BasinResearch BasinmodellingoftheLimo¤nBack-arcBasin

Basinmodellingof the Limo¤ n Back-arc Basin(Costa Rica): burial historyand temperatureevolution of an islandarc-related basin-systemC. Brandes,n A. Astorga,w R. Littkez and J. Winsemannn

nInstitut fˇr Geologie, Leibniz Universit�t Hannover, Hannover, GermanywEscuela Centroamericana de Geolog|¤a, San Jose¤ , Costa RicazLehrstuhl fˇr Geologie, Geochemie undLagerst�tten des Erd˛ls und der Kohle, Aachen, Germany

ABSTRACT

The Limo¤ n back-arc basin belongs to the southern Central American arc-trench system and issituated at the east coast of Costa Rica.The basin- ¢ll consists of Late Cretaceous to Pleistocenesedimentary rocks. A northern and a southern sub-basin can be de¢ned, separated by the E^W-trendingTrans Isthmic Fault System.The North Limo¤ n Basin is nearly undeformed, whereas theSouth Limo¤ n Basin is characterized by a fold-and-thrust belt. Both sub-basins have a very similarsedimentary ¢ll and can act as a natural laboratory for reconstructing controlling factors of arc-related sedimentary basins as well as the in£uence of deformation on a basin system.Modellingfocused on burial history and temperature evolution.Two-dimensional simulations were carried outwith the software PetroMods.The geohistory curve of the North Limo¤ n Basin is overall linear,indicating continuous subsidence.The South Limo¤ n Basin is also characterized by continuoussubsidence, but rates strongly increased at the beginning of the Neogene. Despite a rapid Plio-Pleistocene deformation of the fold-and-thrust belt, the present-day temperature ¢eld is notdisturbed in that area.The modelling results indicate a mean heat £ow of 60mWm� 2 for the NorthLimo¤ n Basin and 41mWm� 2 for the South Limo¤ n Basin.These values are low comparedwith otherback-arc basins.The lower values are attributed to the following e¡ects: (1) underlying basaltic crust,(2) the lack of an initial rift phase, (3) the low extension rates, (4) absence of volcanic activity and (5)insulation e¡ects of a thick sediment pile.The reasons for the locally lower heat £ow in the southernsub-basin can be found in the low-angle subduction of theCocosRidge.Owing to the low subductionangle, the cool fore-arc mantle-wedge below the island-arc is pushed backwards increasing thecooled area.

INTRODUCTION

The thermal history of sedimentary basins is one of themost important factors for the hydrocarbon potential andalso in£uences the £uid £ow and sediment compaction ina basin as well as the physical properties of £uids and sedi-ments (e.g. Lerche et al., 1984; Gosnold & Fischer, 1986;McCulloh&Naeser,1989;Yalcin,1991). In the last 20 years,basin modelling has become a widely accepted tool for re-constructing the burial history and related temperatureevolution of sedimentary basins (e.g. Mareschal, 1987;Jensen, 1997; Rodriguez & Littke, 2001). Modern softwarecan be used for the simulation of £uid £ow and the genera-tion and expulsion of hydrocarbons within the basin (e.g.Welte & Y̌ kler, 1981; Welte & Yalcin, 1987; Ungerer et al.,1990; Zhou & Littke, 1999; Yahi et al., 2001). Basin model-

ling is extensively used in the oil industry and has beenapplied to academic problems in the last decade (e.g.Petmecky et al., 1999; N˛th et al., 2001; Lutz et al., 2004;Rodon & Littke, 2005).

In recent times, complex geological settings such asforeland basins and fold-and-thrust belts have become in-teresting exploration targets.The structural style and evo-lution of fold-and-thrust belts have been discussed bymany authors (Boyer&Elliott,1982;Davis etal.,1983;Dah-len,1984;Dahlen etal., 1984;Verge¤ s etal., 1992; Price, 2001).Numerical modelling studies were carried out on thestructural development (e.g. Burbridge & Braun, 2002)and on the temperature evolution of fold-and-thrust belts(Schneider, 2003). The e¡ect of folding and thrusting onthe thermal ¢eld and on the £uid £ow in fold-and-thrustbelts has received special emphasis (Wygrala et al., 1990;Schmidt & Erdogan, 1993; Husson & Moretti, 2002).In general, folding and thrusting can be regarded as a com-petition between advection and conduction of heat. Ifadvection is faster than conduction, the thermal ¢eld will

Correspondence: Christian Brandes, Institut fˇr Geologie,Leibniz Universit�t Hannover, Callinstr, 30, 30167 Hannover,Germany. E-mail: [email protected]

BasinResearch (2008) 20, 119–142, doi: 10.1111/j.1365-2117.2007.00345.x

r 2008 The Authors. Journal compilation r 2008 Blackwell Publishing Ltd 119

be disturbed. If the ratio between di¡usion and advectionvelocities is �1, there is no distortion (Husson&Moretti,2002). Whether the folding has an e¡ect on the thermal¢eld depends on the relationship between the wavelengthand amplitude of the folds, as well as the velocity of foldingand the thermal conductivity of the rocks (Sleep, 1979). Inthe case of large wavelengths, large amplitudes and highdeformation rates, the isotherms will be folded with thebedding planes. Nevertheless, there are still many openquestions especially with respect to a quantitative under-standing of the balance between conductive and advectiveheat transport in compressive settings.

The Limo¤ n back-arc basin is a complex basin-system.An extensional back-arc area (the North Limo¤ n Basin)and a compressional retro-arc foreland basin (the SouthLimo¤ n Basin) can be observed side by side. Both sub-ba-sins are very similar regarding the stratigraphy and lithol-ogy of their ¢ll, but di¡er signi¢cantly in the structuralstyle and basin morphology. The setting of the Limo¤ nBasin provides the opportunity to identify the controllingfactors of arc-related sedimentary basins such as subsi-dence rates and deformation.Our modelling study focuseson reconstructing burial history and temperature evolu-tion. By comparing the thermal evolution from both sub-basins, we evaluate the impact of folding on the evolutionof the thermal ¢eld.

GEOLOGICAL SETTING

The geology of southern Central America has been dis-cussed in many previous studies (Weyl, 1980; Astorga,1988; Lundberg, 1991; Seyfried et al., 1991; Winsemann &Seyfried,1991;Weinberg,1992;Winsemann,1992; Amann,1993;Krawinkel & Seyfried,1994; vonHuene&Flˇh,1994;Barboza et al., 1995; Werner et al., 1999; Krawinkel et al.,2000; Abratis & W˛rner, 2001; Gr�fe et al., 2002; Calvo,2003; Krawinkel, 2003). Most of the recent work was car-ried out in the ¢eld of marine geology/marine geophysics(Ranero & von Huene, 2000; Ranero et al., 2000a, b; Barc-khausen et al., 2003). Several publications focus on platetectonic reconstructions (e.g. Pindell et al., 1988; Ross &Scotese, 1988; Frisch et al., 1992; Astorga, 1997; Maresch etal., 2000; Meschede et al., 2000). Other studies deal withthe petroleum potential of Costa Rica (Sheehan et al.,1990; Astorga et al., 1991; Barboza et al., 1997; Barrientos etal., 1997; Petzet, 1998; Lutz, 2002; Lutz et al., 2004).

Central America is characterized by the interaction of¢ve lithospheric plates, including the oceanic Cocos, Naz-ca and Caribbean Plates and the continental North and

South American Plates (Fig. 1a). The Cocos and NazcaPlates are subducting beneath the Caribbean Plate alongthe NW^SE-trending Central America trench. The pre-sent-day subduction velocity o¡ Costa Rica, relative tothe Caribbean Plate, is 8.5 cmyear�1 (DeMets, 2001).TheCocosPlate is characterized by the largeNE^SW-trendingCocos Ridge, an asesimic ridge, which is a more than1000-km-long, 250^500-km-broad structure that rises2000m above the adjacent sea£oor. It has been subductedbeneath southern Costa Rica since at least 3.6Ma (Collinset al., 1995;Walther, 2003).

TheCentral American land-bridge is a complex assem-blage of four crustal blocks. From SE-Guatemala to NW-Colombia the Maya, Chortis, Chorotega and ChocoBlocks can be distinguished (Donnelly, 1989; Weinberg,1992; Di Marco et al., 1995; Campos, 2001) (Fig. 1a). TheMaya and Chortis Blocks have a continental basement,whereas the Chorotega and Choco Blocks represent is-land-arc segments underlain by a Mesozoic oceanic crust(Escalante & Astorga, 1994). The Costa Rican part of theisland-arc (Chorotega Block) can be subdivided into anorthern and a southern arc segment (Seyfried et al.,1991).The northern arc segment is bounded to the northby theHess Escarpment and to the south by theTrans Isth-mic Fault System (Fig. 1a). Both features are importanttectonic elements. The Hess Escarpment is a NE^SW-trending bathymetric feature in the Caribbean Sea,separating the continental Chortis Block from the oceanicColombia Basin (Krawinkel & Seyfried, 1994; Campos,2001). In previous studies, theHess Escarpmentwas inter-preted to represent a lateMesozoic plate boundary, whichacted as a strike^slip zone to compensate the movementsbetween Chortis Block, Chorotega Block and the Carib-bean Plate (Krawinkel, 2003).TheTrans Isthmic Fault Sys-tem is an E^W-trending active lineament. It mainly showssinistral movements and links a transtensional grabenstructure of southeastern Nicoya with the Limo¤ n ThrustBelt (Krawinkel & Seyfried, 1994). The southern CostaRican arc segment is located south of theTrans IsthmicFault System and belongs to the Panama Microplate. Animportant structural element of the southern arc seg-ment is the North Panama Deformed Belt, dominatingnorthern Panama. The development of this fold-and-thrust belt has largely been controlled by the colli-sion of Panamawith SouthAmerica sinceMiocene times,the forward movements of the Nazca Plate and the orocl-inal bending of the arc (Kolarsky et al., 1995; Silveret al., 1995).

In general, the evolution of most of the arc-related sedi-mentary basins started in Cretaceous times in response to

Fig.1. (a)Tectonic map ofCosta Rica.TheLimo¤ nBasin extends along theCaribbean coast.The basin is separated into a northern and asouthern sub-basin by the E^W-trendingTrans Isthmic Fault System (modi¢ed after Ross &Scotese,1988;Donnelly,1989; Barboza etal.,1995;Meschede & Frisch, 1998; Campos, 2001).The North Limo¤ n Basin is an extensional back-arc basin; the South Limo¤ n Basin is acompressional retro-arc foreland basin. Both sub-basins are very similar regarding the stratigraphy and lithology of their ¢ll, but di¡ersigni¢cantly in the structural style and basin morphology. (b) SW^NE-trending section of the o¡shoreNorthLimo¤ nBasin.This sectionshows characteristics of passive margins.The large listric normal faults give evidence for a Plio-Pleistocene extensional phase. (c) SW^NE-trending cross-section from the o¡shore South Limo¤ n Basin, showing the structure of the Limo¤ n fold-and-thrust belt.

r 2008 The Authors. Journal compilation r 2008 Blackwell Publishing Ltd,Basin Research, 20, 119^142120

C. Brandeset al.

1s

2s

3s

4s

3s

2s

16 km

1s

19 km

PacificOcean

CaribbeanSea

11°00´

9°00´

10°00´

Moín High

11°00´

200 m

83°00´

10°00´

San José

Nicaragua

Back-arc

Intra-arc

NicaraguaGraben

Fore-arc

Hess Escarpment

San CarlosBasin

North LimónBasin

South LimónBasin

SarapiguíHigh

Isthmic

100 km

North Costa Rican arc segment

South Costa Rican arc segment

Cocos Ridge84°00´ 83°00´

Panama

Outer-arc

Fore-arc

olcanic Arc

Costa Rica

TransFault System

(a)

(b)

(c)

North AmericanPlate

SouthAmerican

PlateNazca Plate

Cocos Plate

CaribbeanPlate

r 2008 The Authors. Journal compilation r 2008 Blackwell Publishing Ltd,Basin Research, 20, 119^142 121

Basinmodelling of the Limo¤ n Back-arc Basin

the onset of subduction of the Farallo¤ n Plate beneath theCaribbean Plate. The evolving tholeiitic island-arc shedlarge amounts of volcaniclastic material into the adjacentsedimentary basins (Seyfried et al., 1991).

