T41C-0697
Seismic Evidence for Widespread Serpentinized Forearc Mantle Along the Mariana Convergence Margin
We use P-to-S converted phases from teleseisms recorded at broadband stations in the Mariana Islands to image the forearc and arc regions of the Mariana convergence margin. The Moho in the subducting Pacific plate is observed at depths between 75 and 110 km beneath the region extending from Rota to Saipan. The S-wave velocity in the subducting crust is inferred to be ~10% slower than the surrounding mantle. This demonstrates that the crust has not yet undergone conversion to eclogite at these depths, in agreement with observations made for other arcs. A low velocity zone (LVZ), approximately 10--25 km thick, whose upper boundary is imaged at about 40--55 km depth, is detected in the forearc region of the mantle wedge along the entire margin. The anomaly is located too shallow to represent subducted oceanic crust. We interpret the LVZ as a serpentinized region in the forearc mantle, resulting from hydration by slab-expelled water. The occurrence of the serpentinized zone along the entire margin suggests that serpentinization of the forearc mantle is a widespread phenomenon in the Mariana arc. The inferred S wave velocity in the LVZ of as low as ~3.6 km/s represents a level of serpentinization of 30--50%, corresponding to a water content of about 4--6 wt%. http://epsc.wustl.edu/seismology/MARIANA
T41C-0698
The relationship between the seismic characteristics of crustal structure in Shikoku Basin and en-echelon arrangement
Detailed crustal structure information of a back-arc basin must be obtained to elucidate the mechanism of its opening. Especially, the Shikoku Basin, which occupies the northern part of the Philippine Sea Plate between the Kyushu-Palau Ridge and the Izu-Ogasawara Arc, is an important area to elucidate the evolution of the back-arc basins as a part of the growth process of the Philippine Sea. Japan Agency for Marine-Earth Science and Technology (JAMSTEC) carried out multi-channel seismic reflection survey using 12,000 cu.in. air gun and streamer with 204 ch hydrophones in the Izu-Ogasawara region since 2004. The total length of survey lines is more than 10,000 km until 2006. We investigate the crustal structure beneath the Shikoku Basin along 10 survey lines, which are across to the strike of the en-echelon seamount chains in the rear arc. From the seismic profiles, some faults and intrusion structures are obtained in the Shikoku Basin. The deformation structure with acoustic basement is widely distributed between the Shikoku Basin and the Izu-Ogasawara arc. Some intrusions structure is identified in the Shikoku Basin are exposed on seafloor. The intrusions structure is assumed to locate in the extended region of the en-echelon arrangement. The strike-slip faults with flower structure cutting whole sediments are located in the arc-backarc transition zone in the northern Shikoku Basin, suggesting that this region is in share stress. On the other hand, these structures indicating the deformation and intrusions are not recognized in the eastern side of the Kyushu-Palau Ridge. The Izu-Ogasawara arc is colliding to the Japan Island arc in the Sagami Bay. In the Nankai Trough, the Philippine Sea plate is subducting to the Japan Island arc. Therefore, the strike-slip and reverse fault would be developed by the compression stress in the eastern side of Philippine Sea plate. If the en-echelon arrangement is developed along these faults, the intrusions structure obtained by our surveys correspond the tip of en-echelon arrangement. We will also discuss the effect with regard to back-arc opening and post volcanism.
