Journal of Conference Abstracts

Combining Digital Spatial Data with Hydrologic Measurements to Interpret Controls of Stream
Chemistry in Large Watersheds

Yvonne H. Baevsky¹ (yhalpern@usgs.gov), Gregory B. Lawrence¹ (glawrenc@usgs.gov),
David M. Wolock² (dwolock@usgs.gov), Douglas A. Burns¹ (daburns@usgs.gov) &
Peter S. Murdoch¹ (pmurdoch@usgs.gov)

¹ U.S. Geological Survey, 425 Jordan Road, Troy, New York, 12180-8349, U.S.A.

² U.S. Geological Survey, Quail Crest Place, Lawrence, Kansas, 6604, U.S.A.

Investigations of natural processes that affect stream chemistry have generally focused on small watersheds (drainage areas less than 10 km²), however, technological advances in mapping and geographic information systems (GIS) have provided new approaches which have enabled these studies to be done at larger scales. Through the U.S. Geological Survey's Neversink Watershed Study, controls of stream chemistry have recently been investigated in the 166 km² Neversink River Basin, in the Catskill Mountains of New York. Within this undeveloped basin are two subbasins that are drained by the East Branch (drainage area 58 km²) and West Branch (drainage area 85 km²) tributary rivers, which exhibit considerable spatial variability in stream chemistry. Atmospheric deposition of sulphur and nitrogen received by the Neversink Basin are among the highest amounts in the northeastern United States. The objective of this study was to utilise digital spatial data in combination with hydrologic measurements to identify processes that cause spatial variations of stream chemistry in the two subbasins.

Stream-water samples were collected at 5 locations along each branch during low-flow conditions in midsummer. Seep-water samples (ground water discharging at the surface) were collected at 90 locations throughout both subbasins during all four seasons. Chemical analyses of solution samples were done by the methods of Lawrence et al. (1995). Digital spatial data coverages were used to evaluate specific physical, chemical, and biological factors that could be related to the observed spatial differences in stream chemistry. Coverages included topography, hydrography, soils, type of forest cover, land use/land cover and a network of stream and seep sampling sites. Digital elevation models with 10 m² and 30 m² grids were developed from the topography coverage to compute distributed hydrological indices that were used to estimate subsurface-water residence time.

Stream water is most acidic in the headwaters of both the East Branch and West Branch and generally increases in acid-neutralising capacity (ANC) downstream to the confluence of the tributaries. The downstream increase in ANC in both branches is consistent with downstream increases of subsurface-water residence time, derived from a topographic index and soil properties (Wolock et al., in press). Despite similar longitudinal trends in the two branches, pH, Ca concentrations, and ANC of stream water are considerably lower in the East Branch than in the West Branch. The similarity of flow-duration curves and values of subsurface-water residence time for the two subbasins indicate that, although differences in residence time explain the longitudinal trends in stream chemistry, they do not explain the chemical differences between the two tributaries. The two subbasins are underlain by the same bedrock (Devonian age sandstone and conglomerate) and locally derived Pleistocene till, however, the West Branch subbasin was more extensively covered by the most recent glaciation, and as result has younger till deposits than the East Branch subbasin (Rich, 1934).

Analyses of seep water suggest that the chemical differences between the two branches are the result of different weathering rates along subsurface flowpaths. Higher ratios of Ca:Si in seep water in the West Branch than in the East Branch indicate that Ca release from weathering in the West Branch is greatest. An inverse correlation between ratios of Ca:Si and dissolved organic carbon (DOC) concentrations in seep water throughout the basin (P < 0.001), however, indicates that weathering release of Ca is highest where DOC immobilization is greatest. Because DOC immobilization increases with longer flow paths, this correlation suggests that flow path length rather than variations in rock type is controlling weathering rates in both subbasins. If Ca release was a function of minerological variations associated with different rock types, a correlation between Ca:Si ratios and DOC concentrations would be unlikely. The more recent glaciation in the West Branch subbasin exposed mineral surfaces that have had less time to weather than those in the East Branch subbasin. Differences in stream chemistry between the two basins can therefore be attributed to present geochemical differences along subsurface flow paths that are the result of a different weathering history, rather than differences in subsurface residence time or variations in rock types.

References

Lawrence, G.B., Lincoln, T.A.,Horan-Ross, D.A., Olson, M.L.,Waldron, L.A. Open-File Report 95 416 (US Geol. Surv. Troy, New York, 1995).

Rich, J.L. Bulletin 299, New York State Museum, Albany, New York (1934).

Wolock, D.M., Fan, J. & Lawrence, G.B. Hydrol. Proc. In Press.