Alaska Tephra Data, 2018

All samples were processed at the U.S. Geological Survey Alaska Tephra Laboratory and Data Center, Anchorage, Alaska. Coarse-grained samples (greater than 2 mm in diameter) were hand-selected to make mounts for electron probe microanalysis (EPMA). Finer grained samples were wet sieved into three size fractions mainly to remove fine-grained ash from the surfaces of the coarser material. Material in the size fraction greater than or equal to 0.125 mm was used to make bulk grain mounts for EPMA. Standard polished probe mounts were made by Mann Petrographics, New Mexico.

Sample analysis - Major-element glass analyses were conducted using wavelength dispersive techniques with a 5-spectrometer JEOL 8900 R electron probe microanalyzer at the USGS Electron Microprobe Lab, Menlo Park, California. Concentrations were determined with the CIT-ZAF reduction scheme (Armstrong, 1995). Glass analyses used a 5-micrometer-diameter beam with 5 nA current and 15 kV accelerating potential. Reported glass compositions are the averages of 15-30 spot analyses, or fewer if multiple populations were found within a single sample; background intensities were determined 1-3 times for each grain. Count times were 10 seconds for Na (which was analyzed first to reduce Na-loss), 10 seconds for S and Cl, and 30 seconds for all other elements.

Data Reduction - During analysis, sets of a minimum of 5-10 replicate analyses of glass standards RLS-75 and RLS-132 (rhyolite glasses), VG-2 (basaltic glass), and GSC (synthetic glass)(Jarosewich and others, 1980) were performed to monitor instrument drift. Natural glass and mineral standards were used for calibration: RLS-132 for Si; basaltic glass VG-2 for Fe, Mg,and Ca; Orthoclase 1 for K and Al; Tiburon albite for Na; Mn2O3 for Mn; TiO2 for Ti; sodalite for Cl; and Wilberforce apatite for P.Standard deviations of averages of multiple spot analyses for single unknown samples are generally within those listed for working standards. Once electron microprobe data are acquired and downloaded as raw data in a spreadsheet, they are filtered to to eliminate accidental analyses of minerals, microlites and epoxy. The workflow describing how raw data is filtered for final submission is presented below. All geochemical data acquired by EMPA were subject to rejection base on the following criteria. 1. Analyses totaling <95 percent (bad analysis), 2. Analyses with MgO >15 percent (mafic mineral), 3. Analyses with Al2O3 >20 percent (feldspar), 4. Analyses with FeO >15 percent (mafic mineral). These criteria were used as a first approximation of the data and allowed for the identification and elimination of most micro-mineral analyses. In addition, all sample data were inspected individually to identify other micro-mineral phases. Geochemical mineral identification was based on values given in Deer and others, 1996. Although glass standards were analyzed prior to each analysis session, it was not uncommon to acquire unusable analyses in instances such as: 1) analyzing thin regions were the electron beam passes through the material and into the underlying epoxy or mineral, 2) inadvertently analyzing micro-mineral phases, and/or 3) analyzing near the edge of a shard resulting in analyses of nearby material (minerals or epoxy). Most of these data were identified using the filtering criteria andeliminated from the data set.