T41C-0699
Distribution and nature of the crust-mantle transition layer deduced from amplitude modeling of wide-angle seismic data along the Izu-Bonin island arc
The Izu-Bonin island arc formed by subducting of the Pacific plate beneath the Philippine Sea plate is a region of the crustal growth. This arc is divided into the northern and the southern part by the Sofu-gan tectonic line, due to the difference of the geological and geophysical characters (Yuasa, 1985). In the seismic velocity structure along the volcanic front in the northern Izu-Bonin island arc, it is clarified that this arc has not only the middle and the lower crust but also the 7.2-7.6 km/s layer (crust-mantle transition layer) underlying the lower crust (Kodaira et al., 2007, accepted). However, since this velocity structure along the volcanic front in this arc is calculated by the tomography method, the nature of the crust-mantle transition layer and uppermost mantle is unknown. To understand the nature of the crust-mantle transition layer in the Izu-Bonin island arc on the crustal growth, it is also important to know the seismic reflectivity at the top and bottom of this transition layer in this arc. In this study, we clarify the distribution of the seismic reflectivity at the top and bottom of the crust-mantle transition layer along the volcanic front in this arc using the velocity contrast values at these reflectors estimated by the amplitude modeling of wide-angle data. In 2004 and 2005, seismic refraction/reflection surveys using ocean bottom seismographs (OBSs) and controlled sources were conducted along the volcanic front in the Izu-Bonin island arc from the Sagami Bay to Kaitoku Seamount (Kodaira et al., accepted). In record sections of several OBSs, not only the first arrived phases but also later phases reflected from interfaces in the crust and uppermost mantle can be observed. The velocity contrast values at the top and bottom of this transition layer were estimated from the comparison of the observed and synthetic wave forms computed by a finite difference wave propagation program code "e3d" (Larsen and Grieger, 1998). Along the northern Izu-Bonin island arc, the top of the crust-mantle transition layer has a velocity contrast value of about 0.4 km/s ranging from Kurose hall to Kayo Seamount. On the other hand, the velocity contrast of the bottom of the crust-mantle transition layer is weaker southward along the northern Izu-Bonin island arc. From these results of this study and the petrological study (e.g., Tatsumi, 2000), this transition layer along the northern Izu-Bonin island arc is interpreted as the mixture material of the mafic restites and olivine cumulate during the crustal growth. In addition, the variation of the velocity contrast at the bottom of this transition layer along the northern Izu-Bonin island arc implies that this transition layer has more olivine cumulate southward along this arc. Besides, record sections of some OBSs are characterized by fuzzy reflections from the top and bottom of the crust-mantle transition layer looking southwards bounded on the Sofu- gan tectonic line, suggesting that the character of the crust-mantle transition layer of the northern Izu-Bonin island arc differs from that of the southern arc.
T41C-0700
Crustal and uppermost mantle structure suggesting the influences of arc and rifting activities at the northernmost part of Mariana Trough
Izu-Bonin and Mariana Arcs (IBM) are oceanic island arcs growing by the arc activity associated with the subduction of the Pacific Plate beneath the Philippine Sea Plate. The seismic studies in the IBM have been revealed the existences of andesitic middle crust, high-velocity (~7 km/s) lower crust, and low-velocity uppermost mantle beneath the arcs (e.g. Suyehiro et al., 1996; Takahashi et al., 2007). Based on the seismic structure and petrological calculation, the lower crustal volumes between the two results are different and requiring the process transforming the denser crustal materials into the mantle (Takahashi et al., 2007). Although, the results suggesting the formation of continental crust have been discovered in the IBM region as mentioned above, the influences of rifting activities to the oceanic island arcs have not been revealed yet. Therefore, we will show the results of a seismic study at the northernmost part of Mariana Trough and discuss the influences of arc and rifting activities. The seismic structure at the northernmost part of Mariana Trough shows anomalously thick crust (~25 km) and relatively lower P-wave velocity (~7.6 km/s) beneath the thick crust than normal mantle velocity. The peak of anomalously thick crust is located not beneath the volcanic front but beneath the extension of trough axis. This implies that the accretion at the base of crust is larger at the rift zone than that of volcanic front. Therefore, the crustal thickness beneath the volcanic front is not remarkable. From the seismic structure at the northernmost part of the Mariana Trough, the crustal accretions of the rifting activities seem to be larger than those of the arc activity.
