V33B-1464 1340h
Eruptive History of Arenal Volcano, Costa Rica
Tephra-stratigraphy, volcanic history, eruption types, eruptive dynamics and area distribution of tephra from the most important eruptions of Arenal volcano, are reviewed and updated. Deposits of explosive eruptions are named AR-1 to AR-22 (from older to younger). All previous and new Arenal and neighbouring Chato volcanoes' radiocarbon dates were calibrated: the last Chato eruption occurred 3720$\pm$150 B.P and the first known eruption of Arenal (AR-1), 7010+170-130 B.P. Isopachs and characteristics of the most relevant and recognized airfall deposits are presented. Area distributions of the key layers are oriented toward W, WSW and SW, except the lowest layer of AR-17, which is distributed northward. According to lithic isopleths, an eruption column height of 23 km (subplinian) was calculated for AR-20. Most fall layers in Arenal show associated pyroclastic flow deposits up to a distance of 6.5 km from the summit. They are restricted to river valley paths. Among Arenal eruptions, plinian-subplinian type events like AR-20, AR-15, AR-12 and AR-9, are the most violent and destructive, with tephra volumes between 0.2 and 1 km3. There are also violent strombolian eruptions which were sustained for a considerable time, resembling subplinian eruptions (the AR-19 case, total volume 0.45 km3), and pelean type eruptions, as AR-22, which erupted 0.026 km3 of tephra. A temporal correlation of explosive events after AR-8 shows that the four most important dacitic eruptions (AR-20, 15, 12 and 9), are separated by an average period of 800 years ("long term"). Three of them (AR-20, 15 and 9) have been preceded by two important explosive events with a more basic chemistry, which occurred around 300 years or less previously to the dacitic eruption ("short term" periods). Nevertheless, between cycles AR-13/14/15 and AR-18/19/20, there are two events (AR-16 and 17) that do not belong to any cycle. In fact, between the major eruptions AR-15 and 20, the highest eruptive frequency of all volcano history occurred, which notes that "cyclicity" is relative. The longest periods of explosive inactivity follow the dacitic events AR-9, 12 and 20. The most recent lava cycles have been temporally well determined, erupting volumes of 0.5 - 1 km3, following the most destructive plinian-subplinian eruptions. Lava cycles older than 1300 years B.P. have not been detailed because of their poor exposures. The tephra sequence previous to AR-9 ($\sim$3300 B.P. - 7010 B.P.) has no dark soils, while the overlying sequence ($\sim$3200 B.P. - present) does. It is presumed to be a consequence of regional climatic changes from a dry environment to a rainy tropical one around 3300 years B.P.
V33B-1465 INVITED 1340h
Decompression Induced Crystallization of Basaltic Andesite Magma: Constraints on the Eruption of Arenal Volcano, Costa Rica.
Arenal Volcano is a small stratovolcano located 90 km NW of San Jose, Costa Rica. In 1968 current activity began with a Plinian phase, and has continued to erupt lava flows and pyroclastic flows intermittently since. Samples from the Plinian, pyroclastic flow, strombolian, and effusive phases have been studied texturally. Little variation in crystallinity occurs amongst the different phases. Number density of crystals, both 2D and 3D are 50-70 mm$^{-2}$ and 30,000-50,000 mm$^{-3}$ in the Plinian sample, compared to the lesser values in other eruptive types. Characteristic crystal size also increases as explosivity decreases. Two samples, both lava flows collected while warm, overlap with the Plinian sample. This suggests that the variations seen may be a result of cooling history. Plagioclase differs between the Plinian sample, in which they are only tabular in shape, and the other eruptive types, which contain both tabular and equant crystals. To link decompression paths of the Arenal magma to possible pre-eruptive conditions, we have carried out hydrothermal experiments. The experiments were preformed in TZM pressure vessels buffered at a fugacity of Ni-NiO and water saturation. Phase equilibria results in conjunction with mineral compositions and temperature estimates by previous workers from active lava flows and two-pyroxene geothermometry, constrain the likely pre-eruptive conditions for the Arenal magma to 950-1040$\deg$C with a water pressure of 50-80 MPa. Samples that started from conditions that bracket our estimated pre-eruptive conditions were decompressed in steps of 5-30 MPa and held for various times at each step until 20 MPa was reached, approximating average decompression rates of 0.25, 0.025, 0.0013 MPa/s. Comparison of textures found in the natural samples to the experimentally produced textures suggest that the Plinian eruption likely was fed by magma ascending at 0.05-1 m/s, whereas the less explosive phases were fed by magma ascending at 0.05 m/s or less.
V33B-1466 1340h
The Times Scale of Andesite Differentiation at Arenal Volcano, Costa Rica (1968-2003), Indicated by U-Th-Ra Disequilibria.
