resulting errors may beparticularly important when estimates are based on sea-level data fromthe North Atlantic margin of the United States of America UNITED STATES OF AMERICA. The name of this country. The United States, now thirty-one in number, are Alabama, Arkansas, Connecticut, Delaware, Florida, Georgia, Illinois, Indiana, Iowa, Kentucky, Louisiana, Maine, Maryland, Massachusetts, Michigan, Mississippi, Missouri, New Hampshire, where theglacio-isostatic effects are significant. * A detailed description of the growth and decay of the ice sheets,required to calculate the glacio-isostatic term. For localities near orwithin former ice-sheet limits, a detailed description of the evolutionof the ice load is required whereas for regions away from the icemargins, such as the Mediterranean, approximate spatial descriptions ofthe changing ice-sheets suffice. * A model of the Earth's rheology, describing the response ofthe planet to surface loading on time-scales of thousands of years. Theelastic behaviour of the Earth is well known from seismological seis��mol��o��gy?n.The geophysical science of earthquakes and the mechanical properties of the earth.seis studiesbut the viscous properties, less well known, are estimated fromcomparisons of isostatic-model predictions with observations fromregions of otherwise tectonic stability. * A description of the coastline geometry and the shallow-waterbathymetry, required to estimate the palaeo-shorelines through therelation (5). The time-dependence of this coastline must also beconsidered in the calculation of the hydro-isostatic term, particularlyif high accuracy local solutions are sought for a region such as theeastern Mediterranean. This is usually achieved by iterating betweenequations 4 and 5.The eustatic sea-level curvePublished estimates of the eustatic sea-level curve are varied, thereasons for which are several. Some are associated with interpreting ofthe observational evidence. Corals, as for example used to estimate thesea-level curve at Barbados (Fairbanks 1989), provide estimates of alower limit to the sea-level curve, while submerged in situ In place. When something is "in situ," it is in its original location. terrestrialvegetation such as fresh-water peats or tree stumps provide an upperlimit. Estimates based on the depth-age relations of molluscs are oftenunreliable: their original depth-range relative to mean sea-level islarge, or they may have been transported from their growth position tonew levels. The different published results may also reflect theneglect, or inadequate elimination, of the glacio-hydro-isostaticfactors. For example, a number of the still-quoted estimates used insea-level studies in the Mediterranean are based on old observationsfrom the Atlantic coast of North America (e.g. Curray 1965; Milliman& Emery 1968) where the relative sea-level is influenced by theglacio-isostatic rebound of the Laurentide region such that the observedlevels in Late Glacial time generally lie above the eustatic curve[ILLUSTRATION FOR FIGURE 1C OMITTED].Across continental margins, corrections for the hydro-isostaticfactors in Late Glacial times are also significant, reaching 20-30 m atthe time of the Last Glacial Maximum (Lambeck & Nakada 1990).Observations from islands far from both continental margins and icesheets provide better estimates of the eustatic sea-level, but eventhese are not immune to the isostatic factors (e.g. Nakada & Lambeck1987; Mitrovica & Peltier 1991). As the establishment of a globaleustatic sea-level curve requires models for the isostatic corrections,the procedure adopted here is to apply the model (4) to areas of theworld where other tectonic processes producing vertical movements arebelieved insignificant, or where corrections for these movements can beindependently made (e.g. Lambeck 1993; 1995a). FIGURE 3 illustrates thebest estimate of the eustatic sea-level curve since the time of the LastGlacial Maximum; it is based on analyses of sea-level change fromlocations far from the ice sheets, along the continental margin ofAustralia and elsewhere. This result is consistent with the upward coralreef coral reefRidge or hummock formed in shallow ocean areas from the external skeletons of corals. The skeleton consists of calcium carbonate (CaCO3), or limestone. A coral reef may grow into a permanent coral island, or it may take one of four principal forms. growth-rate curve from Barbados (Fairbanks 1989) when that iscorrected for the isostatic perturbations, as well as with the resultsfrom the rapidly uplifting Huon Peninsula Huon Peninsula is a large peninsula in Morobe Province, eastern Papua New Guinea, at . It is named after French explorer Jean-Michel Huon de Kermadec. The peninsula is dominated by the Saruwaged Range. of Papua New Guinea Papua New Guinea(păp`ə, –y (Chappell& Pollach 1991) once tectonic and isostatic corrections are applied.Major contributions to this globally integrated rate of melting comesfrom the Laurentian, Fennoscandian, Barents Sea Barents Sea,arm of the Arctic Ocean, N of Norway and European Russia, partially enclosed by Franz Josef Land on the north, Novaya Zemlya on the east, and Svalbard on the west. and Antarctic icesheets.One characteristic of this sea-level curve is that the bulk ofmelting ceased at about 6000 years ago when the last ice from Laurentiavanished. (The Fennoscandian ice sheet vanished soon after 9000 yearsb.p.) But a small amount of additional meltwater continued to be addedinto the oceans after this time, probably from a small reduction in theAntarctic ice volume, so as to raise the eustatic level by about 2-3 mduring the past 6000 years (Nakada & Lambeck 1988). Anothercharacteristic of this eustatic sea-level curve is that the function isdevoid of rapid oscillations oscillationsSee Cortical oscillations. . Detailed studies from the same or close-bysites of sea-level change during the past 6000 years indicate littleevidence for such oscillations once the major ice sheets have melted orapproached their present volumes (Chappell 1983; Hyvarinen 1980).Likewise, estimates of the eustatic sea-level function for the periodbetween 12,000 and 6000 years b.p. suggest that any oscillations in thisinterval are likely to be less than about 2 m (Lambeck 1993; in press).Because of the viscous nature of the response of the Earth tochanging surface loads, the isostatic contributions to the sea-levelchange are effective long after the changes in loading have occurred, asis best seen in the on-going uplift of the Fennoscandian area[ILLUSTRATION FOR FIGURE 1A OMITTED]. The prediction of sea-level at anyperiod, therefore, requires information about the earlier ice sheethistories. For studies of sea-level near the time of the Last GlacialMaximum, this requires a knowledge of the ice sheets during their growthphase. As this information is generally scanty, reliable models for thechanging volumes of the individual major ice sheets before the LastGlacial Maximum do not exist. Estimates of the early eustatic sea-levelcurve also cannot be established from sea-level information alonebecause of the great paucity of older observations preserved in thegeological record and which can be reliably dated. Instead, anapproximate estimate of this function is obtained from oxygen isotoperecords of the sea-floor sediments scaled by the more directly observedchanges for the past 18,000 years (e.g. Shackleton 1987; Chappell &Shackleton 1986). FIGURE 3 illustrates this result back to 150,000 yearsb.p. The last time sea-levels were near their present value occurredabout 120,000 years ago when climate and ice volumes were last similarto those of the latter