South Florida Holocene coral sea-level database for samples collected from 1977 to 2017

Radiometric dating and screening procedures: Samples in this database have been analyzed by uranium-series (U-series) and radiocarbon dating methods. U-series ages were measured using either thermal ionization mass spectrometry or inductively-coupled plasma mass spectrometry. Ages reported in the original studies were recalculated using the most recent decay constants reported by Cheng and others (2013) for 230Th and 234U and by Jaffey and others (1971) for 238U, where appropriate. U-series data were screened for: open-system behavior indicated by 234U initial values outside of 137–157 per mil (‰) and incorporation of detrital thorium indicated by 232Th values >2 parts per billion (ppb). In addition, researchers screened the 238U data for values outside of the typical ranges of 2800–3800 ppb for acroporid corals and 2000–3200 ppb for massive corals similar to Toth and others (2017). Samples that did not pass the screening procedure have the potential for diagenetic alteration, which could lead to erroneous age calculations. Radiocarbon ages of corals were determined using standard, bulk radiocarbon dating or accelerator mass spectrometry. If no conventional 14C ages or δ13C values were originally reported then USGS scientists corrected 14C ages to conventional 14C ages, using a δ13C value of 0 ± 4 ‰ with the Calib spreadsheet "d13ccorr.xls" (http://calib.org/calib/fractionation.html). Conventional 14C ages were converted to calibrated calendar years before present (where “present” is 1950 Common Era [CE]) using the Marine 13 calibration curve in the CALIB 7.0.2 computer program accessible at, http://calib.org/calib/. The time-varying ΔR (local marine reservoir correction offsets) values calculated by Toth and others (2017; https://doi.org/10.3133/ofr20161074) were used for this calculation.

Characterization of the samples: Every dated sample identified from the literature was first cross-referenced with the USGS Core Archive (https://doi.org/10.5066/F7319TR3) to determine if the core material was physically available for analysis. If samples were not present in the archive, researchers relied on the information and interpretations provided in the original publication (see “Reference” attribute) to evaluate sample quality. Samples that were stored in the archive were visually inspected to identify taphonomic characteristics that aided in determining whether the samples are in situ, and ultimately, whether they are suitable sea-level indicators. The coral samples were assessed for the presence of: basal attachment surfaces (which form when a coral grows and encrusts its substrate), normally-oriented (in other words, vertical) corallites, normally-oriented geopetals (cavities filled with sediment), and presence and/or thickness of intergrown vermetid worms, crustose coralline algae (CCA), and the benthic foraminifer Homotrema rubrum. USGS researchers also described the 50 centimeters (cm) of core material above and below each dated sample to better characterize the section that the sample was emplaced within and to determine the dominant framework composition of that section (in other words, massive vs. A. palmata corals). Furthermore, any obvious discrepancies between observations of the cores and the published literature were corrected and updated in the database (the most common discrepancies were misidentification of coral species). The overall characterization of the samples was used to rank each sample (from 0–3, with 0 being the highest rank) in terms of the quality and reliability of the sea-level data it could provide. Samples were excluded outright (rank = 3) if one of the following criteria were met: a) the core was not available for observation and no explicit mention of in situ characteristics were made in the original publication, b) the sample did not pass U-series screening (either performed in the original publication, or described above), and/or c) the sample did not pass diagenetic screening if performed in the original publication (in other words, X-ray diffraction or Scanning Electron microscope analyses). Samples were also excluded (rank = 2) if researchers could not confidently locate a sample within its section and/or the sample possessed inverted corallites and/or geopetals (indicating that the sample was not in situ). For samples that possessed some ambiguous features (for example, corallite orientation was indeterminate, but possessed a normal geopetal) or were otherwise not clearly definitive, a rank =1 was assigned, whereas those that had clearly distinguished in situ characteristics (for example, normal corallite orientation and no other ambiguous features) were assigned a rank = 0.

Added keywords section with USGS persistent identifier as theme keyword.