Date of Award

Summer 7-16-2021

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Civil and Environmental Engineering

Advisor(s)

Driscoll, Charles T.

Second Advisor

Gao, Peng

Keywords

Acidic Deposition, Adirondack, Climate Change, Critical Load, Modeling, Sulfur Nitrogen

Subject Categories

Biogeochemistry | Civil and Environmental Engineering | Earth Sciences | Engineering | Physical Sciences and Mathematics

Abstract

Despite decreases in acidic deposition since the 1970s, the recovery of surface waters from acidification has been limited primarily due to the depletion of exchangeable base cations, net mineralization of organic sulfur and nitrogen and release of previously retained SO42- and NO3-, and increases in concentrations of naturally occurring organic acids from soil. The future recovery of stream chemistry from acidic deposition may be altered by projected increases in temperature and precipitation associated with a changing climate. The goals of this study were to conduct a modeling analysis of the response of soils and streams in the Adirondack Park, New York, USA to future changes in acidic deposition and climate. I conducted the research for this dissertation in three phases. In phase one, the integrated biogeochemical model PnET-BGC was applied to 25 forested watersheds that represent a range of acid sensitivity in the Adirondack region to simulate the response of streams to past and future changes in atmospheric S and N deposition, and to calculate the target loads of acidity for protecting and restoring stream water quality and ecosystem health. Using measured data, the model was calibrated and applied to simulate soil and stream chemistry at all study sites. Model hindcasts indicate that historically, stream water chemistry in the Adirondacks was variable, but inherently sensitive to acid deposition. Model projections suggest that simultaneous decreases in sulfate, nitrate and ammonium deposition are more effective in restoring stream ANC than individual decreases in sulfur or nitrogen species in deposition. However, the increases in stream ANC per unit equivalent decrease in S deposition is greater than for equivalent decreases in N deposition. Using empirical algorithms, fish community density and biomass are projected to increase under several deposition-control scenarios that coincide with increases in stream ANC. However, model projections suggest that even under the most aggressive deposition-reduction scenarios, stream chemistry and fisheries will not fully recover to pre-industrial values by 2200 due to legacy effects of historical acidification. In phase two, the PnET-BGC model was applied to two montane forested watersheds in the Adirondack region to evaluate the effects of future climate change on the recovery of surface waters from historical acidification in response to future changes in atmospheric sulfur and nitrogen deposition. Statistically downscaled climate scenarios, on average, projected warmer temperatures and greater precipitation for the Adirondacks by the end of the century. Model simulations suggest under constant climate, acid-sensitive Buck Creek would gain more acid neutralizing capacity (ANC) than acid insensitive Archer Creek by 2100 from large reductions in acidic deposition. However, climate change could limit those improvements in stream acid-base status. Under climate change, acid-insensitive Archer Creek is projected to experience less of an ANC increase than Buck Creek by 2100. Calculated target loads for 2150 for both sites decreased when future climate change was considered in model simulations, which suggests further reductions in acid deposition may be necessary to restore ecosystem structure and function under a changing climate. In phase three, the "One-at-A-Time (OAT) first-order sensitivity index method and Monte Carlo method were used to analyze the uncertainty in modeling Adirondack stream ANC. The results of first-order sensitivity analysis indicated that in general the model simulations of stream ANC are most sensitive to variation in precipitation quantity, Ca2+ and Na+ weathering rates, maximum monthly air temperature, SO42- wet deposition, and DOC site density (the moles of organic anions per moles of organic carbon). The results of the first-order sensitivity analysis showed that even if the order of the most sensitive parameters between different research sites were consistent, there were differences in projected uncertainty of stream ANC among sites. Monte Carlo analysis was conducted under the assumption of a 30% interval uncertainty (± 15%) in 16 input factors for 500 simulations that were normally distributed around the original simulated stream ANC for year 2050. The Monte Carlo analysis indicated that the model simulation of ANC is most sensitive to precipitation quantity, Ca2+ weathering rate, Na+ weathering rate, SO42- wet deposition, and maximum monthly air temperature. Future simulations could be improved with further research to improve characterization of these inputs.

Access

Open Access

Share

COinS