High-Resolution Glacial Discharge Records From Deep-Water Tidal Rhythmites in an Alaskan Fjord

In this study we have compiled and analyzed two high-resolution records of deep-water tidal rhythmites derived from glacial discharges. The rhythmites contain an average of 1 cm of sediment thickness per week during the melt-season. Two sediment cores over 17-m-long were collected from Muir Inlet, Glacier Bay National Park, Southeast Alaska, aboard the R/V Maurice Ewing in 2004 (EW0408). One core (core 60JC) was collected just north of the mouth of Wachusett Inlet. The other (core 62JC) was collected 3.8 km due north of 60JC in a separate basin more proximal to Muir and McBride Glaciers. In Glacier Bay, glacial retreat since the Little Ice Age Maximum is well constrained by historical mapping of glacial temini, and local climate records exist as well. The next step after establishing glacial discharge records is to test hypotheses relating glacial discharge to temperature and rainfall. The cores were subsampled with ODP-style u-channels and scanned for magnetic susceptibility and bulk density. X-radiographs of u-channels were collected to observe small-scale (42 microns/pixel) density differences between silt and mud laminae. Spring-neap tidal packages, representing 2-week periods during summer, are visible because of closely-spaced bounding silt laminae. Seasonality is marked by winter gravelly mud (diamicton) beds and/or spring plankton blooms. Spring-neap packages and seasonal markers were considered jointly in order to construct two mostly continuous records of spring-neap packages and melt- season deposits in each core. Secondarily, melt-season duration, which we define as the number of spring- neap packages occurring in a melt-season, was determined. We found that core 60JC extends back 83 years. 75 of those years were identified as having complete, contiguous melt-season deposits that are, on average, 17 cm thick and contain 8.3 spring-neap packages. Core 62JC extends back 62 years. 57 of those years were identified as having complete, contiguous melt- season deposits that are, on average, 24 cm thick and contain 8.5 spring-neap packages. The multi-taper method and singular spectrum analysis spectral estimation techniques were used to calculate statistically significant periodicity in both the melt-season deposit thicknesses and spring-neap package thicknesses for each core. Periodicity in melt-season deposit thickness shows that both cores contain significant ENSO signals and that 60JC contains significant PDO signals. Significant periodicity in the thickness of spring-neap packages may reflect dominant sedimentation trends that occur in the average melt-season. 60JC shows significant periods around 4.8 spring-neap packages, supporting a sediment distribution model with bimodal peaks: an early sediment pulse related to spring snowmelt and a later pulse during the peak melt-season. 62JC shows significant periods around 8.1 spring-neap packages, supporting a unimodal distribution model for sediment delivery most likely related to the peak melt-season. Data from high-resolution glacial discharge records may act as reliable proxies for past magnitude and duration of annual meltwater discharge into Alaskan glacial fjords, thus adding to the climate change record in a sensitive area.