The Limo¤ n back-arc basin belongs to the southernCentral American arc-trench system and is situated be-neath the present-day coastal plain and continental shelfof eastern Costa Rica (Fig. 1a). The northern boundary istheHess Escarpment.To the west and to the south, the ba-sin is fronted by the volcanic arc. The eastward extent isde¢ned by the 200m bathymetric contour line of the Car-ibbeanSea and by the extent of theLimo¤ n fold-and-thrustbelt, respectively. The ¢ll of the Limo¤ n Basin consists ofLate Cretaceous to Pleistocene deep-marine to continen-tal volcaniclastic rocks (Sheehan et al., 1990; Coates et al.,1992, 2003; Amann, 1993; Bottazzi et al., 1994; Fernandezet al., 1994; McNeill et al., 2000; Campos, 2001; Mende,2001) (Fig. 2). Deposition of carbonates occurred duringthe Late Cretaceous, Eocene and Oligocene.

The North Limo¤ n Basin belongs to the North CostaRican arc segment and is undeformed. Outcrops are rareand most information comes from two-dimensional (2D)seismic re£ection lines (Fig.1b) and some wells. Aeromag-netic data imply that the central basin is more than 7000mdeep (Barboza et al., 1997). To the NW, the Sarapiqu|¤High separates the North Limo¤ n Basin from the SanCarlos intra-arc basin (Barrientos et al., 1997). Someauthors, however, integrate the San Carlos Basin into theNorth Limo¤ n Basin (Sheehan etal., 1990).The basin is stillsubsiding (Mende, 2001).

The South Limo¤ n Basin belongs to the South CostaRican arc segment and contains approximately 6^8 km ofsedimentary rocks. The basin- ¢ll has been deformed byNE-directed folding and thrusting (Fig.1c).The resultingfold-and-thrust belt is referred to as the Limo¤ n fold-and-thrust belt in this study. All stratigraphic information forthe South Limo¤ n Basin is based on onshore outcrops andwell data (Amann, 1993; Bottazzi et al., 1994; Fernandezet al., 1994).The oldest sediments are �1280-m-thick pe-lagic limestones and intercalated volcaniclastic rocks ofLate Campanian to Maastrichtian age (Changuinola For-mation, Fig. 2). The overlying Tu|¤ s Formation consists of�3000m of Palaeocene to Lower Eocene coarse-grainedvolcaniclastic turbidites, debris- £ow deposits, lava- £owsand tu¡s, representing a prograding deep-water apron-system (Mende, 2001). An early compressional phasea¡ected the island-arc during Eocene to Oligocene timesand led to the formation of signi¢cant topographic relief,as implied by the simultaneous deposition of 150^200-m-thick shallow-water limestones of the Las AnimasFormation on local structural highs (Amann,1993;Mende,2001), and 700^900-m-thick hemipelagic mudstones,calcareous turbidites and carbonate debris- £ow depositsof the Senosri Formation in adjacent basinal areas(Mende, 2001). On top of these carbonates �2000-m-thick shallow-water volcaniclastic sediments of the UpperOligocene to Upper Miocene Uscari Formation weredeposited (Amann, 1993; Mende, 2001). The Uscari

Formation is overlain by the shallow-water limestonesand volcaniclastic rocks of the 400^1800-m-thick R|¤ oBanano Formation (Amann, 1993; Bottazzi et al., 1994;Mende, 2001).The subduction of theCocosRidge beneaththe island-arc began during the Late Miocene to EarlyPliocene (Collins et al., 1995; Abratis &W˛rner, 2001).Thisled to increased NE-directed folding and thrusting andthe development of small intramontane piggy-backbasins. Subsequently, shallow-marine and continental

200

400

600

800

1000

1200

1400

1600

1800

2000

2200

2400

2600

2800

3000

3200

3400

3600

3800

4000

4200

4400

4600

4800

5000

5200

5400

5600

5800

6000

6200

6400

6600

6800

7000

Key

Shale

Siltstone

Sandstone

Conglomerate

Limestone

Basalt

3.03.8

4.2

5.5

6.3

8.2

9.2

10.5

15.5

16.5

21.0

25.5

30.0

36.0

39.5

42.5

49.5

68.0

75.0

Fig. 2. Stratigraphy of the South Limo¤ n Basin (modi¢ed afterCampos, 2001).


C. Brandeset al.

rocks of the Plio-Pleistocene Suretka Formation were de-posited (Amann, 1993; Bottazzi et al., 1994;Mende, 2001).

DATABASE ANDMETHODS

The database includes a grid of 2D seismic re£ection linesaquired during onshore and o¡shore campaigns in the1970s and 1980s by the national oil company of Costa Rica(RECOPE) (Fig. 3). The grid consists of NE^SW- andNW^SE-directed seismic lines. Our interpretation of theseismic sections was performedwith the software packageKingdomSuiter. Seismic sections are one of the buildingblocks of the conceptual model for the basin simulation.Therefore, the interpretation focused on the reconstruc-tion of the basin- ¢ll architecture. Stratigraphic and litho-logic control for the interpretation is derived from oneonshore well for each sub-basin.These wells were drilledby RECOPE in the late1980s as stratigraphic tests in eachsub-basin (Well #1 and #2).Well #1 is the key well for

the interpretation of the seismic section from the NorthLimo¤ n Basin. It penetrates Pleistocene to Eocene sand-stones, shales and limestones. Well #2 is the key wellfor the interpretation of the seismic sections of theSouth Limo¤ n Basin. It penetrates Pleistocene to Miocenesandstones, shales and limestones. The Cretaceous toPalaeocene portions of the basin- ¢ll that were not drilledwere interpreted on the basis of published outcrop data(Astorga et al., 1991; Campos, 2001; Mende, 2001). Depthconversion was performed for the two modelled sectionsand for the sections, which were used for the reconstruc-tion of balanced cross-sections. The depth conversionwas performed on the basis of interval velocities. Basinmodelling focused on burial history and temperature evo-lution, carried out with the software PetroMods 2D forone cross-section of each sub-basin.The modelled cross-sections are indicated by the bold lines in Fig. 3.We chosethese lines because they cover a large part of each sub-ba-sin and the nearby well provides the input and calibrationdata for the simulations. In addition, a 1D simulation was

PuertoLimón

Río Matina

Río Vizecaya

Río Pacuare

Caribbean Sea

Costa Rica

10°

83°

10°

Well #2

Well #1

25 km

Moín High

12

3

4

5

7

8

10

9

12

13

1415

16

17

11

18

19

6

Fig. 3. Location map of the seismic lines and the two key wells.The key wells (Well#1and#2) provide the calibration data for thesimulation.The bold seismic lines indicate the foundation for the basin modelling.



carried out for each sub-basin based on well data to checkthe results derived from the 2D simulations. All numericalmodels were calibrated with vitrinite re£ectance data andbottom-hole temperatures. These data are unpublishedand were provided by the Ministry of Environment andEnergy (MINAE).Additional input data like the variationsofwaterdepth over timewere also taken from these unpub-lished reports. For the structurally complex South Limo¤ nBasin, balanced cross-sections have been reconstructed toestimate the amount of horizontal shortening.

SEISMIC INTERPRETATION ANDBASIN-FILL ARCHITECTURE OFTHE LIMOŁ N BASIN

The North Limo¤ n Basin is characterized by continuousdepositional units, which can be laterally traced for severaltens of kilometres on the NW^SE-directed seismic lines.The NE^SW-directed lines, which run perpendicular tothe slope, display a wedge-shaped geometry of the accu-mulated sediments (Brandes et al., 2007a). The sedimen-tary units display a pronounced seaward dip with thehighest sediment thickness close to the present-day coast-line (Fig. 1b). This depositional style is characteristic ofpassive margin settings in the circum Atlantic area (e.g.Hutchinson et al., 1982; Bond et al., 1995). Large listric nor-mal faults can be observed in the shelf area (Fig.1b).Thesefaults trend NW^SE and belong to small graben struc-tures. The major normal faults of the graben structuressole into a common detachment (Brandes et al., 2007b).Borehole data indicate that the position of the detachmentis probably controlled by the lithological change fromshale of the Uscari Formation to limestone of the SenosriFormation (Fig.1b).

An antiformal structure, referred to as Mo|¤ n High, se-parates the North and South Limo¤ n Basin and was inter-preted to represent a basement high in previous studies(Sheehan et al., 1990; Barboza et al., 1997).The Mo|¤ n Highhas a convex, mound-like shape (Fig. 4). The internalstructure is di⁄cult to image with seismic methods. Astrong re£ector envelopes the Mo|¤ n High and separates itfrom the surrounding sedimentary rocks. Below the re-£ector the Mo|¤ n High is weakly layered (Brandes et al., inpress). On the northern £ank of the structure a more con-tinuous re£ector pattern can be observed.TheMo|¤ n Highis draped with Middle Miocene and younger deposits.Upper Eocene to Lower Miocene units display an onlappattern or pinch out against the Mo|¤ n High.Well #1wasdrilled on the northern £ank of the structure and termi-nates in Middle Eocene carbonates. These rocks are di-rectly overlain by Early Miocene shales. Upper Eoceneand Oligocene deposits are absent at the drill site. Thewedging of re£ector packages against the Mo|¤ n High im-plies that these units are present in the deeper parts ofthe basin.Therefore, the unconformity is interpreted as alocal feature, limited to the £ank of the structure. Fromthis data set, theMo|¤ n High is interpreted as an anticline,

which evolved betweenMiddle Eocene and EarlyMiocenetimes (Brandes et al., in press).

The South Limo¤ n Basin is characterized by the Limo¤ nfold-and-thrust belt (Fig.1c).The seismic grid is denser intheSouthLimo¤ nBasin and allows a detailed assessment ofthe structural style of the deformed belt.The best insightinto the architecture of the fold-and-thrust belt is pro-vided by the NE^SW-directed seismic lines. These linesrun roughly parallel to the direction of maximum com-pression.The o¡shore part of the Limo¤ n fold-and-thrustbelt is characterized by concentric or asymmetric hang-ingwall anticlines and large southwestward-dipping listricor planar thrusts (Brandes et al., in press).The o¡set at thethrust faults ranges between 50m and1km.All thrusts soleinto a subhorizontal detachment. Below the detachmentno signi¢cant deformation can be observed.The detach-ment lies within the sedimentary rocks, resulting in athin-skinned fold-and-thrust belt. Lithological logs de-rived from Well #2 indicate that the position of the de-tachment occurred at the change from shale of the UscariFormation to limestone of the Senosri Formation(Brandes et al., in press).Therefore, the rheologic contrastfrom shale to limestone probably controlled the positionof the detachment. Remarkable features associated withthe fold-belt are small and lenticular piggy-back basins.These basins contain some of the youngest sediments ofthe South Limo¤ n Basin.

Deformation phases of the Limo¤ n Basin

The Central American island-arc was a¡ected by threemajor deformation phases since the Cretaceous (Seyfriedet al., 1991; Krawinkel et al., 2000). Based on our seismicdata, two di¡erent deformation phases could be recon-structed in the Limo¤ n Basin since the Palaeocene. TheMo|¤ n High is a key element in the structural analysis. On-lap geometries in the vicinity of the structure imply that a¢rst compressional phase a¡ected that area between theMiddle Eocene and the Early Miocene (Brandes et al.,2007a). This phase caused the initial uplift of the Mo|¤ nHigh. Upper Eocene and Oligocene deposits are absenton the northern £ank of the Mo|¤ n High. Early to MiddleMiocene strata show a distinct onlap against the structureor pinch out.A small graben on the top of theMo|¤ nHigh isinterpreted as a crestal collapse structure that is related tothe formation of the anticline. Based on the sedimentthickness distribution across theMo|¤ n High, it can be de-duced that a second but weaker phase of uplift occurredbetween the late Middle Miocene and the Pliocene(Brandes et al., 2007a). Since then, the Mo|¤ n High wasdraped with sediments. No further vertical movementsare documented.North of theMo|¤ nHigh theLimo¤ nBasinshows large listric NW^SE-trending normal faults, whichcan be attributed to a Pleistocene extensional phase(Brandes et al., 2007a).

For the South Limo¤ n Basin, thickness variations acrossthe thrust faults provide evidence of the timing of defor-mation-improving estimates of the temporal evolution of


C. Brandeset al.

the fold-and-thrust belt. The o¡shore part of the SouthLimo¤ n Basin was a¡ected by compression since the Plio-cene. However, most of the deformation occurred duringthe Pleistocene (Brandes et al., in press).