part of the Holocene. In between these twointerglacial in��ter��gla��cial?adj.Occurring between glacial epochs.n.A comparatively short period of warmth during an overall period of glaciation. periods, the oxygen isotope information shows the peakglacial conditions persisting for only a relatively short interval,although for much of the time between about 70,000 years b.p. and theGlacial Maximum, ice volumes significantly exceeded those of today suchthat the global sea-levels did not rise above 40-50 m below presentlevel during this interval.Ice sheet modelsThe main sources of meltwater were the ice-sheets over North America,northern Europe including the Barents Sea, and a larger-than-presentice-sheet over Antarctica. These ice caps contributed about 70, 25 and25 m respectively to the eustatic sea-level change since the time of theLast Glacial Maximum. The retreat of the northern ice sheets isreasonably well constrained over the continents by geological andgeomorphological observations (e.g. Denton & Hughes 1981), but theextent over the shallow seas such as the Barents Sea is less well known,although observations of raised shorelines from islands in such areascan be used to constrain the ice volumes (Lambeck 1995b). Even over thecontinents the ice thickness is usually not well constrained fromindependent observations and is usually inferred from ice-modelconsiderations and from the interpretation of raised shorelines withinthe limits of the areas of former glaciation (e.g. Tushingham &Peltier 1991; Lambeck 1993). Observational evidence of sea-level changewithin and immediately outside the former areas of glaciation hasgenerally been sufficient to arrive at consistent northern ice sheetmodels since the time of the Last Glacial Maximum. More problematic hasbeen the estimation of the volume changes in the Antarctic ice sheet The Antarctic ice sheet is one of the two polar ice caps of the Earth. It covers about 98% of the Antarctic continent and is the largest single mass of ice on Earth. The total ice mass on the Earth covers an area of almost 14 million square km and contains 30 million cubic km of since this time. Some observations of raised shorelines from the marginof this ice sheet suggest a reduction in ice volumes since the glacialmaximum (Zwartz et al. in preparation) but the evidence is inadequate toestablish a quantitative model. Instead, inferences of past Antarcticice volumes are based on more indirect indicators, such as the need forthe total ice volume to correspond to the observed eustatic sea-levelcurve and from the pattern of Late Glacial and Postglacial sea-levelchange in southern latitudes.Earth model parametersEssential to predictions of the isostatic sea-level corrections is amodel for the elastic and viscous properties of the Earth. The elasticparameters and the radial density distribution are obtained from theanalysis of seismic wave seismic waveVibration generated by an earthquake, explosion, or similar phenomenon and propagated within the Earth or along its surface. Earthquakes generate two principal types of waves: body waves, which travel within the Earth, and surface waves, which travel along the velocities through the planet, quantitiessufficiently well known for the elastic deformation to be calculated andtested against known short-period tidal deformations of the Earth. Theviscosity structure, less well known, has to be inferred from theisostatic analysis itself. It is this geophysical problem that hasprovided the major impetus for the study of past sea-level (Cathles1975; Peltier & Andrews 1976; Nakada & Lambeck 1987; Mitrovica& Peltier 1993). The assumed viscosity models are relatively simple,comprising a mantle with a few layers of different viscosity values. Formost predictions of palaeosea-levels and shorelines, a three-layeredmodel gives adequate predictive capabilities; studies from differentlocalities produce largely consistent results (Lambeck et al. 1990;Lambeck & Nakada 1990; Lambeck 1993). Such models comprise anessentially elastic lithosphere of 60-80 km thickness, an upper mantle(down to the 670 km depth seismic discontinuity) with an effectiveviscosity of (2-5)[10.sup.20] Pa s, and a lower mantle Noun 1. lower mantle - the deeper part of the mantlelayer - a relatively thin sheetlike expanse or region lying over or under anothermantle - the layer of the earth between the crust and the core with a viscosityof about [10.sup.22] Pa s.Whether the lithospheric thickness and mantle viscosity parametersare laterally uniform is an important geophysical question and someevidence suggests they may vary from region to region, depending on thetectonic history of the crust. For tectonically active areas such as theAegean, the lithosphere can be expected to be thinner than for thestable tectonic terrains of northern Europe; likewise, the upper mantleviscosity for the tectonic regions are likely to be less in the Aegeanas well. Ideally the mantle parameters should be estimated fromsea-level data from the region of interest. But for the Aegean theobservational data-base is limited and, furthermore, the evidence iscontaminated by tectonic contributions. But - by a largely fortuitoustrade-off between the earth-model parameters - sea-level predictions forthin-lithosphere, lower-mantle-viscosity models are very similar tothose for thick-lithosphere, higher-mantle-viscosity models (Lambeck1995a; Lambeck et al. 1996) and it appears that parameters estimatedfrom well-constrained rebound problems for northern Europe constrainreasonably well the sea-level change models for the easternMediterranean. Until a better observational record is established, theearth models for the northern European region are used for the Aegeanpredictions below.The Aegean regionObservations of sea-level changeThe evidence for sea-level change along the Greek coastline isinferred from a variety of geological and archaeological indicators, thelatter being more plentiful for about the last 4000 years. Thearchaeological evidence has been examined in detail by Flemming (1978),particularly for the coast of the Peloponnese and southwestern Turkey.The evidence consists of the positions of structures that, at their timeof construction, are believed to bear a known relation to the positionof the sea. In other instances, the evidence provides only constraintson sea-level. For example, offshore from Franchthi Cave Franchthi cave (or Frankhthi cave, Greek Σπήλαιον Φράγχθη) in the Peloponnese, in the southeastern Argolid, is a cave overlooking the Argolic Gulf opposite the Greek village of Koilada. , near Koilas(2)in the southern Argolis Peninsula, a Neolithic site points to the seahaving been at least 11 m lower than today between 7610[+ or -]150 and6220[+ or -]130 years b.p. (Jacobsen & Farrand 1987; van Andel1987). Likewise, at Saliagos on the Cycladean island of Andiparos,now-submerged Early Bronze Age Bronze Age,period in the development of technology when metals were first used regularly in the manufacture of tools and weapons. Pure copper and bronze, an alloy of copper and tin, were used indiscriminately at first; this early period is sometimes called the remains point to sea-levels before 5000years b.p. having been at least 5 m lower than today (Morrison 1968).The geological evidence for past sea-levels in the Aegean andadjacent areas takes a number of forms, including the depths ofsubmerged terrestrial or lagoonal vegetation and sediments, inferencesdrawn from seismic reflectors in shallow offshore sediments, and theage-height relation of marine solution notches. The evidence from thelagoonal and terrestrial sediments comes mainly from the