T41C-0701
Seismic Velocity structures in Northern Izu-Bonin arc derived from passive OBS observations
The Izu-Bonin Island arc is an oceanic island arc, where the Pacific plate subducts beneath the Philippine Sea plate. Recent active seismic surveys in the Izu-Bonin arc show significant variations in thickness of the middle crust along the volcanic front [Kodaira et al, 2007]. To understand the crustal evolution in the oceanic island arc, we have to clarify structures in the mantle wedge along the arc in addition to the oceanic island arc crust. We conducted seismicity observations to investigate structure variations in northern Izu-Bonin arc using natural earthquakes. A temporal ocean bottom seismograph (OBS) network consists of 40 pop-up type OBSs was deployed in April 2006 between Tori-shima and Hachijo-jima islands. These OBSs were retrieved in July after about 80-day observations. We used continuous seismic data at 36 OBSs and three F-net and Hi-net seismic stations on Hachijo-jima and Aoga-shima islands operated by National Research Institute for Earth Science and Disaster Prevention. During the OBS observations, about 1600 earthquakes were located. These earthquakes clearly show double seismic zone along the subducting Pacific plate. We estimated 1D and 3D P- and S-wave seismic velocity structure using arrival time data of these earthquakes. The 1D velocity model shows that a layer with low Poissonfs ratio of 0.24 and high Poissonfs ratio of 0.28 corresponds to middle and lower crust, respectively. The low Poissonfs ratio layer suggests the granitic middle crust with Vp of ~6km/s. The high Poissonfs ratio layer agrees with the gabbroic lower crust as suggested by Kodaira et al. [2007]. Three-dimensional Vp and Vs structures were estimated by 3D tomographic inversion method by Kamiyra and Kobayashi [2000] using the 1D model as an initial model. The estimated 3D model shows structure variations along the volcanic front. We will discuss relationships between the seismic velocity variations and the island arc crust structures in northern Izu- Bonin arc.
T41C-0702
The seismic structure of the Rivera subduction zone
The subduction zone of western Mexico is a unique region on Earth where microplate capture and overriding plate disruption are occurring today. The small Rivera plate is subducting beneath western most Mexico primarily beneath Jalisco state while to the east it is the Cocos plate that is subducting. Above the Rivera plate the Jalisco block of Mexico is bounded by the north trending Colima Rift and the northwest trending Tepic-Chapala Rift and may form a microplate in its own right. Magmatism is present throughout the region and is unusual for a subduction zone in that geochemical analyses indicate an ocean island basalt component to some of the lavas. Also, Colima volcano is offset trenchward from other volcanoes in the Mexican Volcanic Belt. Little is known of the subducting Rivera plate geometry due to the paucity of seismicity within the plate yet the geometry of the Rivera and Cocos plates at depth are likely critical for understanding the tectonic evolution of western Mexico. The MARS (MApping the Rivera Subduction zone) project consists of the deployment of 50 broadband seismometers covering the Jalisco block from the coast to the Tepic-Chapala rift in the north and about 150 km to the west of the Colima rift. The instruments were deployed in January, 2006 and removed in June, 2007. The goal of the project is to seismically image the subducting Rivera and Cocos plates at depth as well as the mantle wedge above the plates. We present the results of a P-wave tomography inversion using teleseisms recorded by MARS. The inversion used 10,495 residuals and finite frequency theory to back project the kernels through the model. At shallow depths it is difficult to discern the subducting Cocos and Rivera plates but at depths deeper than about 80 km both plates are clearly imaged in the tomography model. Below a depth of 150 km, a clear gap between the Rivera and Cocos slabs is apparent that increases in size at further depths. The images indicate that the deeper Rivera plate is subducting more steeply than the adjacent Cocos plate and also has a more northerly trajection. The break between the two plates is just to the west of Colima graben. The Rivera and western most Cocos plates reach depths exceeding 200 km to the south of the main Mexican Volcanic Belt. The Rivera plate is at roughly 140 km depth beneath the young (<3Ma) central Jalisco Volcanic lineament indicating the Rivera plate has likely recently rolled back and steepened.