BASIN MODELLING, CONCEPTUALMODEL ANDMODELLING RESULTS

The 2D thermal simulations are based on one seismic linefrom each sub-basin. Lithologic information and calibra-tion parameters are derived from key wells (Well #1 and#2), located on or close to each of the seismic lines. Addi-tional data from unpublished exploration reports were in-tegrated.The locations of the modelled sections are shownin Fig. 3.Table 1 shows the physical properties of the rocksthat were used in the simulation.

Basinmodelling

The basin modelling part of this project was carried outwith the software PetroMods 1D/2D, which was devel-oped by IES GmbH, Germany. PetroMods allows us tostudy the burial history and temperature evolution of a se-dimentary basin and the related generation and migrationof hydrocarbons. It is a dynamic forward simulation, basedon the ¢nite element method.The ¢nite element grid usedfor the 2D simulations has rectangular cells de¢ned byver-tical grid lines and horizontal event lines.The horizontalevent lines are de¢ned by time lines derived from the seis-mic interpretation.The vertical grid lines are placed equi-distantly along the section to generate a mesh. During thesimulation, the ¢nite element grid is ¢xed to the objectsand subsides with the layers (Friberg, 2001). A basin mod-elling study with PetroMods follows a standard work£ow.

16 km

Moín High

Pleistocene

Pliocene

Miocene

Paleocene-Eocene

Basement

UpperCretaceous

10 km

NW SE

0 m

2000 m

4000 m

6000 m

8000 m

20 km 30 kmWell #1

40 km

Lithologies

Pseudo well at grid point 71 (Fig. 7 and Fig. 13a)

Fig.4. Conceptual model for the North Limo¤ n Basin.The section runs parallel to the coastline and is characterized by continuousdepositional units with a constant thickness.The close-up shows the seismic data of the convexMo|¤ n High.The graben on top of theMo|¤ n High is interpreted as a crestal collapse structure that evolved during the rise of the anticline.



At ¢rst a conceptual model must be created,which is basedon the results of basin analysis. The conceptual model isthe description of the geologic evolution of a basin (Poel-chau etal.,1997). For such a conceptual model, the complexreal systemmust be simpli¢ed and reduced to a few funda-mental parameters (Welte & Y̌ kler, 1981).The conceptualmodel is designed on the basis of the basin- ¢ll architec-ture.Then the depositional history of the basin is dividedinto distinct events. Each event can represent a phase ofdeposition, non-deposition or erosion. It is necessary toreconstruct the complete evolution of a sedimentary ba-sin. Therefore episodes of non-sedimentation or evenerosion must be quanti¢ed (Poelchau etal., 1997). Absoluteages and, in a second step, lithologies must be assigned toeach event. The absolute age is important to reconstructthe temporal evolution.The lithologydetermines the ther-mal conductivity of the basin- ¢ll and the behaviour dur-ing compaction.

The next step is to de¢ne a set of boundary conditionsthat describe the state of the model at its boundaries.Theboundary condition can be constant or variable during thesimulation, depending on the requirements of the study.The most important boundary conditions are the basalheat £ow and the temperature at the earth surface (Yalcin,1991; Yalcin et al., 1997).The basal heat £ow is chosen ac-cording to the geological setting.The surface temperatureis determined by the geographic position of the basin andby the water depth. These parameters are not stable andchange during the evolution of the basin (Poelchau et al.,1997). Therefore, these changes must also be consideredin the simulation. PetroMods has a tool to reconstructthe surface temperature based on palaeolatitude and pa-laeo-water depth. Plate drift and related climatic changesare incorporated into the simulation.After the simulation,the resulting output must be compared with the real sys-tem in order to calibrate the model. The most importantcalibration parameters are vitrinite re£ectance values and

present-day temperature data. Both parameters are de-rived fromwells close to the modelled section.Vitrinite re-£ectance is an indicator of the maturity of organic matter.It can be used for the determination of the palaeoheat £uxand to optimize the thermal history models (e.g. Lercheet al., 1984; Durand et al., 1986).With increasing maturityof the organic material, the re£ectivity of the vitrinites alsoincreases. The maturation of vitrinites can be describedwith reaction kinetics (Larter, 1989; Sweeney & Burnham,1990). In this study, we used the widely accepted approachof Sweeney & Burnham (1990), which is based on the Ar-rhenius equation. It describes the maturation of vitriniteswith increasing temperature, considering the loss ofwater,carbon dioxide and hydrocarbons.The rate of this reactionis a function of the activation energy and the frequencyfactor.The advantage of the kinetic approach of Sweeney&Burnham (1990) is itswide range. It can be used to calcu-late vitrinite re£ectance values from 0.3 to 4.5%. Duringbasin simulation, the program calculates the evolution oftemperature and maturity parameters parallel to the evo-lution of the basin. After each run the simulated valuescan be compared with real data. If the simulated para-meters ¢t to the measured ones, the simulation can bestopped.The calibrated model can now be used to predictfurther parameters, which were not directly measured inthe real system. If the simulated and measured parametersdiverge, the input must be adjusted until a sensible cali-bration is achieved.

In this study, two standard types of basin modellingtechniques were applied. 1D basin modelling is based onwell data. 2D basin modelling is performed for one cross-section of each sub-basin.

1D basin modelling

The 1D basin simulations were used to reconstruct theburial history and temperature evolution for a single point

Table1. Physical properties of the rocks that were used in the simulation

Lithology Density (kg m� 3) Porosity (%)

Compressibility(10� 7 kPa)

Thermalconductivity(WmK�1)

Heat capacity(kJ kgK�1)

Max Min 20 1C 100 1C 20 1C 100 1C

Shale 2680 65 60 000 10 1.98 1.91 0.213 0.258Silty shale 2677 62 25 000 10 2.05 1.94 0.21 0.254Shale/siltstone 2674 59 13 000 10 2.09 1.97 0.207 0.251Shale/limestone 2695 53 1500 20 2.39 2.24 0.208 0.246Siltstone 2672 56 8000 10 2.14 2.03 0.201 0.242Siltstone/shale 2674 59 13 000 10 2.09 1.97 0.207 0.251Tu⁄tc siltstone 2666 56 9000 10 2.27 2.13 0.201 0.242Sandstone 2660 42 500 10 3.12 2.64 0.178 0.209Conglomeratic sandstone 2663 35 330 10 2.93 2.63 0.184 0.217Sandstone/siltstone 2665 50 1900 10 2.59 2.31 0.192 0.229Limestone 2710 24 150 10 2.83 2.56 0.195 0.223Shaly limestone 2700 37 550 10 2.51 2.31 0.203 0.237Marl 2687 47 940 10 2.23 2.11 0.208 0.248Basement 2750 5 2 1 2.72 2.35 0.188 0.223


C. Brandeset al.

in the basin. Input data for 1Dmodelling are derived fromthe key wells. As described above, the most importantparameters are the age and thickness of the sediments aswell as their lithologies. Boundary conditions for the si-mulation are the temperature at the base and at the surfaceof the model.

2D basin modelling

The 2D simulations of this study are based on seismic sec-tions.The basin- ¢ll architecture is integrated into the si-mulation. Before basin modelling the seismic lines wereinterpreted. Based on the key wells, which were tied tothe seismic lines, the whole section was subdivided intoas many layers as possible.The geometry and arrangementof the depositional units were transferred into a concep-tual model.The lithology for each layer was derived fromthe key wells or published outcrop data. PetroMods has aset of standard lithologies. Each lithology is characterizedby several physical parameters (Table 1). Based on the de-scription of the rocks, we chose the appropriate lithologyin the modelling software. Again, the boundary conditionsfor the simulation are the temperature at the base andat the top of the model. Because of the integration of thebasin- ¢ll architecture, the 2D simulation allows us toreconstruct the basin evolution in a more realistic way.For example the lateral variations of the thermal ¢eld canbe derived.

There is a di¡erence in the input parameters betweenthe 1D and 2D simulations in this study. Owing to wellcontrol in the upper parts of the stratigraphic sequence,the 1D models were split into more events.Therefore, thetemporal and lithological resolution is higher. It was notpossible to have similar detailed 2Dmodels, because manyof the lithologic intervals were below the resolution of theseismic sections. For the 2D models several intervals hadto be merged.To avoid mistakes due to the simpli¢cations,we used the 1D simulations to check the results of the 2Dsimulations. The heat £ow derived from the 2D simula-tions is 3^4mWm� 2 higher than the values derived fromthe 1D simulations. This di¡erence in the results is re-garded as insigni¢cant.

The PetroMods version used in this study is not able tosimulate horizontal shortening. Therefore, folding of thestrata was modelled as di¡erential burial.The beds on thelimbs of the anticlines were buried faster than the materialon the crest of the anticlines.This approach also generatesa realistic geometry of the basin- ¢ll for each time step andallows a reasonable simulation of the thermal ¢eld.

North Limo¤ n Basin

Conceptual model

The cross-section for the 2D simulation is based on aNW^SE-trending onshore seismic line.The line is 40 kmlong and runs parallel to the coastline north of Puerto Li-mo¤ n (Fig. 3).We chose this section because it is one of thelongest seismic sections from the North Limo¤ n Basin and

awell (Well#1) is located on the seismic line that provideslithologic information for the conceptual model and thecalibration parameters for the simulation.The section cov-ers the southern part of the North Limo¤ n Basin and thetop and the northwestern £ank of the adjacent Mo|¤ n High(Fig. 4). In the central part of the section the basin reachesa thickness of approximately 8000m.Well #1was drilledonshore north of Puerto Limo¤ n on the northern £ank oftheMo|¤ n High close to the present-day coastline. It pene-trates Pleistocene to Middle Eocene sandstone, shale andlimestone. The terminal depth is 3356m. At a depth of2910m, Middle Eocene limestone is directly overlain byEarly Miocene limestone. Upper Eocene and Oligocenedeposits are missing.This unconformity is interpreted asa local feature, which resulted from vertical movementsof theMo|¤ n High.The amount of vertical uplift is di⁄cultto quantify, because it is not clear whether the UpperEocene and Oligocene sediments were eroded or whetherthey were never deposited on theMo|¤ nHigh.The concep-tual model for the 2D simulation consists of 21 layers(Table 2).The layers are based onwell data and the seismicinterpretation. The ¢nite element mesh has 132 verticalgrid lines.

The initial phase of evolution of the North Limo¤ nBasin started in Campanian time, with the depositionof ca. 300m of conglomeratic sandstones, followedby ca. 1000m of pelagic carbonates (Campos, 2001). Theseismic interpretation shows that these sediments weredeposited with a relatively constant thickness along themodelled section.The PalaeoceneÊarly Eocene period ischaracterized by conglomeratic sandstones that representa deep-water slope-apron system (Mende, 2001). Thewater depth decreases from 2000m (Campanian) to 150m(Early Eocene) (MINAE, unpublished report). Duringthe Early Eocene the evolution of the Mo|¤ n High started.From the seismic line it can be determined that EarlyEocene and Middle Eocene sediments were deposited onthe Mo|¤ n High with reduced thickness, compared withthe deeper parts of the basin. Close to the northwesternmargin of the Mo|¤ n High, a small depocentre evolved,where sediments accumulated with slightly greater thick-nesses than further to the northwest. Late Eocene to EarlyMiddle Miocene units show an onlap against the Mo|¤ nHigh.The period EoceneÔligocene is characterized by arelatively shallow water depth of 200m (MINAE, unpub-lished report). During theMiocene there was still tectonicactivity at the Mo|¤ n area. Middle to Late Miocene rocksstill have a reduced thickness on theMo|¤ n High comparedwith the deeper parts of the basin. During this period thewater depth increased from 200m (Late Oligocene) to1000m (Late Middle Miocene) (MINAE, unpublishedreport). Whereas the Late OligoceneÊarly Miocene wascharacterized by carbonate deposition, the subsequentMiddleMioceneÊarly Pleistocene periodwas dominatedby shale deposition (Mende, 2001). The seismic sectiongives evidence for an increased sediment thickness of thePliocene units in the crestal collapse graben on top of theMo|¤ n High.We interpret this as an indicator for ongoing



tectonic activity.The Pleistocene shows very high sedimen-tation rates (ca. 1100m sandstone, siltstone and limestonewere deposited in 1.8Ma). From the seismic line, it can bedetermined that the Pleistocene to Recent deposits displaya constant thickness distribution across the section.Thereis no evidence for di¡erential tectonic activity in the Mo|¤ narea. The water depth decreased continuously from 800m(LateMiocene) to 0m (Holocene) (Campos, 2001).