shallow upperregions of the Gulfs of Messenia (Kraft & Rapp 1975), Lakonia andArgolis (Kraft et al. 1977), and from the Bay of Navarone (Kraft et al.1980) near Pylos, all in the Peloponnese. At these sites sea-levelsappear to have been rising relative to the land for at least the last10,000 years, at rates that may have been temporally and spatiallyvariable, although the quality of the data is generally insufficient toquantify more than the general trends. An important inference fromseismic stratigraphy stratigraphy,branch of geology specifically concerned with the arrangement of layered rocks (see stratification). Stratigraphy is based on the law of superposition, which states that in a normal sequence of rock layers the youngest is on top and the oldest on the is that sea-level at about 18,000 years agooffshore from the southern Argolis Peninsula was about 120 m lower thantoday (van Andel & Lianos 1983). The solution-notches produced bymarine borers are preserved only when rapid uplift events occur suchthat the notch, formed at about mean sea-level, is lifted beyond thetidal range. Well-developed sequences occur, for example, in westernCrete, Rhodes, Karpathos and the Kythera islands (Flemming 1978;Thommeret et al. 1981) as well as in the Gulf of Corinth Noun 1. Gulf of Corinth - inlet of the Ionian Sea between central Greece and the PeloponnesusGulf of LepantoIonian Sea - an arm of the Mediterranean Sea between western Greece and southern Italy (Pirazzoli etal. 1994), at elevations of up to 10 m and with estimated ages of up to4000-5000 years.Tectonic contributions to sea-level changeTectonic movements and deformation of the crust occur over a widespectrum of length and time-scales, from the slow and nearly uniformglobal plate tectonic motions to localized and sudden earthquakedisplacements. These movements are surface expressions of geophysicalmechanisms operating internally to the Earth, and their records containmuch information on the processes shaping the planet. The geologicalrecord provides many examples of these surface displacements,particularly in the vertical direction which contribute to the relativechange in the positions of the land and sea surfaces. The easternMediterranean is a region of active tectonics and Greece in particularis one of the most rapidly deforming continental areas on Earth (Jackson1994). Other examples of vertical movements occur in response to changesin the surface loads when crustal material is eroded and depositedelsewhere. This process, similar to the glacial unloading problem, isusually more localized. In Greece, the sediment transport into theshallow bays and gulfs is comparatively small and this loadingcontribution is unlikely to have been important on time-scales of 10,000or so years. The deposition of sediments may, however, have contributedlocally to the build-up of the land surface and migration of shorelinesin some bays.Usually the processes producing these vertical movements are onlyqualitatively understood; quantitative models with sufficient accuracyto correct observed sea-level oscillations for the tectonic componentsrarely exist. Analyses of relative sea-level observations for isostaticparameters or for the eustatic function are therefore less accurate iftectonic movements of the crust are also important. But some separationis made possible by the different characteristics of the twocontributions to sea-level change. While the tectonic displacements onthe short term may be episodic episodicsporadic; occurring in episodes. e. falling a paroxymal disorder described in Cavalier King Charles spaniels in which affected dogs, starting at an early age, experience episodes of extensor rigidity, possibly brought on by stress. e. and localized, they persist for longperiods of time, and on time-scales of 100,000-1,000,000 years, theyappear as slow but persistent movements. In contrast, changes associatedwith the growth and decay of the ice sheets are relatively uniform andglobal on short time-scales ([approximately]1000 years), but they are ofa cyclic nature on the longer time-scale. Thus, in the absence ofchanges in ocean volume, a tectonic uplift Tectonic uplift is a geological process most often caused by plate tectonics which increases elevation. The opposite of uplift is subsidence, which results in a decrease in elevation. Uplift may be orogenic or isostatic. or subsidence of 1 mm/yearmeans that shorelines formed 10,000 years ago would now be 100 m aboveor below present sea-level. An indicator of long-term tectonic stabilityis therefore provided by the present position of the Last Interglacialshoreline which formed about 120,000 years ago. Ocean and ice volumes atthat time were similar to those of today, and sea-levels were near theirpresent level. In the Mediterranean Sea Mediterranean Sea[Lat.,=in the midst of lands], the world's largest inland sea, c.965,000 sq mi (2,499,350 sq km), surrounded by Europe, Asia, and Africa.GeographyThe Mediterranean is c.2,400 mi (3,900 km) long with a maximum width of c. , the Last Interglacial shoreline- known regionally as the Tyrrhenian Shoreline - occurs in manylocalities to within a few metres above present sea-level. This isreported to be the case, for example, in the upper regions of the Gulfsof Messenia, Lakonia and Argolis of the southern Peloponnese (Kelletatet al. 1976; van Andel 1987). Although a new evaluation of the evidenceis desirable, little vertical tectonic movement Noun 1. tectonic movement - movement resulting from or causing deformation of the earth's crustcrustal movementgeology - a science that deals with the history of the earth as recorded in rocks appears to have occurredin these localities during the past 120,000 or so years. Along thesouthern shore of the Gulf of Corinth, where the Last Interglacialshoreline is found at elevations of 30-70 m (Keraudren & Sorel Sorel(sôrĕl`), city (1991 pop. 18,786), S Que., Canada, at the confluence of the St. Lawrence and Richelieu rivers. It is a grain-shipping center with an important shipbuilding industry. 1987;Collier et al. 1992), tectonic uplift at average rates of 0.25 to 0.6 mmper year appears to have occurred during the same interval. Thus, wherethis shoreline can be identified, it becomes possible to determinewhether vertical crustal tectonics has been important and - if it has -to correct for it on the not always substantiated assumption that thetectonic rates have been uniform through time for at least the last120,000 years.Predicted sea-levels in the Aegean regionGenerally, the sea-level observations for Greece and the Aegean areinadequate to estimate both the parameters describing theglacio-hydro-isostatic theory and the tectonic movements of the crust.Instead, the former part is estimated using the above outlined theoryand model parameters; then it is tested against data from areas inferredto be tectonically stable from observations of the position of the LastInterglacial shoreline. If these comparisons are satisfactory, thenpredictions are also made for the tectonically active sites to estimatethe rates of vertical uplift or subsidence (Lambeck 1995a).FIGURE 4 illustrates the predicted past sea-levels across the regionin the form of contours of equal relative sea-level position for a fewselected epochs. The contours at 18,000 b.p., for example, indicate theamount by which sea-levels at that time were lower than the present daymean sea-level. For all epochs there is a subsidence of the offshoreareas of the Aegean, Ionian and Mediterranean Seas relative to themainland, the contour values increasing with distance from the mainland.This reflects the hydro-isostatic effect with