T41C-0703
Imaging of Cocos Plate Beneath Southern Costa Rica From Receiver Function Analysis
A transect of 19 seismological broadband stations crossing the Talamanca Mountain Range in Southern Costa Rica was operated from March 2005 to April 2007 as a part of the Collaborative Research Center SFB 574 "Volatiles and Fluids in Subduction Zones". The aim of the seismological subproject A2 was to gain insight into the structure of the Central American subduction zone and possible pathways for fluid migration. Previous studies of active seismics and local seismicity suggested to explain the gap of volcanism in the Talamanca range with the lack of a subducting slab. They assumed that the Cocos Ridge underlies the overriding plate at a shallow dip. In contrast, our receiver function analysis of 322 teleseimic earthquakes is able to image the subducting Cocos Plate down to depths of at least 100 km. The dip angle of the slab closer to the trench is outside the network but appears to be shallow, consistent with former studies. Below 40 km, the dip increases to more than 45 deg. This is supported by accurately located seismicity from a tomography study also performed by our group. Crustal structure could also be resolved by the receiver function analysis in agreement with tomography and active seismic investigations. The existence of the subducting slab poses the question why volcanism stopped 4 Ma ago; several possible scenarios are discussed.
T41C-0704
Tomographic Results From the Nicaragua Subduction Zone: Evidence for a Slab Tear
The Nicaragua convergent margin shows different structures than Costa Rica such as a steeper slab, northward shift in the volcanic front and different geochemical traces in the volcanoes. It is also known as the wettest subduction zone in the world. To have a better understanding of the geometry of the slab and structural changes in the incoming and the overriding plate, local earthquake tomography was performed using 860 local earthquakes recorded by an amphibious network. After iterative simultenous inversion of Vp, Vp/Vs and hypocenters, the cold and dense slab is identified as a high velocity dipping structure anomaly with an angle of approx. 70 deg between depth of 20 - 150 km. Upper parts (0 - 20 km) of the incoming plate (between the trench and coast) and margin wedge show strong low velocity anomalies, which can be explained by eroded thick sediments and serpentinized oceanic crust and mantle due to high amount of fluid infiltration down to several km depths along bend-faults. The slowest mantle velocities are found directly beneath the volcanoes indicating a zone of partial melting extending to 100 - 150 km depth. The most striking observation is the alignment of hypocenters. The Wadati-Benioff zone dips with an angle of approx. 35 deg down to 70 km depth and steepens abruptly to approx. 70 deg below 70 km. An abrupt shift of seismicity towards the trench is observed near the border of Costa Rica and Nicaragua possibly indicating a slab break. This abrupt change is also recognizable in the velocity anomalies laterally correlated with the jump of the volcanic chain towards the trench between Maderas volcano in Nicaragua and Orosi volcano in Costa Rica.
T41C-0705
Investigation on subduction erosion of the Central Costa Rica margin with seismic wide- angle data
Seismic wide-angle investigations along the Pacific margin off Central Costa Rica were carried out using closely spaced ocean bottom hydrophones and seismometers along two parallel strike and two parallel dip lines, intersecting at the mid slope. The structure and the P-wave velocities of the subducted oceanic Cocos Plate and overriding Carribean Plate were determined by modeling the wide-angle seismic data combined with the analysis of coincident reflection seismic data and the use of synthetic seismograms. Detailed velocity-depth distributions of two dip-lines and two strike-lines on the continental slope will be presented. Below the slope sediment, a wedge-shaped body, the margin wedge is defined by high velocities (4.3-6.1 km/s). This wedge shows a high velocity gradient zone in the uppermost one to two km, underlain by a low velocity gradient to the plate boundary. Between the subducted plate and overriding plate the low velocity zone including a lense-type structure is seen. This Megalens (4.0-4.3 km/s) and the subducted sediment comprise a low velocity zone (LVZ) all along the plate boundary. This LVZ is constrained by joint analysis of reflection seismic data and wide-angle data. The thickness of the wedge varies along the strike, this is associated with the subduction of the extension of Quepos Plateau, which also resulted in uplift of the margin. The extensional forearc environment is manifested by the normal faults indicated on the the multi-channel seismic (MCS) data. The Megalens is most probably comprised of material transferred from upper margin wedge at the tip of the wedge. The velocity structure within the Megalense resembles the velocities at the tip of the wedge, and is clearly lower than the oceanic crust, but higher thn subducted sediment. If this interpretation is valid, this material has been transported 16 km landward, which implies it was detached from the upper plate 0.2 Ma ago.