Calibration

The simulation was calibrated with vitrinite re£ectancedata and bottom-hole temperatures (Fig. 5). The mea-sured vitrinite trend has an anomalous shape and shows akink at 1500m. Below the kink the trend is relatively steep.The calibration was performed using the Easy%Ro ap-proach of Sweeney & Burnham (1990). We carried out aminimum/maximum scenario with a heat £ow range be-tween 30 and 85mWm� 2. The simulated vitrinite trendwas adjusted to the vitrinite re£ectance measurementsfrom the lower part of the sequence (2200^2800m). A best¢t of measured and simulated vitrinite re£ectance trendswas achieved with a mean heat £ow of 60mWm� 2. Froma 1D simulation a mean heat £ow of 56mWm� 2 was de-rived. This di¡erence in the results between the 1D and2D simulation is not signi¢cant. A calibration of the vitri-nite re£ectance values in the upper part of the sequence(500^2000m) is also possible if a heat £ow of 85mWm� 2

is used.However, the data in the lower part of the sequencedo not ¢t in this case. Furthermore, the recent bottom-

hole temperatures are not met. Therefore, it is assumedthat the ¢rst calibration (60mWm� 2) is more realistic. Atleast two possibilities exist to explain the high vitrinite re-£ectance data for the sequence of 500^2000m: the ¢rstpossibility is convective heating by hot £uids. In our opi-nion, this can be ruled out due to the low permeability ofthe shales.The secondpossibility is the presence of resedi-mentedvitrinite derived from erosion of older rocks, as hasbeen described (e.g. Radke et al., 1997; Taylor et al., 1998).Clearly, geothermal gradients are never linear due to thedi¡erent conductivities of the rocks and are also in£u-enced by compaction (Yalcin et al., 1997).

Table 2. Stratigraphy and input parameters of the conceptual model of the North Limo¤ n Basin at grid point 31

North Limo¤ n Basin

EventNo. Event name

Age at thebase inMa Lithology

Thickness in m(at GP 31)

Waterdepth (m)

SWIT( 1C)

21 Holocene 0.01 Siltstone 105 0 23.120 Late Late Pleistocene 0.45 Sandstone and Siltstone 154 100 21.119 Middle Late Pleistocene 0.66 Limestone 92 50 22.118 Early Late Pleistocene 1.3 Sandstone and Siltstone 448 200 20.117 Early Pleistocene 1.8 Shale 363 500 8.816 Late Pliocene 2.9 Shale 108 300 16.315 Early Pliocene 5.3 silty Shale 228 800 514 LateMiocene 7.4 Shale 327 800 513 LateMiddleMiocene 12.5 Shale 511 1000 512 MiddleMiddleMiocene 18.0 Shale 551 200 21.111 EarlyMiddleMiocene 20.4 Siltstone and Shale 417 700 6.010 EarlyMiocene 23.0 Limestone 127 500 10.49 Late Oligocene 28.4 Limestone 386 200 21.78 Early Oligocene 33.9 Sandstone 387 200 22.47 Late Eocene 37.2 Sandstone 187 200 236 Middle Eocene 41.4 tu⁄tic Siltstone 261 200 23.75 Early Eocene 48.6 tu⁄tic Siltstone 258 150 24.54 Paleocene 65.5 congl. Sandstone 1253 900 16.53 Campanian-Maastrichtian 79.9 Limestone 866 2000 14.92 Early Campanian 83.5 congl. Sandstone 215 2000 15.21 Basement 90 Basement 4500 2000 15.2

SWIT (sediment water interface temperature) is the upper boundary condition for the simulation.

Vitrinite Reflectance in %VRr

Dep

th in

m

1000

00 0.40 0.60 0.80

2000

3000

4000

5000

0

Temperature in °C

Dep

th in

m

1000

2000

3000

4000

5000

(a) (b)

Fig. 5. Calibration for the North Limo¤ n Basin. (a) Vitrinitere£ectance measurements from keyWell#1 (black dots).Thecalibration was performedwith the Easy%Ro approach ofSweeney & Burnham (1990).The simulated trend is given by theblack line. (b) Bottom-hole temperatures from the keywell (blackdots).The simulated geothermal gradient is shown as a solid line.


C. Brandeset al.

65.5

0M

a

7.43

Ma

(a)

0 M

a

20.4

3M

a(b

)

(c)

(d)

Dis

tanc

ein

m

Dis

tanc

ein

mD

ista

nce

in m

Dis

tanc

ein

m

Fig.6.

Burialhistory

oftheN

orthLim

o¤ nBasin.T

hetemperatureevolutionisshow

nwithacolour

overlay.The

initialphaseofbasinevolutionstartedinCretaceou

stim

es(a).Figu

res(b)and(c)show

Neogene

timesteps.The

present-daytemperature¢eldisshow

nin(d).



Modelling results

The most deeply buried rocks on the modelled section oc-cur in the depocentre close to the northwestern margin oftheMo|¤ nHigh.They reach a temperature of �225^250 1C(Fig. 6). From a pseudo well at grid point 71, the tempera-ture and maturity of the Campanian and Early Miocenelayers were inferred. At that point the basin has a depth of7800m.The Campanian rocks have a recent temperatureof �160 1C and a maturity of �1.20%VRr (Fig. 7a). Forthe EarlyMiocene rocks a recent temperature of �100 1Cand a maturity of �0.60%VRr were derived (Fig. 7b).Thesimulation reveals that the evolution of theMo|¤ nHigh hasno signi¢cant e¡ect on the thermal ¢eld.The present-daytemperature ¢eld is characterized byalmost surface-paral-lel isotherms. Only a minor upwarp of the isotherms canbe observed in the area of the Mo|¤ n High.This indicatesthat the uplift of the Mo|¤ n High was too slow for the tem-perature ¢eld to be disturbed.

South Limo¤ n Basin

Reconstruction of balanced cross-sections

The construction of balanced cross-sections is a geo-metric method used to quantify the amount of shorteningor extension of a deformed basin- ¢ll and to reconstructthe initial geometry (e.g. Dahlstrom, 1969; DePaor, 1988).In this study, the technique was applied to the SouthLimo¤ n Basin (Limo¤ n fold-and-thrust belt) prior to basinmodelling. The basic concept behind the constructionof balanced sections is ‘compatibility’. That is all body

translations, rotations and strains developed in a de-formed mass of rock are controlled by geometric laws. Inaddition, the rock mass is conserved during deformation(Ramsey & Huber, 1987). Several rules and methods havebeen proposed for the construction of balanced cross-sec-tions. According to Wilkerson & Dicken (2001), a cross-section is a balanced cross-section if the following fourcriteria are met: (1) interpreted geometries must re£ectnatural stuctures for the area of interest; (2) the pre-defor-mational strata must appear reasonable once folds are un-folded and slip is removed on faults; (3) mass must beconserved between the deformed cross-section and the re-stored cross-section; and (4) a physically possible kine-matic history must describe the development of thestructures from the undeformed to deformed state. Averycommon redeformation technique is line-length balan-cing (e.g. Dennison & Woodward, 1963; Ramsey & Huber,1987). Line-length balancing was performed for threedepth-converted seismic sections from the South Limo¤ nBasin. On each section the pre-deformation length of sev-eral interpreted seismic re£ectors at di¡erent depth levelswas reconstructed. For the section that is used for basinmodelling the results point to a relatively low shorteningof 8% (Fig. 8). Overall, the thrusts have a small displace-ment. We interpret that the horizontal shortening wasmainly compensated by the development of the anticlinesthrough fault-propagation folding. These types of foldshave been discussed in great detail by e.g. Suppe (1983),Suppe & Medwede¡ (1990) and Mitra (1990). The resultsfrom the other sections imply that there are signi¢cantvariations in the amount of shortening in the Limo¤ n

(b)

(a)

Maturity evolution of the Campanian layer (solid line) and Early Miocene layer (dashed line)

Temperature evolution of the Campanian layer (solid line) and Early Miocene layer (dashed line)T

emp

erat

ure

in °

CE

asy

%R

o

Age in Ma

Age in Ma

North Limón Basin (Pseudo well at grid point 71)

Fig.7. Temperature and maturity evolution for the interval Campanian and EarlyMiocene in the North Limo¤ n Basin.


C. Brandeset al.

fold-and-thrust belt. In general, the shortening increasesfrom the SE towards the NW.This observation can prob-ably be explained by the interaction of the fold-and-thrustbelt with theMo|¤ n High.TheNW^SE-orientated seismicsection shows that the southeastern £ank of theMo|¤ nHighis overthrust by the propagating deformed belt.The o¡setsalong the thrust faults increase towards the Mo|¤ n High.The Mo|¤ n High acted as an obstacle for the fold-and-thrust belt. The propagation was limited and the move-ments were compensated with o¡sets along the thrusts(Brandes et al., 2007b).

Conceptual model

The section for the simulation of the South Limo¤ n Basinwas created from two NE^SW-trending seismic lines(Fig. 3). A shorter onshore line and a longer o¡shore linewere combined. An interpolation between the two lineswas possible, based on seismic interpretation and theprevious work of Bottazzi et al. (1994) and Fernandezet al. (1994). From SW towards NE, the section shows

three pronounced anticline structures, which developedin the hangingwall of the main thrust faults (Fig. 9).Thethrusts have more or less listric geometries and a very lowdisplacement. Because of the low displacement thethrusts are not integrated into the conceptual model.From the structural restoration, it can be derived thatmost of the deformation along the section has beencompensated by folding. In the northeastern-most partof the section, the deposits reach a thickness of approxi-mately 8000m (Fig. 9). We choose this section for thebasin modelling, because it is the longest seismic linefrom the South Limo¤ n Basin that runs parallel to thedirection of maximum compression. In addition, Well#2, a key well in the South Limo¤ n Basin is very close tothis section. The well penetrates Pleistocene to LowerMiocene sandstones, shales and limestones. The term-inal depth is 4202m.The well data are used for the con-ceptual model and for calibrating the simulation. Theconceptual model (Fig. 9) consists of 17 layers (Table 3).These layers were de¢ned on the basis of the seismicinterpretation and data derived fromWell 1. Altogether,

Pin line

Pin line

Line drawing of the depth converted seismic line

Restored section

Pin line

1.3 km shortening

1 km

2 km

3 km

4 km

5 km

1 km

2 km

3 km

4 km

5 km

16 km

1 km

1 km

Fig. 8. Line-length balancing of a depth-converted seismic section from the South Limo¤ n Basin. For this section, whichwas also usedfor basin modelling, a shortening of1.3 km (8%) is derived.The thrusts have a small displacement. Horizontal shortening was mainlycompensated by the development of the anticlines throughout fault-propagation folding.



148 vertical grid lines were used to de¢ne the ¢niteelement mesh.

The evolution of the South Limo¤ n Basin started 83Maago in the Campanian with the deposition of ca. 300m ofconglomeratic sandstone and �1300m of pelagic lime-stone (Mende, 2001). During the Palaeocene, �1400m ofsandstone and conglomeratic sandstone were deposited(Campos, 2001). The seismic interpretation shows thatCretaceous and Palaeocene sediments have a constantthickness along the section. The Oligocene unit is stillsandy, whereas the Early^MiddleMiocene period is domi-nated by carbonate deposits (marl, limestone and shalylimestone) (Campos, 2001). The water depth decreasesfrom 2000m (Campanian) (Campos, 2001) to 200m (EarlyMiocene) (MINAE, unpublished report). The MiddleMiocene^Pliocene period is characterized by shaly andsilty deposits. From the seismic line it can be seen that

Early^MiddleMiocene deposits have a constant thicknessacross the section. Deformation and evolution of the Li-mo¤ n fold-and-thrust belt started in the Pliocene (Brandesetal., in press).During theLatePliocene the ¢rst anticlinesevolved. Most of the deformation occurred during thePleistocene. Plio-Pleistocene sediments consist of sand-stone, siltstone and shale and reach their greatest thick-ness of �2000m in the piggy-back basins. On the crestof the anticlines these deposits show a reduced thickness.

Calibration

The vitrinite trend from the South Limo¤ n Basin is quitesteep and has a nearly linear shape (Fig. 10). The calibra-tionwas performedwith the Easy%Ro approach (Sweeney& Burnham, 1990). To reconstruct the heat £ow, a mini-mum/maximum scenario with a range of 25^65mWm� 2

. 5

Middle Miocene

Lower Miocene

Basement

Cretaceous to Eocene

16 km

3s

NE

Quaternary

Pliocene

Upper Miocene

1s

2s

SW

Oligocene4s

5s

Basement

Pleistocene

Oligocene

Pliocene

Miocene

Paleocene-Eocene

UpperCretaceous

10 km 20 km

0 m

2000 m

4000 m

6000 m

8000 m

Lithologies

Well #2

Pseudo well at grid point 5 (Fig. 12 and Fig. 13b)SW NE

Fig.9. Conceptual model for the South Limo¤ n Basin.The well shown on the section isWell#2.MiddleMiocene to Pleistocene rocksare part of the thin-skinned Limo¤ n fold-and-thrust belt.The enlargement of the seismic line shows the o¡shore part of the fold-and-thrust belt.The thrusts are not integrated into the conceptual model because of the low displacement.


