Stream-Aquifer Interactions: Its Effects on Nutrient Variability, its Role on Streamflow Permanence and the Effects of Temperature Measurement Resolution on its Quantification

Abstract

In semi-arid river systems, the connectivity and interaction between the stream and its aquifer is an important hydrologic linkage that affects the presence and quantity of streamflow, and the biogeochemical processes occurring in the stream, its streambed and aquifer. The work presented here builds upon and intends to expand on the knowledge base of stream-aquifer interaction by focusing on three main study areas. First, a descriptive and process based study evaluates how changes in basin groundwater (BGW) contribution to streamflow in the San Pedro River translate to changes in its chemical composition (nitrogen, metals and dissolved organic carbon) and proposes mechanisms that could explain the observed relationships. Second, an experimental and process based study evaluates how an empirically derived predictive model can use and predict streamflow permanence (wet/dry areas) on the San Pedro River from a set of geomorphic and hydrologic variables. Third, a methodological investigation of an idealized system that evaluates the effect of temperature measurement resolution on the estimations of stream-aquifer flux using temperature as a tracer. Major findings of this work show that in the San Pedro River: 1) BGW contribution and its variability increases with downstream distance and time since last flooding; 2) the concentrations, variability and spatial dependence of several chemical species are associated and follow similar patterns with BGW contribution; and 3) an empirical model using variables that describe bedrock elevation, the shape and width of the floodplain, the land surface elevation, and late spring streamflow can be used to correctly predict 80-87% of the wet/dry location in the river. In addition, this study found that temperature measurement resolution introduces large estimation errors when temperature amplitude is used as a predictor of stream-aquifer flux, and that it limits the range over which vertical stream-aquifer fluxes can be accurately estimated more than temperature measurement, placement and thermal parameter uncertainty.