An Empirical Approach to Predicting Effects of Climate Change on Stream Water Chemistry

Abstract:

Climate change may affect stream solute concentrations by three mechanisms: dilution associated with increased precipitation, evaporative concentration associated with increased temperature, and changes in solute inputs associated with changes in climate-driven weathering. We developed empirical models predicting base-flow water chemistry from watershed geology, soils, and climate for 1975 individual stream sites across the conterminous USA. We then predicted future solute concentrations (2065 and 2099) by applying down-scaled global climate model predictions to these models. The electrical conductivity model (EC, model R² = 0.78) predicted mean increases in EC of 19 µS/cm by 2065 and 40 µS/cm by 2099. However predicted responses for individual streams ranged from a 43% decrease to a 4x increase. Streams with the greatest predicted decreases occurred in the southern Rocky Mountains and Mid-West, whereas southern California and Sierra Nevada streams showed the greatest increases. Generally, streams in dry areas underlain by non-calcareous rocks were predicted to be the most vulnerable to increases in EC associated with climate change. Predicted changes in other water chemistry parameters (e.g., Acid Neutralization Capacity (ANC), SO₄, and Ca) were similar to EC, although the magnitude of ANC and SO₄ change was greater. Predicted changes in ANC and SO₄ are in general agreement with those changes already observed in seven locations with long term records.