subsidence of the seafloor and uplift of the land areas [ILLUSTRATION FOR FIGURE 2 OMITTED].Superimposed upon this is a north-south trend imposed by theglacio-isostatic effect, mainly from the Fennoscandian ice lead in LateGlacial time [ILLUSTRATION FOR FIGURE 2 OMITTED]. Throughout the regionsea-level continues to rise, initially rapidly as the ice sheets aredisintegrating and then relatively slowly for the last 6000 years whenthe main contributions to the change come from the isostatic component.FIGURE 5 illustrates the changes in sea-level at four sites in theregion. Predictions for the areas of Thrace and Crete [ILLUSTRATION FORFIGURE 5A OMITTED] differ because Crete is further from the ice sheetand the hydro-isostatic effect is that of an island, whereas theprediction for Thrace is characteristic of a continental margin site. Atthe time of the Last Glacial Maximum the predictions for the two sitesdiffer by about 20 m, with the Thrace curve lying above the Creteprediction at all times. The east-west variability is illustrated by thecomparison for a site in the Argolis Gulf with a site in the Cycladeanisland group [ILLUSTRATION FOR FIGURE 5B OMITTED]. The east-west spatialvariation, largely the result of the hydro-isostatic correction, remainssignificant, about 5 in at 10,000 years b.p. At both sites thepredictions lie well below the eustatic sea-level function.If no other tectonic processes were active, at no epoch, particularlyduring the past 6000 years, are the sea-levels here predicted to haverisen above their present levels. As a whole, the region would appear toexperience a gradual subsidence at average rates ranging from about -0.6to -1.2 mm per year. This is largely consistent with the observations ofan apparent encroachment of the sea in the Argolis, Lakonia, Messenia,and Navarone Bay areas. Tectonic subsidence does not appear to have beensignificant, in agreement with the observation of the location of theTyrrhenian shoreline at only a few metres above present sea-level atseveral of these localities. The model also predicts a rising sea-levelthroughout the Cycladean Islands, consistent with the observation of thesubmerged Early Bronze Age site near Saliagos (Morrison 1968); there isno need to invoke tectonic subsidence of this area of sea floor.Where the observations point to relative uplift the estimated ratesof tectonic displacements are increased over and above the rates thatwould be inferred if the glacio-hydro-isostatic corrections are ignored.Likewise, areas of small apparent subsidence, at rates that are lessthan the isostatic rates, will also be undergoing tectonic uplift albeitat slow rates. The main area of systematic and substantial tectonicuplift forms an arc extending from Rhodes in the east to Karpathos,Crete, Kythera and the southernmost peninsulas of the Peloponnese(Lambeck 1995a). This pattern follows closely the convergent boundary In plate tectonics, a convergent boundary – also known as a convergent plate boundary or a destructive plate boundary – is an actively deforming region where two (or more) tectonic plates or fragments of lithosphere move toward one another. between the major tectonic units in the area and the maximum rate ofuplift, exceeding 4 mm per year, occurs in southwestern Crete close tothe plate margin. The other area of major tectonic uplift occurs alongthe southern shore and eastern end of the Gulf of Corinthos, withuplift-rate estimates approaching 1.5 mm/year. For central Evvoia, therates of uplift are also estimated to be about 1-1.5 mm per year.Elsewhere in the Peloponnese and Aegean no systematic patterns oftectonic uplift or subsidence appear; those movements are either smallerin amplitude or of shorter horizontal length scales than theglacio-hydro-isostatic changes. In the south of the Argolis Peninsulanear Porto Kheli, subsidence appears to have occurred at a rate of about1 mm per year for the last few thousand years (Flemming 1978) but justto the north, at the Franchthi site, little tectonic movement appears tohave occurred if the van Andel & Lianos (1983) estimate of LastGlacial Maximum sea-level of about -120 m is correct.Palaeo shorelines in the Aegean regionOnce a predictive model for sea-level exists that adequately reflectsthe observed changes, it becomes possible to predict the location ofshorelines through time using the relation (5) for areas where tectonicmovements are inferred to have been small - as appears to have been thecase for the central Aegean islands Aegean IslandsGreek islands in the Aegean Sea, particularly the Cyclades, Sporades, and Dodecanese groups. The Cyclades consist of about 30 islands. The Dodecanese, or Southern Sporades, include Kálimnos, Kárpathos, Cos, Léros, Pátmos, Rhodes, and and much of the Peloponnese. As thetopography and water depth vary much more rapidly than the sea-levelchange function, a much higher spatial description of the former isrequired for precise palaeogeographic reconstructions. Thehigh-resolution data compiled here are based on: * digitized bathymetric ba��thym��e��try?n.The measurement of the depth of bodies of water.bathy��met charts for the central Aegean sea-floor atscales of 1:100,000 and 1:150,000 (HNHS HNHS Havelock North High School (Havelock North, New Zealand)HNHS Huntington North High School (Huntington, IN)HNHS Holy Name High School (Parma Heights, OH)1993), * a 500-m resolution digital terrain model for central Greece andthe Aegean islands provided by the National Technical University inAthens, and * for an area around Franchthi on the Argolid Peninsula, digitizeddata (including bathymetry) from the 1:50,000 topographic map (Data West Research Agency definition: see GIS glossary.) A map depicting terrain relief showing ground elevation, usually through either contour lines or spot elevations. The map represents the horizontal and vertical positions of the features represented. for thearea (HMGS HMGS Hellenic Military Geographical ServiceHMGS Hazardous Material Guide Sheet 1992). Beyond the marine areas covered by the digitized1:100,000 and 1:150,000 maps, the bathymetry included in the Greekdigital terrain model is used.These three data-sets have been combined and gridded on to a 0.01[degree] grid using the Delaunay triangulation (mathematics, graphics) Delaunay triangulation - (After B. Delaunay) For a set S of points in the Euclidean plane, the unique triangulation DT(S) of S such that no point in S is inside the circumcircle of any triangle in DT(S). DT(S) is the dual of the voronoi diagram of S. and natural neighbourelement methods (Sambridge et al. 1995). Bathymetric features largerthan about 0.02 [degrees] (or about 2 km) are generally well resolved inthis gridded database.Late Palaeolithic and Neolithic shorelinesResults for these palaeoshorelines are illustrated in FIGURE 6 forthe Aegean region based on the above bathymetric data base. At the timeof the Last Glacial Maximum, as pointed out by van Andel &Shackleton (1982), much of the area of now-shallow waters would havebeen exposed at this time. For example, the central part of theSaronikos Gulf would have been a fresh- or brackish brack��ish?adj.1. Having a somewhat salty taste, especially from containing a mixture of seawater and fresh water: "You could cut the brackish winds with a knife/Here in Nantucket" water environment,and the area between Evvoia and Ellas - forming the present-day northernand southern parts of the Evvoikos Gulf and the Petalion Gulf - wouldhave consisted of a series of topographic depressions beyond marineinfluence. Also the Cycladean group of islands formed an extensive landarea, extending from Andros in the north to