At each location and sampling time, we collected cores in triplicate to calculate an average and examine the range of variability for a given location at a given time. At a site, three cores were collected in a 2 m equilateral triangle. For the locations with repeat sampling in the same season, cores were collected in the same location with the triangle rotated in order to collect cores adjacent to those of the previous sampling. All coring was done with a SIPRE hand-auger. Cores are 79 mm (3 in) in diameter and varied in length depending on ice quality. We attempted to collect ice up to 1 m, but were sometimes limited by ice temperature and density differences resulting in a crumbly, unusable core or auger frozen and stuck in the glacier. Upon drilling and extracting a core, it was laid out in a plexiglass lined tray were it was measured and unique characteristics were noted. All cores were handled with clean nitrile gloves and the tray was wiped clean between cores. If a core broke into sections, the section lengths were noted, and sections were bagged individually in clean whirlpacks. All cores were stored in coolers and kept frozen for transport off the glacier where they were immediately placed in a -20º C freezer to be stored at the field camp or at McMurdo Station.

All cores were processed at McMurdo Station. In the lab, cores were sectioned into 5 cm increments from 0 to 25 cm and 25 cm increments from 25 cm to end of the core. Cores were cut on a band saw in a -20º C freezer. Before cutting ice cores each day, several cuts of frozen deionized water (DIW) “ice cores” were made to clean the band saw blade. The ice from the last cuts of the DIW core was saved, melted, and analyzed for ions to evaluate the cleanliness of the saw. Cores were cut in sections and heat sealed into bags made from polyethylene tubing. The cores were then allowed to melt in the sealed bags at room temperature overnight.

The sediment in the core section settled to the bottom of the bags when the ice was fully melted. Water was carefully poured off the top of the bag leaving the sediment settled at the bottom into one 20 ml glass scintillation vial and two triple DI-washed 60 ml polyethylene bottles. Meltwater and all sediment was then filtered through a pre-weighed GF/F filter. The bags were thoroughly rinsed with DIW in order to ensure all sediment was captured on the filter. In the rare cases of very high sediment loads in a core section (several grams), the sediment was scooped into a pre-weighed aluminum weigh boat before the remaining sediment was washed onto a filter. Total weight was tracked for the scooped and filtered sediment to calculate sediment concentrations for that core section. Filters were placed in a weigh boat and dried overnight at 55º C. Filters and associated sediment were weighed on an analytical balance (± 0.1 mg). Sediment and filters were then archived by storing them in individual sealed 50 mm plastic petri dishes. Sediment mass was calculated by subtracting the filter weight from the total weight.

The scintillation vials was stored at 4º C and all polyethylene bottles were frozen at -20º C. All samples were shipped back to Boulder, Colorado. In some seasons, samples were analyzed by ion chromatography in McMurdo Station for both anions and cations. In other seasons analysis was done by analyzed by ion chromatography for Cl^-, SO₄^-, and NO₃^-. Atomic absorption spectroscopy was used to analyze samples for Ca²⁺ Na⁺ K⁺ Mg²⁺ and by Latchet for NH₄⁺ in the Arikaree Lab at the University of Colorado Boulder.

Samples were analyzed for water isotopes using cavity ring-down spectroscopy on a Picarro in the Barnard Lab at the University of Colorado Boulder.