Pumice vesicularity

Floating and washed-up pumices have been reported by scientific expeditions along the Caribbean Sea coast of the Central American Isthmus and the northern coast of South America since at least 1947. Local coastal communities have been utilizing this resource for many years. The rounded and buffered morphology of hand specimens is consistent with water-borne transit. The volcanically active Caribbean and Central American regions provide a number of candidates for source volcanoes and eruptions. We have attempted to identify this source using samples collected from Carrie Bow Cay and Placencia Beach, Belize; Tulum Beach, Mexico; Morrosquillo Bay, Colombia; and Galeta Point, Panama. We have tracked possible transport routes through the use of river drainage and ocean current maps. The criteria for comparing the products of potential source volcanoes (including Atitlán Caldera in Guatemala and Caribbean sources such as Mt. Pelée, Martinique and Soufrière Hills, Montserrat) were developed from the whole rock major and trace element geochemistry and the compositional and textural characteristics of pumice and their constituent minerals and glasses.

Bubble growth in rhyolitic melts is a primary control on some of the largest explosive eruptions, but growth dynamics remain controversial. We have used hot-stage microscopy to directly observe vesiculation of a Chaiten rhyolite melt (containing ~1.38 wt. % H 2O) at atmospheric pressure. Thin wafers of obsidian were held from 5 minutes up to 2 days in the hot-stage at temperatures between 575 oC and 875 oC. The growth of many individual bubbles was measured using image tracking code within MATLAB. We found that bubble growth rates increased with both temperature and bubble size. The average growth rate at the highest temperature of 875 oC is ~1.27 Î_m s-1, compared with the lowest observed growth rate of ~0.02 Î_m s-1 at 725 oC; below this temperature no growth was observed. Average growth rate V r follows an exponential relationship with temperature and melt viscosity where V r ~ exp (0.0169T) and V r ~ exp (-1.202η). Comparison of these measured rates with existing bubble growth models (e.g. Navon, Proussevitch and Sahagian) indicates slower growth than expected at the highest temperatures. The extent of diffusive degassing of H 2O and OH- from wafer surfaces during experiments was estimated with simple diffusion models. It was found to be negligible during brief high-temperature experiments but became increasingly important for slower, lower-temperature experiments. Several stages of bubble growth were directly observed, including initial relaxation of deformed existing bubbles into spheres, extensive growth of spheres, and, at higher temperatures, close packing and foam formation. An advantage of the imaging techniques used here is that bubble-bubble interactions can be observed in-situ at relatively high resolution. Bubble deformation due to bubble-bubble interaction and coalescence was observed in most experiments. Evolving bubble number densities (BND) with time were determined, allowing nucleation rates to be estimated. Maximum observed BNDs were 3.4 x 1012 m-3 with maximum increases of around 143 % observed in samples with a lower initial vesicularity <5.7 x 1011 m-3. The experiments described can be used to effectively retrace the vesiculation history of samples, providing a useful tool for aiding in the interpretation of end member products.