T41C-0706
A Summary of Chlorine Stable Isotopes as a Volatile Tracer in the Central American and Izu- Bonin-Mariana Volcanic Arcs
The Cl isotope composition of volcanic gases (gas fumarole and gas condensate samples), hydrothermal waters, and lava and ash samples from volcanic centers along the Central American (CA) and Izu-Bonin-Mariana (IBM) arcs have δ37Cl values ranging from -5 to +12‰. Gas and hydrothermal waters from CA span almost this entire range, whereas lava and ash samples span -2.6 to +3.0‰. In contrast, gas and hydrothermal waters from IBM only range from -5.4 to -0.1‰; overlapping with ash and lava samples ranging from -2.6 to +0.4‰. The high δ37Cl values of volcanic gases are found only at volcanic centers in CA with large associated hydrothermal fields and/or crater lakes, such as Poás, Momotombo, and Santa Ana. It is thought that these enriched 37Cl enriched gases are due to liquid-vapor fractionation in acidic systems and record information regarding the plumbing system of a volcano (see Sharp et al., this meeting). The chlorine isotope composition of the ash and lava samples are probably more representative of the ascending magma, which may serve as a tracer of chlorine sources. δ37Cl values of ash and lava samples in Costa Rica are positive (+0.4 to +3.0‰), with the exception of one Arenal sample. In S. Nicaragua samples are negative and become slightly positive in N. Nicaragua. Samples from at then northern end of the arc are near zero (El Salvador: +0.1 to +0.9‰; Guatemala: -0.5‰). The slightly positive values in Costa Rica may indicate a mantle component with some contribution from isotopically heavy serpentinites. The negative values in Nicaragua are consistent with subducting sediments (structurally bound Cl (SBC) in offshore sediments average -0.5 ± 0.8‰; n = 11), sedimentary pore fluids, and/or isotopically light serpentinites associated with plate bending on the outer rise. Samples from the end of the arc in El Salvador and Guatemala indicate more of a mantle-like signature. In the IBM, there is no variation in δ37Cl values along the length of the arc in either gas or ash samples. In contrast, there are distinct variations across the arc, from the forearc seamounts, frontal arc gases and ashes, and the cross chain, implying variations in the fluid source at different depths within the subduction zone. δ37Cl values for SBC of serpentinite clasts and serpentine clays from the Conical, S. Chamorro, and Torishima seamounts are slightly positive (average = +0.3 ± 0.4‰; n = 19), identical values to seafloor serpentinites, suggesting a serpentinite (crysotile to antigorite transition) source. Volcanic gases and ashes are negative, consistent with a sediment-breakdown source. Subducting sediments have isotopically negative SBC δ37Cl values, ranging from -0.6 to -2.5‰ (n = 10). Basalts from the Guguan cross- chain range from 0.0 to +0.5‰ (n = 3), again indicating a serpentinite source, such as fluids derived from antigorite breakdown.
T41C-0707
Apatite volatile barometry at Volcan Irazu, Costa Rica
The 1723 and 1963 basaltic-andesite eruptions of Irazu volcano, Costa Rica, contain abundant fluorapatite phenocrysts with both OH-apatite and Cl-apatite components. The OH and Cl content of apatite may provide information about volatile contents and processes occurring in a magma chamber prior to eruption. SIMS measurements of Cl and H in apatite phenocrysts from these two eruptions demonstrate systematic differences in volatile composition in apatite from 1723 to 1963: Chlorine and hydroxyl contents are higher in apatite erupted in 1723 (Cl = 7000 - 8500 ppm; OH = 4000 - 5500 ppm) than in those from the 1963 eruption (Cl = 4500 - 6000 ppm; OH = 2500 - 4200 ppm). In a qualitative sense, these results are in agreement with the data of Benjamin et al. (JVGR, in press), which demonstrate a decline in Cl and H from 1723 to 1963 in melt inclusions from the same samples. However, in order to interpret apatite volatile content in terms of quantitative magma volatile composition, knowledge of apatite/melt partition coefficients is required. Mathez and Webster (GCA, v. 65, no.5, 2005) have demonstrated that apatite partition coefficients are likely a function of melt chemistry for Cl and F, and their data suggest that the same is true for OH partitioning between silicate melts and apatite. Using the apatite/melt partition coefficients based on the data of Mathez and Webster, without accounting for compositional differences between the basaltic andesites of Irazu and the basaltic melts represented in their experiments, results in model magmatic volatile contents for Irazu that are far in excess of measured melt inclusion values for both Cl and H2O for both 1963 and 1723 eruptions (Benjamin et al.). This lack of agreement is not surprising, as the two melt compositions (the experiments and the natural samples) are quite different. We have developed a multivariate linear regression model for Cl and OH partition coefficients allowing estimation of partition coefficients for Irazu melts. This model suggests that Cl in Irazu magmas decreased from 4500 ppm to 2400 ppm from the 1723 to 1963 eruption events, while H2O decreased from 3.4 wt% to 1.6 wt%. These estimates agree very closely with the melt inclusion data of Benjamin et al. Although this model for apatite is strongly dependent on the scarce partitioning data that exists, it demonstrates that microanalyses of volatile elements in apatite can provide robust quantitative information about magmatic Cl and H2O. However, it also emphasizes the need for additional experiments on a wide range of melt compositions in order to study diverse magmas.