Arenal Volcano in Costa Rica has continuously erupted since 1968 evolving in a complex fashion involving crystal fractionation, magma mixing, degassing, and wall-rock interaction (Reagan et al., 1987; Cigolini, 1998). We have measured trace-element concentrations and U-series disequilibria in whole rocks and mineral separates (pyroxene, plagioclase, magnetite) over the course of the eruption from 1968 to 2003 by ICP-MS, TIMS and PIMMS techniques. Whole rock and mineral separate analyses (n$>$20) show only minor variation in ($^{230}$Th)/($^{232}$Th) (1.10 to 1.18). In contrast, ($^{230}$Th)/($^{238}$U) range from 0.91 to 1.04 reflecting the moderate spread in Th/U. Th/U consistently change from low values in the early samples (2.4) to a constant, higher value from 1986 to present (2.65). The observed U-Th disequilibria are consistent with a time scale of mineral formation being much less than that of $^{230}$Th decay ($<$10$^{4}$ years -they are not old xenocrysts) for all samples in the study. A further implication is that mantle-derived recharge magmas entering the Arenal magma chamber either have changed through time from more U-enriched to less U-enriched, or that assimilation has changed the Th/U of the system but not ($^{230}$Th)/($^{232}$Th) (but the assimilant has to be very young Arenal cumulates or fortuitously have the same ($^{230}$Th)/($^{232}$Th)). Ra isotope data are being analyzed to evaluate whether differentiation and mineral formation in this system operate on a time scale closer to the half life of Ra (1600 years).
V33B-1467 1340h
Detection of SO2, HCl and CO2 in Arenal Volcano Eruptive Plume Using MASTER Multispectral Images
The Costa Rica Airborne Research and Technology Applications (CARTA) Mission developed in March-April, 2003 was a join effort between the National Program for Airtransported Missions of the National Center for High Technology of Costa Rica formed by the four public universities in Costa Rica, and NASA. This mission took aerial infrared photography and multiespectral images using the MASTER sensor of 70% of the national territory of Costa Rica. Multiespectral images were taken from Arenal volcano at high (13780 m) and at low (3450 m) altitude. The MASTER images have an aperture ranging between 0.44 micrometers in the visible and 13 micrometers in the thermic infrared. In addition, the distribution of the 50 channels of the MASTER sensor, have been arranged to avoid the influence of water vapor, always present in large quantities in the atmosphere and also in volcanic gases. We determined that SO2 is clearly visible between 8.5-9.3 micrometers (LWIR), and CO2, H2S and HCl in a smaller bandwidth 3.5-4.4 micrometers (MWIR). Another gas detected at Arenal volcanic plumes is methane, in a bandwidth between 7.7-8.1 micrometers (LWIR). When both multispectral images were taken, Arenal volcano had an active lava flow descending its NE flank and no significative winds were blowing, so the eruptive plume was rising almost vertically from the active vent. Profiles of gas concentration collected from the above mentioned bands were performed on the image using the software ENVI to detect different species present in volcanic gases. The concentration of volcanic gases in the multiespectral image was largest above the active crater (north vent of Crater C), and lower on the short active lava flow whose blocks were cascading down up to 1 km on the NE flank. Significant amounts of SO2 were measured above the lava flow and the fan of cascading blocks suggesting that the cooling lava continues to release magmatic gases as their cascading blocks move down flank of the volcano. The higher resolution of the low altitude multispectral image assured better results when measuring volcanic gas species.
V33B-1468 1340h
Grainsize, Mass per day and Morphological Characteristics of Tephra from Arenal Volcano, Costa Rica: 1987-2000.
We studied tephra samples from collectors located at various locations on the flanks of Arenal Volcano. They correspond to monthly samples collected between 1987-2000. In addition, we studied samples collected at two sites from the fallout events that occurred July 9, 1987 and February 3, 1996. Glass fragments of the most abundant fractions were examined with a scanning electron microscope to determine the range in clast morphologies and allow discussion on the nature of the fragmentation mechanisms involved in their generation. Tephra fall from single explosions, cumulative fallout was subject to grainsize analyses. Explosive activity clearly diminished in 1997 and remained low between 1998 and 2000. Isomass maps were calculated for a number of months across the reporting period. Assuming eruption frequencies of 20 - 40 explosions day -1 for the majority of the reporting period, a mean explosion volume (fraction < 2 mm only) can be estimated. This data yield total tephra masses of between 1 and 10 m3 per explosion. These volumes do not, however, consider the coarser tephra fraction, including ballistics. The grain size of the cumulative monthly tephra varies relative to the location within the dispersal axis. Generally more distal and peripheral samples (those off-axis) are finer grained and more poorly sorted than those along axis. Samples of monthly tephra collected 2 km west of the vent (Caseta C) are consistently coarser grained and better sorted than those from 2.9 km west-south west (Salida del Bosque) of the vent. In addition 85% of the samples from the Salida del Bosque site show a subtle bimodality with one strong mode at +2 phi and another much weaker mode at +6phi. Less than 30% of the samples from Caseta C show such bimodality. Tephra, both from cumulative monthly samples and single explosions, was separated into crystals, lithics (altered grains) and glassy fractions and fragments which were composed of approximately equal proportions of crystals and glass. The scanning electron microscopy allowed subdivision of the glassy fraction into one of four types: vesicular, fluidal, blocky and clasts with both blocky and fluidal portions. The results showed that the blocky type form the most abundant clast morphology forming between 50 and 71%. Part fluidal and part blocky types form between 21 and 40%. Wholly fluidal types form up to 6% and vesicular up to 5 %. Proportions of different clast types for single explosions were similar for cumulative monthly samples. The different clast morphologies are thought to represent the different nature of the state of the magma on fragmentation. Therefore vesicular clasts represent vesicular, degassing magma. Fluidal clasts represent wholly molten fragments of magma. Blocky clasts are interpreted as being cool, rigid parts of the magma that had degassed and part blocky- part fluidal represent fluidal, molten magma that was fragmented as the clast cooled during eruption. This would indicate that the fragmentation process was sequential occurring over tens of seconds and not instantaneous.