C. Brandeset al.

was carried out. Similar to the North Limo¤ n Basin, the si-mulated vitrinite trendwas adjusted to the vitrinite re£ec-tance measurements from the lower part of the sequence(2900^4000m). A best ¢t between measured and simu-lated trends was achieved with a mean heat £ow of41mWm� 2. A mean heat £ow of 38mWm� 2 was derivedfrom a 1D simulation.The di¡erence in the results is notsigni¢cant. For the South Limo¤ n Basin itwas also possibleto calibrate with the vitrinite re£ectance values of theupper part of the sequence (1000^2700m) when if a heat£ow of 58mWm� 2 is applied. However, the data in thelower part of the sequence do not ¢t in this case.The ¢rstcalibration (41mWm� 2) is therefore more realistic. Againdi¡erent possibilities exist to explain the high vitrinite re-£ectance data for the upper part of the sequence.The ¢rst

one is convective heating by circulating hot £uids. Thiscan be ruled out due to the shaly nature of the sequence.The second possibility is the presence of resedimented vi-trinite derived from erosion of older rocks as described bye.g. Radke et al. (1997), and Taylor et al. (1998). Clearly,geothermal gradients are also in£uenced by the conduc-tivity of the rocks and are therefore not linear. From the si-mulations an average geothermal gradient for the uppercrust of 2 1C100m�1 can be deduced, which is equal tothe present-day geothermal gradient that was measuredin the key well (Astorga et al., 1991).

Modelling results

The most deeply buried rocks in the section reach a tem-perature of �180^200 1C (Fig. 11). From a pseudo well atgrid point 5, the temperature and maturity of the Campa-nian and Early Miocene layers were inferred (Fig. 12). Atthat point the basin has a depth of 8000m.TheCampanianrocks show a recent temperature of �170 1C and a matur-ity of �1.30%VRr. For the Early Miocene rocks a recenttemperature of �80 1C and a maturity of �0.50%VRrcan be inferred.The simulation suggests that the evolutionof the Limo¤ n fold-and-thrust belt has no signi¢cant e¡ecton the thermal ¢eld.There is no distortion due to the de-formation.The present-day temperature ¢eld is character-ized by surface-parallel isotherms (Fig.11).

Comparison of the burial history of the North and SouthLimo¤ n Basin

The software PetroMods 2D allows the extraction of geo-history curves at any location of the 2D section. This 1Dextraction was performed for each sub-basin (Fig. 13a

Table3. Stratigraphy and input parameters of the conceptual model of the South Limo¤ n Basin at grid point 103

South Limo¤ n Basin

EventNo. Event name

Age at thebase (Ma) Lithology

Thickness inm at (GP103)

Waterdepth (m)

SWIT( 1C)

17 Late Late Pleistocene 0.4 Sandstone 297 130 21.316 Eraly Late Pleistocene 1.2 Sandstone 493 130 21.315 Early Pleistocene 1.8 Siltstone 310 200 21.314 Pliocene 5.3 Shale and Siltsone 692 200 19.913 LateMiocene 7.4 Shale and Siltsone 223 350 15.212 LateMiddleMiocene 11.8 Shale 480 500 8.911 MiddleMiddleMiocene 16.1 silty Shale 1128 400 11.710 EarlyMiddleMiocene 20.4 Shale and Limestone 182 200 20.89 Late EarlyMiocene 20.8 shaly Limestone 270 200 20.28 Middle EarlyMiocene 21.5 Shale and Limestone 251 300 19.37 Early EarlyMiocene 22.3 Marl 136 300 18.46 Earliest EarlyMiocene 23.0 Marl 378 300 17.55 Oligocene 33.9 Sandstone 788 800 9.34 Paleocene-Eocene 65.5 congl. Sandstone 1426 900 14.53 Campanian-Maastrichtian 79.9 Limestone 1345 2000 14.92 Campanian 83.5 congl. Sandstone 290 2000 14.91 Basement 90 Basement 4500 2000 15

SWIT (sediment water interface temperature) is the upper boundary condition for the simulation.

Temperature in °C

0

Dep

th in

m

1000

2000

3000

4000

5000

0

Dep

th in

m

1000

2000

3000

4000

5000

Vitrinite Reflectance in %VRr

0 0.40 0.60 0.80

(a) (b)

Fig.10. Calibration for the South Limo¤ n Basin. (a) Vitrinitere£ectance measurements from keyWell#2 (black dosts).The calibration was performedwith the Easy%Ro approach ofSweeney & Burnham (1990).The simulated trend is given by theblack line. (b) The solid line shows the geothermal gradientderived from the simulation.The dashed line shows the observedgradient (2 1C100m�1) inWell#2.



(a)

(d)

(b)

(c)

65.5

0M

a20

.43

Ma

0Ma

1.81

Ma

0-20

20-4

040

-60

60-8

0

Tem

per

atu

re in

°C

80-1

0010

0-12

012

0-14

014

0-16

016

0-18

018

0-20

0

200-

220

220-

240

240-

260

260-

280

280-

300

300-

320

320-

340

340-

360

360-

380

380-

400

1000

020

000

1000

020

000

1000

020

000

1000

020

000

Dis

tanc

e in

mD

ista

nce

in m

Dis

tanc

e in

mD

ista

nce

in m

0 0

Depth in m

0

0

0

1300

0

1200

0

1100

0

1000

0

9000

8000

7000

6000

5000

4000

3000

2000

1000

Depth in m

0

1300

0

1200

0

1100

0

1000

0

9000

8000

7000

6000

5000

4000

3000

2000

1000

Depth in m

0

1300

0

1200

0

1100

0

1000

0

9000

8000

7000

6000

5000

4000

3000

2000

1000

Depth in m

0

1300

0

1200

0

1100

0

1000

0

9000

8000

7000

6000

5000

4000

3000

2000

1000

0-20

20-4

040

-60

60-8

0

Tem

per

atu

re in

°C

80-1

0010

0-12

012

0-14

014

0-16

016

0-18

018

0-20

0

200-

220

220-

240

240-

260

260-

280

280-

300

300-

320

320-

340

340-

360

360-

380

380-

400

0-20

20-4

040

-60

60-8

0

Tem

per

atu

re in

°C

80-1

0010

0-12

012

0-14

014

0-16

016

0-18

018

0-20

0

200-

220

220-

240

240-

260

260-

280

280-

300

300-

320

320-

340

340-

360

360-

380

380-

400

0-20

20-4

040

-60

60-8

0

Tem

per

atu

re in

°C

80-1

0010

0-12

012

0-14

014

0-16

016

0-18

018

0-20

0

200-

220

220-

240

240-

260

260-

280

280-

300

300-

320

320-

340

340-

360

360-

380

380-

400

Fig.11.Burialhistory

oftheSouth

Lim

o¤ nBasin.T

hetemperatureevolutionisshow

nwiththecolour

overlay.Basin

evolutionstartedinlateCretaceou

stim

es(a).Su

bsidence

continuedasshow

nin

¢gures(b)and

(c).The

present-daytemperature¢eldisno

tdisturbed

bythefoldingofthestrata(d).


C. Brandeset al.

and b). The North Limo¤ n Basin is characterized by asubsidence rate of �68mm.y.�1 during the Palaeocene(Fig. 13a). At the beginning of the Neogene, there is aminor phase of uplift. During the Middle Miocene thesubsidence rate is �300mm.y.�1 Between 6 and 4Maanother short period of uplift occurred. The last 4m.y.(Late Pliocene^Pleistocene) show a subsidence of�200mm.y.�1 In the South Limo¤ n Basin the period66^34Ma is characterized by a subsidence rate of�56mMa�1. At the beginning of the Neogene there is adistinct change in the trend. The subsidence abruptlyincreased to �675mm.y.�1 for a period of approximately2m.y. The following 9m.y. are characterized by a rate of�150mm.y.�1 The period 12^2Ma shows a subsidenceof 55mm.y.�1; between 2 and 0.5Ma there is an increaseto 280mm.y.�1During the last 0.5m.y. there was uplift.

LIMITATIONS OF THEMODELS

The results of the basin modelling must be regarded as aninitial work because of the limited data. The vitrinite re-£ectance measurements were taken from an unpublishedreport.There is no information provided about the uncer-tainties of the measurements. In addition, the calibrationwith one vitrinite pro¢le and a few measurements of thepresent-day geothermal gradient for each sub-basin ledto relatively unconstrained thermal models. Regionalpredictions are not possible. Nevertheless, the ¢t of thesimulated recent temperature ¢eld and the measuredbottom-hole temperatures show that the simulations arereasonable at least for the modelled sections.

DISCUSSION

The geological setting of the Limo¤ n Basin with the twodi¡erent sub-basins allows us to study the controlling fac-tors of the basin-system in great detail. Although there aredistinct di¡erences in structural style, the burial historycurves from the North and South Limo¤ n Basins show alinear subsidence trend in pre-Neogene times (Fig. 13aand b).This is probably typical for back-arc basins, whichevolved behind island-arcs.There is no evidence for an in-itial rift-phase during the early stage of basin evolution.The subsidence rate is �68 to �56mMa�1.This is con-sistentwithMende’s (2001) conclusion that the Limo¤ n Ba-sin started as a non-extensional back-arc basin behind thevolcanic arc. Mende (2001) observed bimodal volcanismwith dyke intrusions and the formation of pillow lavas thathe interpreted to be an indicator of a Palaeocene to Eocenerifting event. Such a distinct rift-phase cannot be con-¢rmed with the seismic data or by our basin modelling.The geohistory curve for the onshore South Limo¤ n Basinshows a very similar evolution, with an initial linear subsi-dence trend in pre-Neogene times (Fig.13b). In contrast tothe northern sub-basin, there is a pronounced increase insubsidence (675mMa�1) at the beginning of theNeogene.The shape of the burial history curve is consistent with aforeland basin setting (Angevine et al., 1990).The increasein subsidence probably indicates the point when the back-arc basin was transformed into a retro-arc foreland basin.The rapid subsidence is interpreted to have resulted fromcrustal loading due to the evolution of the Cordillera deTalamanca. The increase at 2Ma is probably related tothe evolution of the Limo¤ n fold-and-thrust belt. The

(a)

South Limón Basin (Pseudo well at grid point 5)

(b) Maturity evolution of the Campanian layer (solid line) and Early Miocene layer (dashed line)

Temperature evolution of the Campanian layer (solid line) and Early Miocene layer (dashed line)

Tem

per

atu

re in

°C

Eas

y %

Ro

Age in Ma

Age in Ma

Fig.12. Temperature and maturity evolution for the Campanian and EarlyMiocene intervals in the South Limo¤ n Basin.



sediments deposited in the South Limo¤ n Basin weresuccessively folded and integrated into the propagatingdeformed belt. The di¡erent Neogene geohistory curvesobserved in the North and South Limo¤ n Basin can clearlybe attributed to the di¡erent structural settings. TheNorth Limo¤ n Basin is a back-arc basin.The South Limo¤ nBasin also evolved as a back-arc basin, but in contrast tothe northern sub-basin was later compressed and trans-formed into a fold-and-thrust belt.