T41C-0708
Episodic Volcanism and Geochemistry in Western Nicaragua
The active volcanic arc in western Nicaragua is separated from the Miocene arc by a temporal gap in the volcanic record, during which little volcanic material was erupted. Previous work suggested that this gap lasted from 7 to 1.6 Ma, during which volcanic production in Nicaragua was limited or nonexistent. Because the precise timing and duration of this gap has been poorly constrained, recent fieldwork has focused on locating samples that may have erupted close to or even during this apparent hiatus in activity. Recent 40Ar/39Ar dates reveal pulses of low- level episodic volcanism at 7 Ma and 1 Ma between the active and Miocene arcs with current volcanism beginning ~350 ka. In addition, sampling from an inactive area between Coseguina and San Cristobal yielded two distinct groupings of ages; one of Tamarindo age (13 Ma) and the other around 3.5 Ma-the only samples of that age collected on-strike with the active arc. This raises the possibility the bases of the other active volcanoes contain lavas that are older than expected, but have been covered by subsequent eruptions. The Miocene arc differs from the active arc in Central America in several ways, with the latter having higher Ba/La and U/Th values due to increased slab input and changes in subducted sediment composition. Analysis of sample C-51 and others taken from the same area may shed light on the timing of this shift from high to low Ba/La and U/Th values. More importantly, it may help explain why the arc experienced such a dramatic downturn in volcanic production during this time. We also report 25 new major and trace element analyses that shed some light on the origins of these minor episodes of Nicaraguan volcanism. These samples are currently awaiting Sr and Nd isotopic analyses.
T41C-0709
Source components and intensive parameters of magma genesis in the CentAm and North IBM arcs: analyses using Arc Basalt Simulator model
We have developed a general mass balance model nfor magma genesis in subduction zones and applied it to IBM and Central American arcs. The Arc Basalt Simulator (ABS) model includes: 1) P-T dependent compositional variations of fluids from subducted altered oceanic crust (AOC) and sediment (SED); 2) zone refining chemical modification of slab-derived fluid by interaction with mantle peridotite; and 3) metasomatism and fluid flux melting of mantle peridotite caused by the modified fluid. Application of the model to the northern Izu arc (N-Izu) and Central America arc (CentAm) highlights differences between the two arc systems. The N-Izu basalts are modeled to have derived from a common SED/AOC = 10/90 slab composite with slab fluid dehydration at 880C/4GPa (VF: volcanic front) and 980C/6GPa (RA: rear arc). Mantle wedge peridotite (PERID) composition is assumed to be a common 8% MORB extracted primitive mantle (DMM) with depleted DM isotopic composition for both VF and RA. Melting conditions of the mantle are estimated to be F=24% with 0.4-0.5% fluid flux rate at 1.0 GPa (VF) and 2-4%F with 0.07% fluid flux rate at 2.3 GPa (RA). The CentAm arc basalts require at least three AOC components:Cocos-Nazca Spreading Center (CNS), Northern Galapagos domain (NGD), and Southern Galapagos domain (SGD), with increasing HIMU component in that order. SED component is a unique mixture between hemipelagic and carlcareous sediments. The PERID component ranges from undepleted PM to 4% MORB-depleted PM (DPM) with isotopic composition represented by Utila OIB in the RA of Honduras, which is unaffected by Cocos Plate components. The Guatemala-El Salvador VF basalt requires CNS-AOC, SED, and DPM with SED/AOC = 1/99 slab, dehydrated