Ios in the south over adistance of about 160 km. To the north, beyond the area illustrated inFIGURE 6, large tracts of the Thermaikos Gulf were exposed, and theplain of Thrace extended more than 60 km to the south of the presentshoreline. The broad and shallow area between Limnos and Turkey wouldalso have been exposed, cutting off the Sea of Marmara from marineinfluence; (See [ILLUSTRATION FOR FIGURE 4 OMITTED] for the predictedsea-level change in these areas.)At this time of maximum glaciation the 'Cycladean Island'effectively divides the Aegean Sea into two: the Aegean Sea proper inthe north from the Mirtoan Sea and the Sea of Crete in the south andsouthwest. The two basins are connected via a narrow channel([approximately]6 km wide) between Evvoia and Andros in the north and byfurther channels ([approximately]12 km wide) between the islands ofAmorgos and Kalimnos. On the Cycladean Island itself, a broad centraland relatively flat plain was punctuated by hills and mountains that nowform the residual islands. The average height of the plain was about15-20 m above sea-level; topographic lows within it, between Naxos andMikonos, for example, could have held shallow freshwater environments.To the west, the islands of Milos, Sifnos, Serifos and Kithnos remainedseparate both from each other and from the mainland, although theminimum widths of the sea passages were considerably less than today.The single entity made by the islands of Milos, Kimolos and Poliaigos,for example, were separated from the Cycladean landmass land��mass?n.A large unbroken area of land.landmassNouna large continuous area of landlandmass? by relativelyshallow ([less than]100 m) channels with a minimum water-crossingdistance of about 8 km [ILLUSTRATION FOR FIGURE 7 OMITTED]. Thira(Santorini) would have been separated from the main island landmass byless than 5 km.(3) Elsewhere, the coastal plains at the time of the LastGlacial Maximum are wider than today, as in the Argolis Gulf, around theArgolis Peninsula [ILLUSTRATION FOR FIGURE 8 OMITTED] and along theeastern margin of the Aegean Sea.Initially this geography changes only slowly; even though the largeice sheets are beginning to melt, the delayed response of the Earthmeans that the actual sea-level change at these sites is relativelyslow. By 14,000 years b.p., the coastal geometry has changed only littlefrom that 4000 years earlier [ILLUSTRATION FOR FIGURE 6B OMITTED]although some of the narrow channels separating the various islandgroups have expanded marginally. After about 14,000 b.p. the sea-levelrise is more rapid. The Cycladean Island begins to break up intonorthern and southern parts that are separated by shallow sea, break-upbeing completed soon after 12,500 b.p. At this time the Milos group isseparated from the southern Cycladean island, via Sifnos, by a minimumwater crossing of about 14 km. By 12,000 b.p. the two main parts of theearlier Cycladean Island are separated by a minimum distance of about 17km. At 10,000 b.p. the original Cycladean Island has broken up further,and the geography of the region is beginning to resemble the presentconfiguration. The southern part of the Evvoikos Gulf is first subjectedto marine influence at about 9000 b.p. and the northern sector, throughthe Dhiavlos Strait, before 8000 b.p. The islands of Milos, Kimolos andPoliaigos remain together with decreasing surface area until about 8500b.p. The final separation of Naxos, Paros and Andiparos also occurs atabout this time. Likewise, the area of the coastal plains around theperiphery of the present Argolis Peninsula is gradually reducedthroughout this period [ILLUSTRATION FOR FIGURE 8 OMITTED].Shorelines from Late Neolithic time to the presentDuring Late Neolithic and Bronze Age times (after about 6000 yearsb.p.), the predicted change is a sea-level rising at a rate of about0-7-1.0 mm per year, small compared with that for the earlier periodwhen the rates were as high as 12 mm/year [ILLUSTRATION FOR FIGURE 4OMITTED]. The Late Holocene rates are comparable to the verticaltectonic rates inferred for some localities in the region, an effectiveseparation of the two contributions to relative sea-level change is nowmore problematical. For example, the local modification of shorelines bysedimentation may now become important in areas such as the upperreaches of the Argolis Gulf. Nevertheless, the characteristic regionalpattern is a slow encroachment of the sea throughout this period,consistent with observations of Bronze Age and more recent sites thatare now a few metres below sea-level in, for example, Milos (Bintliff1976), Andiparos (Morrison 1968) and near the Franchthi Palaeolithiccave site (Jacobsen & Farrand 1987). If local bathymetric maps areavailable, the shoreline evolution for this period is best estimated bysuperpositioning upon them the sea-level predictions given in FIGURE 4.Early Palaeolithic to Early Upper Palaeolithic shorelinesFIGURE 3 illustrates the predicted sea-level oscillations for Parosnear the centre of the present Cycladean island group from the time ofthe penultimate pe��nul��ti��mate?adj.1. Next to last.2. Linguistics Of or relating to the penult of a word: penultimate stress.n.The next to the last. Glacial Maximum (Saalian) at about 140,000 years ago tothe present. Spatial variability of sea-level occurring across theregion will generally be small when compared with the largeruncertainties introduced by incomplete knowledge of the ice sheetsbefore about 20,000 years b.p. As the Last Glacial Maximum appears tohave been of relatively short duration, the coastal plains were fullyexposed from only a little before 20,000 years b.p. to about 16,000years b.p. The maximum development of the Cycladean Island wouldtherefore have been restricted to between about 21,000 and 14,000 yearsb.p. The only other time in the last 150,000 years that these low levelswould have been reached is during the penultimate Glacial Maximum beforeabout 135,000 years b.p. From about 70,000 years b.p. to the lead-up tothe Last Glacial Maximum, sea-levels oscillated between about -50 and-80 m and shoreline locations would have been similar to those thatoccurred later between about 12,000 and 10,000 years b.p. [ILLUSTRATIONFOR FIGURE 6 OMITTED]. Between about 110,000 and 70,000 years b.p. thewater depths and shorelines are characteristic of conditions lastexperienced between about 9000 and 7000 years b.p.Limitations of the shoreline reconstructionsThe reconstructions of past topography and shoreline locations aremodel dependent, notably on the assumed eustatic sea-level function, onthe appropriate earth-model parameters that define the isostaticcontributions, and on the assumed absence of other tectonic andsedimentary processes. Constraints on the model parameters are providedby observations of past sea-levels, and the practice is first tocalibrate To adjust or bring into balance. Scanners, CRTs and similar peripherals may require periodic adjustment. Unlike digital devices, the electronic components within these analog devices may change from their original specification. See color calibration and tweak. the eustatic and isostatic models with data from stableregions and then to apply them to the tectonically active areas.Solutions of the sea-level equation (4) have therefore tended to beiterative: observations of sea-level change from both the area ofinterest and from areas outside of it are compared with the modelpredictions to improve the estimates of the parameters that define boththe isostatic model and the tectonic components. Predictions for theAegean point to potential tests to check the correct timing of theshoreline movements. For