The results of the thermal modelling indicate that theNorth Limo¤ n Basin is characterized by a mean heat £owof 60mWm� 2 and the South Limo¤ n Basin by a mean heat£ow of 41mWm� 2. The heat £ow in both sub-basins isunusually low comparedwith other back-arc settings alongthe eastern Paci¢c margin (Hyndman et al., 2005).The average heat £ow in active back-arc basins is65^120mWm� 2 (Allen & Allen, 1992). The geothermalgradient observed in wells in the Limo¤ n area is also

moderate with �3 1C100m�1. Similar geothermal gradi-ents are observed in the Nicaraguan back-arc (MisquitoBasin) (Darce et al., 2000) (Fig. 14). The low heat £ow inthe North Limo¤ n Basin can be explained by the lack ofa rift-phase. As mentioned above, there is no evidencefor important crustal stretching. Another possible expla-nation for the low heat £ow is given by the nature ofthe underlying lithosphere. Data of Darce et al. (2000)show a decreasing geothermal gradient from 3.5 to2.5 1C100m�1 from north to south in the Misquito Basin(Fig. 14). Munoz et al. (1997) concluded that the basementin the Misquito Basin changes from north to south froma continental over transitional to an oceanic crust. In con-trast to the continental crust, the oceanic crust has littleradiogenic heat production. The crustal variations ob-served by Munoz et al. (1997) may explain the decreasein the geothermal gradient in that area.The basins furtherto the south in Costa Rica and Panama are exclusively

(b) South Limón Basin

(a) North Limón Basin

30

30

40

40

50

50

60

60

70

70

Age in Ma

Age in Ma

80

80

10

10

0

0

20

20

2000

2000

Bur

ial d

epth

in m

Bur

ial d

epth

in m

4000

4000

6000

6000

8000

8000

Fig.13. Comparison of the geohistory curves from the (a) North Limo¤ n Basin and (b) South Limo¤ n Basin (see text for detailedexplanation).The geohistory curve from theNorthLimo¤ n Basinwas extracted at grid point 71, and the geohistory curve from the SouthLimo¤ n Basin was extracted at grid point 5.


C. Brandeset al.

underlain by basaltic crust. Radiogenic heat productionis absent.

To explain the low heat £ow in the Limo¤ n Basin, blan-keting e¡ects due to high sedimentation rates must alsobe taken into account. High sedimentation rates can dras-tically in£uence the surface heat £ow. In the North andSouth Limo¤ n Basin, the Pleistocene sediments reach athickness of nearly1km. Husson &Moretti (2002) pointedout that active sedimentation can reduce the heat £ow by afactor of 2 in the uppermost kilometres. In the Gulf ofLyon sedimentation rates of 620mm.y.�1 absorb 32% ofthe surface heat £ow (Lucazeau & Le Douaran, 1985). Inthe North Limo¤ n Basin the sedimentation rates in theEarly Pleistocene locally exceed �1500mm.y.�1 (Fig.15).In the Late Pleistocene the sedimentation rates are in arange of �300^1200mm.y.�1 For the South Limo¤ n Ba-sin, the piggy-back basins show high sedimentation ratesof up to �1650mm.y.�1 (Fig.15).

The heat £ow on the modelled section from the SouthLimo¤ nBasin betweenPuertoLimo¤ n andCahuita is signif-icantly lower than the heat £ow in theNorth Limo¤ n Basin.This lower heat £ow might be a function of the low-anglesubduction of theCocosPlate in the area of southernCostaRica. As mentioned before the low subduction angle iscaused by the bouyant Cocos Ridge, which rests on theCocos Plate. The Central American subduction zone ischaracterized by strong along-trench variations in the dip

angle of the Wadati-Benio¡ zone. Protti et al. (1995) ob-served an angle of 841 underNicaragua, 601 under CentralCosta Rica and a £at slabwith noWadati-Benio¡ zone un-der South Costa Rica. Following Campos (2001), the sub-duction angle in southern Costa Rica has shallowed sinceMiddle Oligocene times. As suggested in the model ofGutscher (2002), a low-angle subduction pushes the man-tle-wedge backwards, resulting in an increase of the coolfore-arc area.The signi¢cant gap in the volcanic chain ofthe Cordillera deTalamanca has been interpreted to resultfrom the subduction of the Cocos Ridge, which stoppedmagma production from the mantle-wedge at 8Ma (Abra-tis & W˛rner, 2001). De Boer et al. (1995) proposed a de-crease in the subduction angle of the Cocos Plate sinceMiocene times, in combination with thickening of theupper plate.

Further reasons for the locally low heat £ow can prob-ably be found in cooling meteoric waters. The prevailingwind in that area transports water-saturated air from theCaribbean Sea towards the southeast. The Cordillera deTalamanca evolved in Miocene times (Campos, 2001).Since several million years ago, the elevated mountainrange has acted as a barrier that induces orographic rainand probably acts as a potential recharge area for ground-water. The topographically low North Limo¤ n Basin isprobably a discharge area.The strong e¡ect of groundwater£ow on the geothermal gradient and heat £ow is well

Managua

San José

Geothermal gradient: 3°C/100 mHeat-flow: 60 mW/m

Geothermal gradient: 2°C/100 mHeat-flow: 41 mW/m

Geothermal gradient: 3.5°C/100 m

Geothermal gradient: 3°C/100 m

Geothermal gradient: 2.5°C/100 m

Cocos Plate

Caribbean PlateMisquito Basin

Limón BasinCosta Rica

Nicaragua

Honduras

Panama

Panama City

82°

82°

84°

84°

86°

86°

8°8°

10°10°

12°12°

14°14°

80°

80°

78°

78°

Fig.14. Geothermal gradient and heat £ow in southern Central America.The map is a compilation of data fromAstorga et al. (1991),Darce et al. (2000), and simulation results from this study.The geothermal gradient decreases from north to south in theMisquito Basin(Nicaragua), probably related to a change in crustal structure.TheCostaRicanback-arc is characterized by amoderate geothermal gradientof 3 1C100m�1 in theNorth Limo¤ n area and a lower value of 2 1C100m�1 in the northeastern part of the Limo¤ n fold-and-thrust belt.



100000

0

0

2000

2000

3000

3000

4000

4000

5000

5000

6000

6000

7000

7000

8000

8000

9000

9000

10000

10000

11000

11000

12000

12000

13000

101

101

20

20

30

30

40

40

50

50

60

60

70

70

80

80

90

90

100

100

110

110

120

120

130

130 140

1000

1000

Dep

th in

mD

epth

in m

Grid points

Grid points

20000 30000 40000

South Limón Basin

North Limón Basin

0-150150-300300-450450-600

Sedimentation rate in m/Myr

Sedimentation rate in m/Myr

600-750750-900900-10501050-12001200-13501350-1500

1500-16501650-18001800-19501950-21002100-22502250-24002400-25502550-27002700-28502850-3000

0-150150-300300-450450-600600-750750-900900-10501050-12001200-13501350-1500

1500-16501650-18001800-19501950-21002100-22502250-24002400-25502550-27002700-28502850-3000

100000Distance in m

Distance in m

20000

Fig.15. Sedimentation rates for theNorth andSouthLimo¤ n Basin. In theNorthLimo¤ n Basin, sedimentation rates of �1500mm.y.�1

occur in the Pleistocene. In the South Limo¤ n Basin, the sedimentation rate in the Pleistocene locally reaches �1600mm.y.�1 In theNorthLimo¤ n Basin, the high Pleistocene sedimentation rates show a quite uniform distribution along the modelled section, whereas inthe South Limo¤ n Basin they are restricted to the piggy-back basins.


C. Brandeset al.

known from the RockyMountain Foothills in Alberta andthe Great Plains in the United States (Majorowicz et al.,1984, 1985; Gosnold, 1985). Gayer et al. (1998) and Nunnet al. (2005) have shown the strong impact of a topographi-cally driven downward £ow of groundwater on the thermal¢eld of a sedimentary basin.The recharge areas are charac-terized by a low temperature and the discharge areas showincreased values.The calibration well in the South Limo¤ nBasin is located relatively close to the possible rechargearea and might be in£uenced by cooling e¡ects. Despitethe shaly nature of the Neogene part of the basin ¢ll, thee¡ect of groundwater £ow must be taken into account.

Crustal thickening as a consequence of the evolution oftheLimo¤ n fold-and-thrust belt can be ruled out to explainthe low heat £ow.The displacement along the thrust faultsis very low and has not led to crustal thickening.

Most of the deformation in the o¡shore part of theLimo¤ n fold-and-thrust belt took place in the Pleistocene(Brandes et al., in press). Despite this young and rapiddeformation, the present-day temperature ¢eld is notdisturbed.The isotherms are still more or less parallel tothe sediment surface. It is assumed that the deformationrate and the wavelength of the anticline structures in theLimo¤ n Basin are too low to have an e¡ect. In the Limo¤ nBelt, the conductive heat transfer was fast enough to com-pensate the e¡ect of folding.Husson&Moretti (2002) haveshown that even thrusting is very unlikely to disturb thethermal ¢eld in fold-and-thrust belts.

CONCLUSIONS

The Limo¤ n back-arc basin is an excellent setting to studythe evolution of the thermal ¢eld and the highly variableheat £ow in arc-related sedimentary basins. The burialhistory curves of theNorth andSouthLimo¤ nBasin recordthe di¡erent tectonic settings. The pre-Neogene burialhistory curves from the North and South Limo¤ n Basinshow a linear subsidence trend. No initial rift-phase canbe observed. In the South Limo¤ n Basin, the subsidencestrongly increased at the beginning of the Neogene. Theshape of the burial history curve is similar to foreland ba-sin settings. The abrupt increase in subsidence is inter-preted to indicate the point when the back-arc basin wastransformed into a retro-arc foreland basin.

Based on the thermal modelling results, it is assumedthat the heat £ow distribution in arc-related basins is veryheterogenous and strongly in£uenced by a combination oflocal parameters.The thermal modelling indicates a meanheat £ow of 60mWm� 2 for the North Limo¤ n Basin and41mWm� 2 for the South Limo¤ n Basin. The heat £ow islow comparedwith other back-arc basins in the circumPa-ci¢c region.The lower values are attributed to the follow-ing e¡ects: (1) underlying basaltic crust, (2) the lack of aninitial rift phase, (3) the low extension rates, (4) absence ofvolcanic activity and (5) insulation e¡ects of a thick sedi-ment pile. Additional reasons for the locally lower heat£ow in the southern sub-basin can be found in the sub-duction geometry. The subduction angle drastically de-

creases in southern Costa Rica because of the buoyantCocos Ridge. Owing to the low subduction angle, the coolfore-arc mantle-wedge below the island-arc is pushedbackwards, increasing the cooled area. Furthermore, cool-ing meteoric waters, which probably in¢ltrate in the Cor-dillera de Talamanca close to the calibration well of theSouth Limo¤ n Basin, may also reduce the heat £ow.

Although there is a rapid deformation in Plio-Pleisto-cene times, the present-day temperature ¢eld is not dis-turbed by the folding.The heat transfer was fast enough tocompensate for the e¡ects of folding.The deformation rateand thewavelength of the anticlinal structures in theLimo¤ nBasin are too low to in£uence the temperature distribution.

ACKNOWLEDGEMENTS

Wewould like to thank the Costa RicanMinistry of Envir-onment and Energy (MINAE) for providing the data base.We are indebted especially to Alvaro Aguilar and GustavoSegura for logistic help. Financial support from the Ger-man Research Foundation (DFG) Project Wi 1844/6-1and a graduate scholarship from the University of Hann-over are gratefully acknowledged. Many thanks are due toStefan Back for help with seismic interpretation. We arevery grateful to Peter Blisniuk for introduction of the tech-niques of balanced cross-sections and toMichaelHaschkefor the discussion about the thermal structure of subduc-tion zones and magmatic arcs. Ulrich Asprion, Franz Bi-not, Lolita Campos, Christoph Gaedicke, Stefan Ladage,Andreas Mende, Klaus Reicherter and Danny Schwarzerhelped with seismic interpretation and basin modelling.Imke Stru� is gratefully acknowledged for constructivediscussions and helpwith seismic interpretation and basinmodelling. The whole manuscript greatly bene¢ted fromcareful reviews by Michelle Kominz, Claire Currie andRobert Harris.

REFERENCES

Abratis,M.&W˛rner,G. (2001) Ridge collision, slab-windowformation, and the £ux of Paci¢c asthenosphere into the Car-ibbean realm.Geology, 29, 127^130.

Allen, P.A. & Allen, J.R. (1992) Basin Analysis. Blackwell,Oxford, 451pp.

Amann, H. (1993) Randmarine und terrestrische Ablagerungs-r�ume des neogenen Inselbogensystems in Costa Rica (Mitte-lamerika) Pro¢l, 4, 161pp.

Angevine,C.L.,Heller, P.L. & Paola,C. (1990) Quantitativesedimentary basin modeling. AAPG Course Note Ser., 32,247pp.

Astorga, A. (1988) Geodina¤ mica de las cuencas del Creta¤ cicoSuperior ^ Paleo¤ geno de la regio¤ n ‘‘forearc’’del Sur de Nicara-gua y Norte de Costa Rica.Rev. Geol. Am. Central, 9, 1^40.

Astorga,A. (1997) El puente-istmo deAme¤ ricaCentral y la evo-lucio¤ n de la Placa caribe (con e¤ nfasis en el Mesozoico). Pro¢l,12, 201pp.

