Abstract

Patterns of water flow paths and time scales are important for nearly all environmental processes within in the Critical Zone. To better understand these hydrological processes, a water stable isotope network was established at the Susquehanna-Shale Hills Critical Zone Observatory to determine spatial and temporal dynamics of the hydrologic pools (precipitation, soil water, groundwater, and stream water) within the catchment. Precipitation samples were collected automatically on an event basis in a clearing at the ridge top. Soil water was collected every two weeks from four distinct transects at varying depths using suction cup lysimeters. Groundwater at two locations in the stream riparian and stream water at the outlet were collected daily using automatic samplers. Groundwater was also sampled every two weeks at 18 spatially distributed wells throughout the catchment. Isotopic analysis was performed using LGR Isotope Analyzer following IAEA guidelines. Results demonstrated the strong seasonality of precipitation isotope compositions and relative stationarity of groundwater isotopic compositions around the annual amount weighted isotope composition of precipitation suggesting groundwater is recharged by precipitation from each season, but that recharge mechanisms appear to differ during the year.  Results strongly demonstrate the ability of the soil profile to attenuate the seasonal isotopic composition of the input to a constant composition at a depth of 1.5 m suggesting the importance of hydrodynamic mixing of precipitation from different seasons. Spatial patterns of soil water isotope profiles showed asymmetric snow melt dynamics between the north and south slopes.  Investigations of standard deviations of seasonal isotope profiles provided evidence of lateral preferential flow along soil horizon and soil-bedrock interfaces during the cold season and vertically through macropores during the warm season.

Investigation of the temporal dynamics of isotopic composition of precipitation yielded interesting results with respect to the influence of precipitation amounts and type on expected frequencies as well as the local meteoric water line. A test case of a small subset of the precipitation record showed that incorporation of precipitation amounts to one-dimensional and two-dimensional kernel density estimates shifted the distribution substantially. Full record unweighted and weighted kernel density estimates revealed that isotope compositions of precipitation were not symmetrical but skewed towards more depleted values for the four year monitoring period. Monthly weighted kernel density estimates showed the importance of snow and tropical storm isotopic composition imposing a seasonal variation to the precipitation record.  Time integration of precipitation isotope compositions using an amount weighing procedure from event to seasonally amount weighted isotope compositions reduced the variability within the record yet preserved the seasonal cycle. Construction of local meteoric water lines using event, daily, weekly, monthly, and seasonally amount weighted isotope composition time series demonstrated the significance of incorporation of precipitation amounts and averaging, with event and daily local meteoric water lines being statistically different from a local meteoric water line based on a precipitation amount weighted least squares regression. These differences in local meteoric water lines become important when investigations of hydrologic pool interactions or other moderate to long-term hydrologic processes are in question.

Utilizing the knowledge of earlier investigations, a robust comparison of two isotope incorporated atmospheric general circulation models against SSHCZO observations was performed with the goal of developing a fully distributed high-resolution data product predicting isotope compositions of precipitation for the Chesapeake Bay. Comparisons were performed on daily, weekly, and monthly amount weighted values as well as monthly values from the Online Isotopes in Precipitation Calculator. Linear regression results showed the best agreement between the monthly amount weighted values yet Nashe-Sutcliffe coefficients were only 0.2807 and 0.5792 for the global GCM and regional GCM respectively. To determine the temporal structure of the time series, singular spectrum analysis (SSA) was performed on both GCM models and observations from SSHCZO. SSA results demonstrated the importance of the annual cycle and its harmonics. Reconstruction of weekly amount weighted isotope compositions for the regional model and SSHCZO recovered 43.44% and 51.78% amounts of the variance respectively suggesting much of the records contain inherent noise.

Citation

Thomas, E (2013): Spatial and Temporal Patterns of Water Stable Isotope Compositions at the Susquehanna-Shale Hills Critical Zone Observatory . Master of Science, Civil Engineering, The Pennsylvania State University, p.118.

This Paper/Book acknowledges NSF CZO grant support.