example, the model predicts that the SaronikosGulf was separated from the Aegean Sea and from marine influence betweenabout 23,000 and 13,000 years ago and sedimentary cores from the centralareas of the depression may contain information on the timing of thefirst marine transgression and - through the sea-level equation - on theeustatic sea-level function. Other potential target areas for testingthe model are the Petalion Gulf, or possibly the now-shallow sea floordepression between Naxos and Mikonos [ILLUSTRATION FOR FIGURE 6OMITTED].As noted above, resolution of the digitized and gridded nauticalcharts gives a spatial resolution (Data West Research Agency definition: see GIS glossary.) A measure of the accuracy or detail of a graphic display, expressed as dots per inch, pixels per line, lines per millimeter, etc. It is a measure of how fine an image is, usually expressed in dots per inch (dpi). of about 2 km, so some very finedetail is lost in the reconstructions. FIGURE 8c, for example, showsIdrha connected to the mainland at about 10,000 b.p., whereas therewould actually have been a narrow channel - less than about 500 m wide -between the two. Similar resolution is lost between those islands of theCycladean group that are separated by very narrow and deep channels.The model predictions, most reliable for the time since the LastGlacial Maximum, are also indicative of conditions throughout Early andMid Palaeolithic time. Thus the reconstruction in FIGURE 6 for theinterval 12,500-14,000 years b.p. also represents conditions betweenabout 25,000-22,000 years b.p. Some question does remain about thetime-scale appropriate to observation and prediction. All ages hererefer to the conventional radiocarbon time-scale since most observationsof sea-level change and of the retreat of ice sheets use radiocarbondeterminations. Insofar in��so��far?adv.To such an extent.Adv. 1. insofar - to the degree or extent that; "insofar as it can be ascertained, the horse lung is comparable to that of man"; "so far as it is reasonably practical he should practice as archaeological determinations forPalaeolithic and Neolithic contexts also refer to this time-scale, themodels are consistent with those data. Care is required when usingcalendar ages for the more recent times because discrepancies betweenthis and the radiocarbon time-scale can be substantial (e.g. Klein etal. 1982; Bard et al. 1990).Shorelines and the human environmentThe examples set out here indicate that reconstructions of thecontemporaneous geography should be integral to discussion of earlycoastal inhabitants of Greece. With some notable exceptions, the fullrecognition of this input has been neglected. That sea-levels during theLast Glacial Maximum were some 100-150 m lower than today is not amatter for dispute in the archaeological and prehistory literature. Noris the fact that at that time extensive coastal plains would have beenexposed, plains that would generally have seen much human activity (e.g.Gamble 1986). Less attention has been paid to the timing of thesea-level rise from then up to about 6000 years ago, and to the factthat this rise occurred over a substantial time interval. Van Andel& Shackleton (1982) and van Andel (1989), in neglected papers,attempted to quantify sea-level and shoreline changes in the easternMediterranean during this Late Glacial period and recognized that therise occurred over an extended period, with important human implicationsas the progressive reduction of the relatively hospitable coastal plainenvironment caused a concomitant loss of resources. The consequenceswould have been particularly severe for the Cycladean Island, where anextensive, relatively flat and low-lying plain was progressively reducedto a few rocky islands over a period of about 6000 years [ILLUSTRATIONFOR FIGURE 6 OMITTED]. Likewise, the coastal plains of the palaeoArgolis Peninsula [ILLUSTRATION FOR FIGURE 8 OMITTED] could have offereda hospitable environment for Palaeolithic and Neolithic dwellers.Certainly throughout the Aegean region, these now-flooded plains arelikely to have been more conducive to human activities than much of thepresent coastal zone and human traffic patterns and migration routes maywell have been quite different in the past. Can this, for example,explain the general paucity of evidence for Palaeolithic occupation ofthe region. Did the encroaching sea cover much of the evidence for theactivities of early people until about 10,000-9000 years b.p.? Even inEarly Bronze Age times, sea-levels were lower than today by as much as 5m and, as the region has been a zone of low tidal range throughout, muchof the archaeological evidence for this interval may also be belowtoday's sea-level.Locally, the coastal changes may have been significant. In Milos thecoastal plain was considerably more extensive at the time when obsidianfrom this island first appears on the mainland in latest Palaeolithictime [ILLUSTRATION FOR FIGURE 7 OMITTED]. Can this explain the apparentabsence of Late Palaeolithic and Early Neolithic settlement on theisland (Renfrew 1972) despite the quarrying that must have occurred toobtain the obsidian found on the mainland? Likewise - as alreadyemphasized by Jacobsen (1969) and van Andel (1987) - the coastalgeometry at the Late Palaeolithic site at Franchthi on the ArgolisPeninsula, would have been distinctively different in the past. At thepeak of the last glaciation, the coast would have been at least 6 km tothe west [ILLUSTRATION FOR FIGURE 9 OMITTED]; if much of the evidencefor the activities of Palaeolithic and Neolithic people is nowsubmerged, Franchthi Cave represents only a partial record of humanactivity in the area. Certainly any interpretation of the cave recordneeds to consider these shifting shorelines. Does the appearance ofmarine molluscs and small fish bones by about 12,000-10,000 b.p., andthe first much larger fish vertebrae fish vertebraeRadiology A descriptor for biconcave, fish-like vertebrae, caused by infarction and central bone collapse due to thrombosis of the vertebral arteries, a finding typical of sickle cell anemia, which often occurs before the 2nd by about 9200 b.p. reflect more onthe coastal evolution than on the fishing methods of the earlyinhabitants, as Jacobsen (1976) suggested?Obsidian from Milos appears in the Franchthi strata after about11,000 b.p. (Perles 1987; Renfrew & Aspinall 1987) (Perles (1979)and Cherry & Torrence (1982), quoting Perles, give an age of 11thmillennium b.c.). With the lower sea-levels of Late Palaeolithic time,the island would have been much more accessible from the mainland thanit became later as sea-levels rose. At the time of the maximumglaciation and until about 16,000 b.p. the minimum water crossingrequired to travel from Milos to the mainland would have been about 8 km[ILLUSTRATION FOR FIGURES 6 AND 7 OMITTED], with the largest crossingbeing between Sifnos and the Cycladean Island; this configuration wouldhave been largely preserved until about 13,000 b.p. By 11,000 b.p., thetime of the oldest obsidian finds recorded so far, the Cycladean Islandhad broken up into northern and southern parts centred onParos-Andiparos and Mikonos-Tinos-Andros although at this time theminimum water separation across shallow seas was only about 20 km, withclear inter-island visibility for a daylight crossing.Acknowledgements. I thank Dr C.L. Smither and Mrs C. Krayshek fortheir major effort in digitizing the nautical charts and for preparingthe figures. I thank Professor G. Veis and Dr Lysandros Tsoulos of theNational Technical University of Athens for making available thenon-marine topographic data.1 All ages are