Astorga, A., Fernandez, J.A., Barboza, G., Campos, L.,Obando, J., Aguilar, A. & Obando, L.G. (1991) Cuencassedimentarias de Costa Rica: evolucion geodinamica y poten-cial de hidrocarburos.Rev. Geol. Am. Central, 13, 25^59.



Barboza, G., Barrientos, J. & Astorga, A. (1995) Tectonicevolution and sequence stratigraphy of the central Paci¢c mar-gin of Costa Rica.Rev. Geol. Am. Central, 18, 43^63.

Barboza, G., Ferna¤ ndez, A., Barrientos, J. & Bottazzi, G.

(1997) Costa Rica: petroleum geology of the Caribbean mar-gin. Lead. Edge, 16, 1787^1794.

Barckhausen, U., Ranero, C.R., von Huene, R., Cande,S.C. & Roeser, H.A. (2003) Revised tectonic boundaries inthe Cocos Plate o¡ Costa Rica: implications for the segmenta-tion of the convergent margin and for plate tectonic models.J. Geophys. Res., 106, 19207^19220.

Barrientos, J., Bottazzi, G., Ferna¤ ndez, A. & Barboza, G.

(1997) Costa Rican data synthesis indicates oil, gas potential.Oil Gas J., 12, 76^80.

Bond, G.C., Kominz, M.A. & Sheridan, R.E. (1995) Conti-nental terraces and rises. In:Tectonics of Sedimentary Basins (Ed.by C.J. Busby & R.V. Ingersoll), pp.149^178. Blackwell Science,Oxford.

Bottazzi,G.,Fernandez,A.&Barboza,G. (1994)Sedimento-log|¤ a e historia tectono-sedimentaria de la cuenca Limo¤ n Sur.In: Geology of an Evolving Island Arc,The Isthmus of Southern Ni-caragua, Costa Rica and Western Panama¤ (Ed. by H. Seyfried &W.Hellmann), Pro¢l, 7, 351^389.

Boyer, S.E. & Elliott,D. (1982) Thrust systems. AAPGBull.,66, 1196^1230.

Brandes, C., Astorga, A., Back, S., Littke, R. & Winse-

mann, J. (2007a) Deformation style and basin- ¢ll architectureof the o¡shore Limo¤ n back-arc basin (Costa Rica).Mar. Petrol.Geol., 24, 277^287.

Brandes, C., Astorga, A., Back, S., Littke, R. & Winse-

mann, J. (2007b) Fault controls on sediment distribution pat-terns, Limo¤ n Basin, Costa Rica. J. Petrol. Geol., 30(1), 25^40.

Brandes, C., Astorga, A., Blisniuk, P., Littke, R. &Winse-

mann, J. (in press) Anatomy of anticlines, piggy-back basinsand growth strata: a case study from the Limo¤ n fold-and-thrust belt, Costa Rica. International Association of Sedimentolo-gists Series 38: Sedimentary Processes, Environments andBasins (Ed. by G. Nichols, F.Williams & C. Paola), pp.91^110.

Burbridge, D.R. & Braun, J. (2002) Numerical models ofthe evolution of accretionary wedges and fold-and-thrustbelts using the distinct-element method. Geophys. J. Int., 148,542^561.

Calvo, C. (2003) Provenance of plutonic detritus in cover sand-stones of Nicoya Complex, Costa Rica: cretaceous unroo¢nghistory of a Mesozoic ophiolite sequence. Geol. Soc. Am. Bull.,115, 832^844.

Campos, L. (2001) Geology and basins history of middle CostaRica: an intraoceanic island arc in the convergence betweenthe Caribbean and the central paci¢c plates.Tˇbinger Geowis-senschaftliche Arbeiten, Reihe A, Band 62, 138 pp

Coates,A.G.,Aubry,M.-P.,Berggren,W.A.,Collins,L.S.&Kunk,M. (2003) Early Neogene history of the Central Amer-ican arc from Bocas del Toro, western Panama. Geol. Soc. Am.Bull., 115, 271^287.

Coates, A.G., Jackson, J.B.C., Collins, L.S., Cronin,T.M.,Dowsett, H.J., Bybell, L.M., Jung, P. & Obando, J.A.(1992) Closure of the Isthmus of Panama: the near-shore mar-ine record of Costa Rica and western Panama. Geol. Soc. Am.Bull., 104, 814^828.

Collins,L.S., Coates, A.G., Jackson, J.B.C. &Obando, J.A.(1995) Timing and rates of emergence of the Limo¤ n and Bocasdel Toro basins: Caribbean e¡ects of Cocos Ridge subduction?In:GeologicandTectonicDevelopment of theCaribbeanPlateBound-

ary in Southern Central America (Ed. by P. Mann),Geol. Soc. Am.Spec. Pap. 295, 263^289.

Dahlen, F.A. (1984) Noncohesive critical coulomb wedges: anexact solution. J. Geophys. Res., 89, 10125^10133.

Dahlen, F.A., Suppe, J. &Davis, D. (1984) Mechanics of fold-and-thrust belts and accretionary wedges: cohesive coulombtheory. J. Geophys. Res., 89, 10087^10101.

Dahlstrom, C.D.A. (1969) Balanced cross-sections. Can.J. Earthsci., 6, 743^757.

Darce,M., Baca, D., Duarte,M., Barboza, G. & Fernadez,A. (2000) New law, seismic information smooth Nicaragua li-censing.Oil Gas J., 14, 72^74.

Davis,D., Suppe, J. &Dahlen, F.A. (1983) Mechanics of fold-and-thrust belts and accretionary wedges. J. Geophys. Res., 88,1153^1172.

De Boer, J.Z.,Drummond,M.S., Bordelon,M.S.,Defant,M.J., Bellon, H. &Maury, R.C. (1995) Cenozoic magmaticphases of the Costa Rican island arc (Cordillera deTalamanca).In:GeologicandTectonicDevelopment of theCaribbeanPlateBound-ary in Southern Central America (Ed. by P. Mann),Geol. Soc. Am.Spec. Pap., 295, 35^55. Boulder, Colorado.

DeMets, C. (2001) A new estimate for present-day Cocos-Car-ibbean plate motion: implications for slip along the CentralAmerican volcanic arc.Geophys. Res. Lett., 28, 4043^4046.

Dennison, J.M.&Woodward,H.P. (1963)Palinspasticmaps ofcentral Appalachians.AAPGBull., 47, 666^680.

DePaor,D.G. (1988) Balanced section in thrust belts part1: con-struction.AAPGBull., 72, 73^90.

Di Marco,G., Baumgartner, P.O. &Channell, J.E.T. (1995)Late cretaceous-early tertiary paleomagnetic data and a re-vised tectonostratigraphic subdivision of Costa Rica and wes-tern Panama. In: Geologic and Tectonic Development of theCaribbean Plate Boundary in Southern Central America (Ed. by P.Mann),Geol. Soc. Am. Spec. Pap., 295, 1^27.

Donnelly,T.W. (1989) Geologic history of the Caribbean andCentral America. In:TheGeology ofNorth America Ân overview(Ed. by A.W. Bally & A.R. Palmer), Geol. Soc. Am. Spec. Pap. A,299^321.

Durand,B.,Alpern,B.,Pittion, J.L.&Pradier,B. (1986)Re-£ectance of vitrinite as a control of thermal history of sedi-ments. In: Thermal Modeling in Sedimentary Basins (Ed. by J.Burrus),Collect. Colloq. Se¤ min., 44, 441^474.

Escalante,G.&Astorga,A. (1994) Geolog|¤ a del este de CostaRica y el norte de Panama. Rev. Geol. Ame¤ r. Central.Vol. esp.Ter-remoto Limo¤ n, 1^14.

Fernandez, J.A., Bottazzi, G., Barboza, G. & Astorga, A.(1994) Tecto¤ nica y estratigra¢a de la Cuenca Limo¤ n Sur. Rev.Geol. Am. Central.Vol. esp.Terremoto Limo¤ n, 15^28.

Friberg, L.J. (2001) Untersuchungen zurTemperatur- und Ab-senkungsgeschichte sowie zur Bildung und Migration vonMethan und molekularem Sticksto¡ im NordostdeutschenBecken.Berich. Forschungszentr. Jˇlich, 3914, 248pp.

Frisch, W., Meschede, M. & Sick, M. (1992) Origin of theCentral American ophiolites: evidence from paleomagneticresults.Geol. Soc. Am. Bull., 104, 1301^1314.

Gayer, R., Garven, G. & Ricard, D. (1998) Fluid migrationand coal-rank development in foreland basins. Geology, 26,678^682.

Gosnold,W.D. (1985) Heat £ow and ground water £ow in thegreat plains of the United States. J. Geodyn., 4, 247^264.

Gosnold,W.D. & Fischer,D.W. (1986) Heat £ow studies in se-dimentary basins. In:Thermal Modelling in Sedimentary Basins(Ed. by J. Burrus),Collect. Colloq. Se¤ min., 44, 199^217.


C. Brandeset al.

Gr�fe, K., Frisch,W.,Villa, I.M. & Meschede, M. (2002)Geodynamic evolution of southern Costa Rica related to low-angle subduction of the Cocos Ridge: constraints from ther-mochronology.Tectonophysics, 348, 187^204.

Gutscher,M.A. (2002) Andean subduction styles and their ef-fect on thermal structure and interplate coupling. J. South Am.Earth Sci., 15, 3^10.

Husson, L. & Moretti, I. (2002) Thermal regime of fold andthrust belts ^ an application to the Bolivian sub Andean zone.Tectonophysics, 345, 253^280.

Hutchinson, D.R., Grow, J.A., Klitgord, K.D. & Swift,B.A. (1982) Deep structure and evolution of the CarolinaTrough.AAPGMem., 34, 129^152.

Hyndman,R.D.,Currie,C.A.&Mazotti,S.P. (2005) Subduc-tion zone backarcs, mobile belts, and orogenic heat. GSATo-day, 15, 4^10.

Jensen, P.K. (1997) Modelling the history of burial, tempera-ture, and hydrocarbon generation of sedimentary basins. Ap-plication to the Aars-1Awell.Geophys.Trans., 41, 65^94.

Kolarsky,R.A.,Mann, P.&Montero,W. (1995) Island arc re-sponse to shallow subduction of the Cocos Ridge, Costa Rica.In:GeologicandTectonicDevelopment of theCaribbeanPlateBound-ary in Southern Central America (Ed. by P. Mann),Geol. Soc. Am.Spec. Pap., 295, 235^262.

Krawinkel, J.J. (2003) Struktur undKinematik am konvergen-ten Plattenrand der sˇdlichen Zentralamerikanischen Land-brˇcke (Zentral- und Sˇd-Costa Rica, West-Panama¤ ). Pro¢l,20, 36pp.

Krawinkel, J.& Seyfried,H. (1994) A review of plate-tectonicprocesses involved in the formation of the southwestern edgeof the Caribbean Plate. In:Geology of an Evolving Island Arc,TheIsthmus of Southern Nicaragua, Costa Rica and Western Panama¤(Ed. by H. Seyfried &W.Hellmann), Pro¢l, 7, 47^61.

Krawinkel,H.,Seyfried,H.,Calvo,C.&Astorga,A. (2000)Origin and inversion of sedimentary basins in southern Cen-tral America. Z. Angew. Geol. SH, 1, 71^77.

Larter,S. (1989)Chemical models ofvitrinite re£ectance evolu-tion.Geol. Rundsch., 78, 349^359.

Lerche, I.,Yarzab, R.F. &Kendall, C.G.ST.C. (1984) Deter-mination of paleoheat £ux from vitrinite re£ectance data.AAPGBull., 68, 1704^1717.

Lucazeau, F. & Le Douaran, S. (1985) The blanketing e¡ect ofsediments in basins formed by extension: a numerical model.Application to the Gulf of Lion and Viking graben. Earth Pla-net. Sci. Lett., 74, 92^102.

Lundberg,N. (1991) Detrital record of the early Central Amer-ican magmatic arc: petrography of intraoceanic forearc sand-stones, Nicoya Peninsula, Costa Rica.Geol. Soc. Am. Bull., 103,905^915.

Lutz, R. (2002) Numerische Simulation der Kohlenwassersto¡-genese an der Subduktionszone vor Costa Rica. Diss. RWTHAachen, 120pp.

Lutz, R., Littke, R., Gerling, P. & B˛nnemann, C. (2004)2D numerical modelling of hydrocarbon generation in sub-ducted sediments at the active continental margin of CostaRica.Mar. Petrol. Geol., 21, 753^766.