at 950C/4GPa, and mantle melting at 2%F/2.4GPa with flux rate at 2%. RA alkali basalt of the same segment (Yojoa volcano) requires same AOC and SED but needs undepleted PM melted with fluid flux from SED/AOC = 3/97 slab dehydrated at 1000C/6GPa, and mantle melting at 3%F/2.3GPa with a flux rate 0.05%. Costa Rica segment VF basalts (strongest HIMU signature) require NGD- AOC, SED, and DPM peridotite for Arenal and PM for Platanar basalts. Slab dehydration conditions are SED/AOC = 2/98 for Arenal and 0.7/99.3 for Platanar at 1000C/6GPa and 1100C/6GPa, respectively. Melting conditions are 2%F/2.3GPa with flux rate 0.2% for Arenal and 5%/2.3GPa with flux rate 2% for Platanar. The most rear arc OIB La Providencia basalt needs the strongest HIMU SGD-AOC with SED/AOC = 10/90 slab dehydrated at 1000/6GPa fluxed to melt PM at 1%F/2.3GPa with 0.01% flux rate. Slab dehydration temperature is generally higher beneath CentAm (950-1100C) than beneath N-Izu (880-980C), especially for Costa Rica (1000-1100), reflecting the much younger age of the CentAm slab. Mantle melting degree is high for N-Izu VF (24%F) but low for CentAm and N-Izu RA (2-5%F). Fluid flux rate is low beneath RAs of both arcs (0.01-0.05%) and larger beneath both volcanic fronts (0.2-2%). VF sub-alkaline basalts have relatively low fluid flux (0.05-0.5%) compared to larger flux for the alkali basalt (2%) in VF (Platanar) and behind VF (Yojoa). Intensive parameters derived from the calculations appear to be consistent with the two different tectonic settings between N-Izu and CentAm.
T41C-0710
A 3D synoptic model of Central America inferred from gravity data interpretation
Large portions of the Central American Isthmus have served as key areas for the collaborative research program (SFB 574) and its goal to understand orogenic processes at convergent margins, such as the volatile and fluid cycle and the relationships between tectonics and magmatism. Gravity data from both on- and offshore has been gathered from various institutions and has been combined in a homogeneous data set. Due to difficult access to the high mountains the coverage by gravity observations remains rather incomplete mainly in the area of southern Costa Rica and eastern Nicaragua. Station complete Bouguer anomalies, Free Air anomalies and isostatic residual anomalies maps were compiled as a result of the homogenization of gravity field data. First analyses of the gravity field using curvature methods helps to separate density provinces in the crust. A comparison with the geological map shows a good correlation with tectonical units in most of the region and provides possibilities for crustal segmentation. Sources of gravity anomalies were investigated by Euler deconvolution and source point clusters in depths of 10 km and 30 km were obtained. For the first time a 3D density model up to the Central American lithosphere has been compiled by combining the results of curvature and Euler analysis with constraining data e.g. geological maps, seismic profiles, earthquake hypocenters and new results from tomographic modeling and receiver function analysis of the seismological task group of the SFB 574. The in-house software package IGMAS was used for modeling visualization of the model structures and gravity effects (e.g. serpentinization of the oceanic lithosphere at the Pacific side); it helps to identify borders between tectonic blocks e.g. the Chortis block in the north or the Chorotega block in the south of the research area. At a more local scale our 3D modeling works hand in hand with a small scale 3D modeling by Lücke and Alvarado and provides insight into the upper crustal parts beneath the volcanoes of Central Costa Rica.