in conventional radiocarbon years unless otherwisespecified.2 Spelling of place-names generally follows that of the Times Atlasof the World The first version of The Times Atlas of the World appeared as The Times Atlas in 1895; more printings followed up to 1900. It was published at the office of The Times newspaper in London, and contained 117 pages of maps with an alphabetical index of 130,000 names. .3 No attempt has been made to reconstruct Thira to its pre-eruptiontopography.ReferencesBARD, E., B. HAMELIN, R.G. FAIRBANKS & A. ZINDLER, 1990.Calibration of the 14C timescale over the past 30,000 years using massspectrometric U-Th ages from Barbados corals, Nature 345: 405-10.BINTLIFF, J. 1976. The history of archaeo-geographic studies ofprehistoric Greece, and recent fieldwork, in Mycenaean geography: 3-16.Cambridge; Cambridge University Press Cambridge University Press (known colloquially as CUP) is a publisher given a Royal Charter by Henry VIII in 1534, and one of the two privileged presses (the other being Oxford University Press). & British Association forMycenaean Studies.CATHLES, L.M. 1975. The viscosity of the Earth's mantle.Princeton (NJ): Princeton University Princeton University,at Princeton, N.J.; coeducational; chartered 1746, opened 1747, rechartered 1748, called the College of New Jersey until 1896.Schools and Research Facilities Press.CHAPPELL, J. 1963. Evidence for a smoothly falling sea-level relativeto north Queensland North Queensland is the northern part of the state of Queensland in Australia. Queensland is a massive state, larger than most countries, and the Tropical northern part of it has been historically remote and underdeveloped, resulting in a distinctive regional character and , Australia, during the past 6000 years, Nature 302:406-8.CHAPPELL, J. & H. POLLACH. 1991. Post-glacial sea-level rise froma coral record at Huon Peninsula, Papua New Guinea, Nature 349: 147-9.CHAPPELL, J. & N.J. SHACKLETON. 1986. Oxygen isotopes andsea-level, Nature 324: 137-40.CHERRY, J.F. & R. TORRENCE. 1982. The earliest prehistory ofMelos, in C. Renfrew & M. Wagstaff (ed.), An island polity: thearchaeology and exploitation in Melos: 24-34. Cambridge: CambridgeUniversity Press.COLLIER, R.E.L., M.R. LEEDER, P.J. ROWE Rowe? , Nicholas 1674-1718.English writer whose works include drama, poetry, and an edition of Shakespeare. He was appointed poet laureate in 1715. & T.C. ATKINSON. 1992.Rates of tectonic uplift in the Corinth and Megara MegaraPort city (pop., 1991: urban agglom., 20,403), Greece. Situated on the Saronic Gulf and west of Athens, it served as the capital of ancient Megaris. A maritime power, by the 7th century BC it had established colonies in Sicily, Chalcedon, Byzantium, Bithynia, and Basins, CentralGreece, Tectonics 11: 1159-67.CURRAY, J.R. 1965. Late Quaternary quaternary/qua��ter��nary/ (kwah��ter-nar?e)1. fourth in order.2. containing four elements or groups.qua��ter��nar��yadj.1. Consisting of four; in fours. history, continental shelves ofthe United States United States,officially United States of America, republic (2005 est. pop. 295,734,000), 3,539,227 sq mi (9,166,598 sq km), North America. The United States is the world's third largest country in population and the fourth largest country in area. , in H.E. Wright & D.G. Frey (ed.), The Quaternaryof the United States: 723-35. Princeton (NJ): Princeton UniversityPress.DENTON, G.H. & T.J. HUGHES (ed.). 1981. The last great icesheets. New York New York, state, United StatesNew York,Middle Atlantic state of the United States. It is bordered by Vermont, Massachusetts, Connecticut, and the Atlantic Ocean (E), New Jersey and Pennsylvania (S), Lakes Erie and Ontario and the Canadian province of (NY): Wiley.FAIRBANKS, R.G. 1989. A 17,000-year glacio-eustatic sea-level record:influence of glacial melting dates on the Younger Dryas event and deepocean circulation, Nature 342: 637-42.FLEMMING, N.C. 1978. Holocene eustatic changes and coastal tectonicsin the northeast Mediterranean: implications for models of crustalconsumption, Philosophical Transactions of the Royal Society The Philosophical Transactions of the Royal Society, or Phil. Trans., is a scientific journal published by the Royal Society.Begun in 1665, it is the oldest scientific journal printed in the English-speaking world and the second oldest in the world, of LondonA289: 405-58.GAMBLE, C. 1986. The Palaeolithic settlement of Europe. Cambridge:Cambridge University Press.HMGS = HELLENIC MILITARY GEOGRAPHICAL SERVICE. 1992. 1:50,000 mapsheet Spetsai. Athens.HNHS = HELLENIC NAVY HYDROGRAPHIC hy��drog��ra��phy?n. pl. hy��drog��ra��phies1. The scientific description and analysis of the physical conditions, boundaries, flow, and related characteristics of the earth's surface waters.2. SERVICE, 1993. 1:100,000 NauticalCharts 412-414 and 150,000 Nautical Charts 415, 421, 423. Athens.HYVARINEN, H. 1980. Relative sea-level changes near Helsinki,southern Finland, during early Litorina time, Bulletin of the GeologicalSociety of Finland 52: 207-19.JACKSON, J. 1994. Active tectonics of the Aegean region, AnnualReview of Earth and Planetary Science planetary scienceor planetology,study of planets and planetary systems as a whole. Planetary science applies the theories and methods of traditional disciplines such as astronomy, geology, physics, chemistry, and mathematics to the study of 22: 239-71.JACOBSEN, T.W. 1969. Excavations at Porto Cheli Port Cheli, also Portocheli, Portoheli, Porto Heli, and rarely Porto Kheli or Portokheli (Greek, Modern: Πορτο Χέλι or Πορτοχέλι, Ancient/Katharevousa: -on and vicinity,preliminary Report, II: The Franchthi Cave, 1967-1968, Hesperia 39:343-81.1976. 17,000 years of Greek prehistory, Scientific American Scientific AmericanU.S. monthly magazine interpreting scientific developments to lay readers. It was founded in 1845 as a newspaper describing new inventions. By 1853 its circulation had reached 30,000 and it was reporting on various sciences, such as astronomy and 234(6):76-87.JACOBSEN, T.W. & W.R. FARRAN. 1987. Franchthi Cave and Paralia.Bloomington (IN): Indiana University Press Indiana University Press, also known as IU Press, is a publishing house at Indiana University that engages in academic publishing, specializing in the humanities and social sciences. It was founded in 1950. Its headquarters are located in Bloomington, Indiana. . Excavations at FranchthiCave, Greece Fascicle fascicle/fas��ci��cle/ (fas��i-k'l)1. a small bundle or cluster, especially of nerve, tendon, or muscle fibers.2. a tract, bundle, or group of nerve fibers that are more or less associated functionally. 1.KELLETAT, D., G. KOWALCZYK, B. SCHRODER & K.P. WINTER. 1976. Asynoptic syn��op��tic? also syn��op��ti��caladj.1. Of or constituting a synopsis; presenting a summary of the principal parts or a general view of the whole.2. a. Taking the same point of view.b. view on the neotectonic development of the Peloponnesiancoastal regions, Zeitschrift der Deutsche Geologische Gesellschaft27:447-65.KERAUDREN, B. & D. SOREL. 1987. The terraces of Corinth (Greece)- a detailed record of eustatic sea-level variations during the last500,000 years. Marine Geology marine geologyor geologic oceanographyScientific discipline concerned with all geologic aspects of the continental shelves and slopes and the ocean basins. Marine geology originally focused on marine sedimentation and the interpretation of bottom samples. 77: 99-107.KLEIN, J., J.C. LERMAN, P.E. DAMON & E.K. RALPH. 