Majorowicz, J.A., Jones, F.W., Lam, H.L. & Jessop, A.M.

(1984)The variability of heat £owboth regional andwith depthin southernAlberta, Canada: e¡ect of groundwater £ow?Tecto-nophysics, 106, 1^29.

Majorowicz, J.A., Jones, F.W., Lam, H.L. & Jessop, A.M.

(1985) Terrestrial heat £ow and geothermal gradients in rela-

tion to hydrodynamics in the Alberta Basin; Canada. J. Geody-namics, 4, 265^283.

Maresch, W.V., St˛ckhart, B., Baumann, A., Kaiser, C.,Kluge, R., Krˇckhans-Lueder, G., Brix,M.R. & Thom-

son, S. (2000) Crustal history and plate tectonic developmentin the southern Caribbean. Z.Angew. Geol. SH1, 283^290.

Mareschal, J.C. (1987) Subsidence and heat £ow in intraconti-nental basins and passive margins. In: Sedimentary Basins andBasin-Forming Mechanisms (Ed. by C. Beaumont & A.J. Tan-kard),Can. Soc. Petrol. Geol., 12, 519^527.

McCulloh,T.H. &Naeser,N.D. (1989) Thermal history of se-dimentary basins: introduction and overview. In:Thermal His-tory of Sedimentary Basins (Ed. by N.D. Naeser & T.H.McCulloh), pp.1^11. SpringerVerlag, NewYork.

McNeill,D.F.,Coates,A.G.,Budd,A.F.&Borne,P.F. (2000)Integrated paleontologic and paleomagnetic stratigraphy ofthe Upper Neogene deposits around Limon, Costa Rica: acoastal emergence record of the Central American Isthmus.Geol. Soc. Am. Bull., 112, 963^981.

Mende, A. (2001) Sedimente und Architektur der Forearc- undBackarc-Becken von Sˇdost-Costa Rica und Nordwest-Pana-ma¤ . Pro¢l, 19, 130pp.

Meschede, M. & Frisch,W. (1998) A plate tectonic model forthe Mesozoic and Early Cenozoic history of the CaribbeanPlate.Tectonophysics, 296, 269^291.

Meschede,M.,Frisch,W.,Chinchilla Chavez,A.L.,Lo¤ pezSaborio, A. & Calvo, C. (2000) The plate tectonic evolutionof the Caribbean Plate in theMesozoic and Early Cenozoic. Z.Angew. Geol. SH1, 275^281.

Mitra, S. (1990) Fault-propagation folds: geometry, kinematicevolution, and hydrocarbon traps.AAPGBull., 74, 921^945.

Munoz, A., Baca,D., Artiles,V. &Duarte,M. (1997) Nicara-gua: petroleum geology of the Caribbean margin. Lead. Edge,16, 1799^1805.

N˛th, S.,Karg,H.&Littke,R. (2001) Reconstruction of LatePalaeozoic heat £ows and burial histories at the Rhenohercy-nian-Subvariscan boundary, Germany. Int. J. Earth Sci., 90,234^256.

Nunn, J.A.,Hanor, J.S.&Lee,Y. (2005)Migration pathways inthe Central North Slope foreland basin, Alaska, USA: soluteand thermal constraints on £uid £ow simulations. Bas. Res.,17, 403^416.

Petmecky, S.,Meier, L., Reiser,H. & Littke, R. (1999) Highthermal maturity in the Lower Saxony Basin: intrusion ordeep burial?Tectonophysics, 304, 317^344.

Petzet, G.A. (1998) Costa Rica awards blocks on Caribbeancoast.Oil Gas J., 12, 76^80.

Pindell, J.L.,Cande, S.C., Pitman,W.C.,Rowley,D.B.,De-

wey, J.F.,Labrecque, J.&Haxby,W. (1988)Aplate-kinematicframework for models of Caribbean evolution. Tectonophysics,155, 121^138.

Poelchau,H.S., Baker,D.R.,Hantschel,T.,Horsfield, B.&Wygrala, B. (1997) Basin simulation and the design of theconceptual model. In: Petroleum and Basin Evolution (Ed. byD.H. Welte, B. Hors¢eld & D.R. Baker), pp. 3^70. SpringerVerlag, Berlin.

Price,R.A. (2001) An evaluation of models for kinematic evolu-tion of thrust and fold belts: structural analysis of a transversefault zone in theFront ranges of theCanadianRockies north ofBan¡, Alberta. J. Struct. Geol., 23, 1079^1088.

Protti,M.,Gˇendel,F.&McNally,K. (1995)Correlation be-tween the age of the subducting Cocos plate and the geometryof the Wadati-Benio¡ zone under Nicaragua and Costa Rica.



In:GeologicandTectonicDevelopment of theCaribbeanPlateBound-ary in Southern Central America (Ed. by P. Mann),Geol. Soc. Am.Spec. Pap., 295, 309^326.

Radke,M.,Horsfield,B.,Littke,R.&Rullk˛tter, J. (1997)Maturation and petroleum generation. In: Petroleum and BasinEvolution (Ed. by D.H.Welte, B. Hors¢eld & D.R. Baker), pp.169^230. Springer, Berlin.

Ramsey, J.G. & Huber, M.I. (1987) The Techniques of ModernStructural Geology,Vol. II. Academic Press, London, 700pp.

Ranero, C.R. & von Huene, R. (2000) Subduction erosionalong the Middle America convergent margin. Nature, 404,748^752.

Ranero, C.R.,von Huene, R., Flueh, E.,Duarte,M., Baca,D. &McIntosh,K. (2000a) A cross section of the convergentPaci¢c margin of Nicaragua.Tectonics, 19, 335^357.

Ranero, C.R., von Huene, R., Flueh, E.R.,Weinrebe,W.,Hinz, K., Leandro, G., Alvarado, G. & Duarte, M.

(2000b) Lower plate control on subduction erosion processesalong the middle America convergent margin. Z. Angew. Geol.SH1, 291^296.

Rodon, S.&Littke,R. (2005) Thermal maturity in the CentralEuropean Basin system (Schleswig-Holstein area): results of1D basin modelling and new maturity maps. Int. J. Earth Sci.,94, 815^833.

Rodriguez, J.F.R. & Littke, R. (2001) Petroleum generationand accumulation in the Golfo San Jorge Basin, Argentina: abasin modeling study.Mar. Petrol. Geol., 18, 995^1028.

Ross,M.I. & Scotese,C.R. (1988) A hierachical tectonic modelof the Gulf ofMexico and the Caribbean region.Tectonophysics,155, 139^168.

Schmidt, F. & Erdogan, L.T. (1993) Basin modelling in anoverthrust area of Austria. In: Basin Modelling: Advances andApplications (Ed. By A.G. Dore¤ ),NPFSpec.Publ., 3, 573^581.

Schneider (2003) Basin modeling in complex area: examplesfrom easternVenezuelan and Canadian foothills. Rev. IFP, 58,313^324.

Seyfried,H., Astorga, A., Amann,H., Calvo, C., Kolb,W.,Schmidt,H.&Winsemann, J. (1991)Anatomy of an evolvingIsland Arc: tectonic and eustatic control in the south CentralAmerican forearc area. In: Sea-Level Changes at Active PlateMargins: Processes and Products (Ed. by D.I.M. MacDonald), Int.Assoc. Sediment. Spec. Publ., 12, 273^292.

Sheehan, C.A., Penfield, G.T. & Morales, E. (1990) CostaRica geologic basins lure wildcatters.Oil Gas J., 30, 74^79.

Silver, E.A.,Galewsky, J. &McIntosh,K.D. (1995) Variationin structure, style, and driving mechanism of adjoining seg-ments of theNorth Panama deformed belt. In:GeologicandTec-tonic Development of the Caribbean Plate Boundary in SouthernCentral America (Ed. by P. Mann),Geol. Soc. Am. Spec. Pap., 295,309^326.

Sleep,N.H. (1979) A thermal constraint on the duration of fold-ing with reference to Acadian geology, New England (USA).J. Geol., 87, 583^589.

Suppe, J. (1983) Geometry and kinematics of fault-bend folding.Am. J. Sci., 283, 684^721.

Suppe, J.&Medwedeff,D.A. (1990) Geometry and kinematicsof fault-propagation folding. Eclogae Geol. Helv., 83, 409^454.

Sweeney, J.J. & Burnham, A.K. (1990) Evaluation of a simplemodel of vitrinite re£ectance based on chemical kinetics.AAPGBull., 74, 1559^1570.

Taylor, G.H.,Teichmˇller, M., Davis, A., Dissel, C.F.K.,Littke,R. &Robert, P. (1998)Organic Petrology. Borntraeger,Berlin, 704pp.

Ungerer, P., Burrus, J.,Doligez, B., Chenet, P.Y. & Bessis,F. (1990) Basin evaluation by integrated two-dimensionalmodeling of heat transfer, £uid £ow, hydrocarbon generationand migration.AAPGBull., 74, 309^335.

Verge¤ s, J.,Mun� oz, J.A.&Mart|¤nez,A. (1992) South Pyreneanfold and thrust belt: the role of foreland evaporitic levels inthrust geometry. In: Thrust Tectonics (Ed. K.R. McClay), pp.255^263. Chapman andHall, London, UK.

von Huene, R. & Flˇh, E. (1994) A review of marine geophysi-cal studies along the Middle America Trench o¡ Costa Ricaand the problematic seaward terminus of continental crust.In: Geology of an Evolving Island Arc,The Isthmus of Southern Ni-caragua, Costa Rica and Western Panama¤ (Ed. by H. Seyfried &W.Hellmann), Pro¢l, 7, 143^159.

Walther, C.H.E. (2003) The crustal structure of the Cocosridge of Costa Rica. J. Geophys. Res., 108, 1^21.

Weinberg, R.F. (1992) Neotectonic development of westernNicaragua.Tectonics, 11, 1010^1017.

Welte, D.W. & Yalcin, M.N. (1987) Basin modeling ^ A newcomprehensive method in petroleum geology. Adv. Org. Geo-chem., 13, 1^3.

Welte,D.H.&Yˇkler,M.A. (1981) Petroleum origin and accu-mulation in basin evolution ^ a quantitative model. AAPGBull., 65, 1387^1396.

Werner, R.,Hoernle,K.,van den Bogaard, P., Ranero, C.& von Huene, R. (1999) Drowned 14-m.y.-old Gala¤ pagos ar-chipelago o¡ the coast of Costa Rica: implications for tectonicand evolutionary models.Geology, 27, 499^502.

Weyl, R. (1980) Geology of Central America. Borntraeger, Berlin,371pp.

Wilkerson, M.S. & Dicken, C.L. (2001) Quick-look techni-ques for evaluating two-dimensional cross sections in de-tached contractional settings.AAPGBull., 85, 1759^1770.

Winsemann, J. (1992) Tiefwasser-Sedimentationsprozesse und-produkte in den Forearc-Becken des mittelamerikanischenInselbogensystems: eine sequenzstratigraphische Analyse.Pro¢l, 2, 218pp.

Winsemann, J. & Seyfried, H. (1991) Response of deep-waterforearc systems to sea-level changes, tectonic activity and vol-caniclastic input in central America. In: Sea-Level Changes atActive Plate Margins: Processes and Products (Ed. by D.I.M. Mac-Donald), Int. Assoc. Sediment. Spec. Publ., 12, 217^240.

Wygrala, B.P.,Yalcin, M.N. & Dohmen, L. (1990) Thermalhistories and overthrustingâpplication of numerical simula-tion technique.Org. Geochem., 16, 267^285.

Yahi,N.,Sch�fer,R.G.&Littke,R. (2001) Petroleum genera-tion and accumulation in the Berkine basin, eastern Algeria.AAPGBull., 85, 1439^1467.

Yalcin,M.N. (1991) Basin modelling and hydrocarbon explora-tion. J. Petrol. Sci. Eng., 5, 379^398.

Yalcin, M.N., Littke, R. & Sachsenhofer, R.F. (1997)Thermal history of sedimentary basins. In:Petroleum andBasinEvolution (Ed. byD.H.Welte, B.Hors¢eld&D.R.Baker), pp.71^167. Springer, Berlin.

Zhou,Y. & Littke, R. (1999) Numerical simulation of the ther-mal maturation, oil generation and migration in the Songliaobasin, Northeastern China.Mar. Petrol. Geol., 16, 771^792.

Manuscript received 22 February 2006; Manuscript accepted12October 2007.


C. Brandeset al.

Documents

BasinResearch BasinmodellingoftheLimo¤nBack-arcBasin