1982.Calibration of radiocarbon dates, Radiocarbon 24: 103-50.KRAFT, J.C., S.E. ASCHENBRENNER & G.J. RAPP. 1977.Paleogeographic reconstructions of coastal Aegean archaeological sites,Science 195: 941-47.KRAFT, J.C. & G.J. RAPP. 1975. Late Holocene paleogeography paleogeographyor palaeogeographyGeography of selected portions of the Earth's surface at specific times in the geologic past. The simplest kind of paleogeography is a map showing the locations of ancient lands and seas, but paleogeographic maps may also ofthe coastal plain of the Gulf of Messenia, Greece, and its relationshipsto archaeological settings and coastal change, Geological Society ofAmerica The Geological Society of America (or GSA) is a nonprofit organization dedicated to the advancement of the geosciences. The society was founded in New York in 1888 by James Hall, James D. Bulletin 86: 1191-1208.KRAFT, J.C., G.R. RAPP & S.E. ASCHENBRENNER. 1980. Late Holocenepalaeogeomorphic reconstructions in the area of the Bay of Navarino:Sandy Pylos, Journal of Archaeological Science Archaeological science (also known as Archaeometry) is the application of scientific techniques and methodologies to archaeology.Archaeological science can be divided into the following areas: 7: 187-210.LAMBECK, K. 1988. Geophysical geodesy geodesy(jēŏd`ĭsē)or geodetic surveying,theory and practice of determining the position of points on the earth's surface and the dimensions of areas so large that the curvature of the earth must be taken into : the slow deformations of theearth. Oxford: Oxford University Press.1993. Glacial rebound of the British Isles British Isles:see Great Britain; Ireland. II: A high resolution,high-precision model, Geophysical Journal International 115: 960-90.1995a. Late Pleistocene and Holocene sea-level change in Greece andsouthwestern Turkey: a separation of eustatic, isostatic and tectoniccontributions, Geophysical Journal International 122: 1022-44.1995b. Constraints on the Late Weichselian ice sheet over the BarentsSea from observations of raised shorelines, Quaternary Science Quaternary science is an inter-disciplinary field of study focusing on the Quaternary period, which encompasses the last 2.6 million years. The field studies the last ice age and the recent interstadial the Holocene and uses proxy evidence to reconstruct the past environments Review14: 1-16.In press. Sea-level change along the French Atlantic coast since thetime of the Lastglacial Maximum, Palaeogeography, Palaeoclimatology Noun 1. palaeoclimatology - the study of the climate of past agespaleoclimatologyarchaeology, archeology - the branch of anthropology that studies prehistoric people and their cultures ,Palaeoecology paleoecology, palaeoecologythe branch of ecology that studies the relationship of ancient plants and animals to their environments. — paleoecologic, palaeoecologic, paleoecological, palaeoecological, adj. .LAMBECK, K., P. JOHNSTON & M. NAKADA. 1990. Holocene glacialrebound and sea-level change in NW Europe, Geophysical JournalInternational 103: 451-68.LAMBECK, K., P. JOHNSTON, C. SMITHER & M. NAKADA. 1996. Glacialrebound of the British Isles III: Constraints on mantle viscosity,Geophysical Journal International 125: 340-54.LAMBECK, K. & M. NAKADA. 1990. Late Pleistocene and Holocenesea-level change along the Australian coast, Palaeogeography,Palaeoclimatology, Palaeoecology (Global and Planetary Change Section)89: 143-76.MILLIMAN, J.D. & K.O. EMERY. 1968. Sea-levels during the past35,000 years, Science 162: 1121-3.MITROVICA, J.X. & W.R. PELTIER. 1991. On postglacial geoid ge��oid?n.The hypothetical surface of the earth that coincides everywhere with mean sea level.[German, from Greek geoeid subsidence over the equatorial oceans, Journal of Geophysical Research Journal of Geophysical Research is a publication of the American Geophysical Union. JGR was formerly titled Terrestrial Magnetism from its founding by the AGU's president Louis A. 96: 20,053-71.1993. The inference of mantle viscosity from an inversion of theFennoscandian relaxation spectrum, Geophysical Journal International114: 45-62.MORRISON, I.A. 1968. Appendix I. Relative sea-level change in theSaliagos area since Neolithic times, in J.D. Evans & C. Renfrew(ed.), Excavations at Saliagos near Antiparos: 92-8. London: Thames& Hudson.NAKADA, M. & K. LAMBECK. 1987. Glacial rebound and relativesea-level variations: a new appraisal, Geophysical Journal of the RoyalAstronomical Society This article is about the British Society. For the Canadian Society, see Royal Astronomical Society of Canada. The Royal Astronomical Society (RAS) is a learned society that began as the Astronomical Society of London in 1820 to support astronomical 90: 171-224.NAKADA, M. & K. LAMBECK. 1988. The melting history of the LatePleistocene Antarctic ice sheet, Nature 333: 36-40.PELTIER, W.R. & J.T. ANDREWS. 1976. Glacial-isostatic adjustment- 1. The forward problem, Geophysical Journal of the Royal AstronomicalSociety 46: 605-46.PERLES, C. 1979. Des navigateurs mediterraneens il y a 10,000 ans, LaRecherche La Recherche is a monthly French language popular science magazine covering recent scientific news. It is published by the Soci��t�� d'��ditions scientifiques (the Scientific Publishing Group), a subsidiary of Financi��re Tallandier. 10: 82-3.1987a. Les industries lithiques tailles de Franchthi (Argolide,Grece), in Perles (ed.): 142-3.(Ed.). 1987b. Presentation generale et industries paleolithiques.Bloomington (IN): Indiana University Press. Excavations at FranchthiCave, Greece Fascicle 3(1).PIRAZZOLI, P.A., S.C. STIROS, M. ARNOLD, J. LABOREL, F.LABOREL-DEGUEN & S. PAPAGEORGIOU. 1994. Episodic uplift deduced fromHolocene shorelines in the Perachora Peninsula, Corinth area, Greece,Tectonophysics Tectonophysics, a branch of geophysics, is the study of rock deformation and of dynamic geophysical processes. This includes small individual crystals as well as the larger scales which include tectonic plates and forces in the mantle. 229: 201-9.RENFREW, C. 1972. The emergence of civilisation: the Cyclades and theAegean in the 3rd millennium BC. London: Methuen.RENFREW, C. & A. ASPINALL. 1987. Aegean obsidian and FranchthiCave, in Perles (ed): 257-70.SAMBRIDGE, M., J. BRAUN & H. MCQUEEN. 1995. Geophysicalparametrization and interpolation of irregular data using naturalneighbours, Geophysical Journal International 122: 837-57.SHACKLETON, N.J. 1987. Oxygen isotopes, ice volume and sea-level,Quaternary Science Review 6: 183-90.THOMMERET, Y., J. LABOREL, L.F. MONTAGGIONI & P.A. PIRAZZOLI.1981. Late Holocene shoreline changes and seismo-tectonic displacementsin western Crete (Greece), Zeitschrift fur Geomorphologie (neue Folge)40: 127-49.TUSHINGHAM, A.M. & W.R. PELTIER. 1991. Ice 3G: a new global modelof Late Pleistocene deglaciation based upon geophysical predictions ofpostglacial sea-level change, Journal of Geophysical Research 96:4497-523.VAN ANDEL, T.H. 1987. The adjacent sea, in T.H. van Andel & S.B.Sutton (ed.), Landscapes and people of the Franchthi region: 3-64.Bloomington (IN): Indiana University Press. Excavations at FranchthiCave, Greece Fascicle 2.VAN ANDEL, T.H. 1989. Late Quaternary sea-level changes andarchaeology, Antiquity 63: 733-46.VAN ANDEL, T.H. & N. LIANOS. 1983. Prehistoric and historicshorelines of the southern Argolid Peninsula: a subbottom profilerstudy, Nautical Archaeology 12(4): 303-24.VAN ANDEL, T.H. & J.C. SHACKLETON. 1982. Late Paleolithic andMesolithic coastlines of Greece and the Aegean, Journal of FieldArchaeology 9: 445-54.ZWARTZ, D., K. LAMBECK, M. BIRD & J. STONE. IN PREPARATION.Constraints on the former Antarctic ice sheet from sea-levelobservations and geodynamic modelling, in Proceedings of the VIIInternational Symposium on Antarctic Earth Sciences, Siena, Italy.