DEPTH- AND TIME-RESOLVED TEMPERATURE OF THE VADOSE ZONE AND A SHALLOW ALLUVIAL AQUIFER BENEATH A COLORADO RIVER FLOODPLAIN

Shallow alluvial aquifers beneath floodplains are thought to play an important role in biogeochemical cycling of detrital and dissolved organic carbon. Subsurface temperature (T) is a major factor affecting cycling and export of carbon from floodplains through influencing biogeochemical, microbiological, and hydrological processes. In this study we seek to understand controls on subsurface T such that we can predict T under conditions of future climate as an input to models of floodplain biogeochemical processes. Approximately one year of data from depths ranging 0.2 to 6.0 m in a floodplain along the Colorado River in Rifle, CO show that the vadose zone at a depth of 0.75 m varies by ~24˚C annually. The T of the upper 1.5 m of the system is sensitive to a combination of infiltration of meteoric water and air temperature changes. Data from a deuterium-amended artificial snow melt experiment demonstrate that, after 6 months, little or no water from the artificial snow crossed the contact between fine-grained, locally-derived loess fill and underlying alluvial sandy gravel. Extreme precipitation or flooding events could exceed the water retention capacity of the fill material, but for the most part at this site, direct T impacts from infiltration are limited to the depth of this contact.

The T of groundwater near the water table varies by ~10˚C annually whereas the deepest regions of the aquifer (~7 m) vary in T by ~5˚C annually. Below the average depth to the water table (~ 3 to 4 m) T variations are damped versions of seasonal sinusoidal T patterns with time. In areas close to the upland edge of the flood plain, slight deviations in the sinusoidal pattern suggest a T impact of horizontal groundwater flow from an upland recharge area. Overall the results demonstrate the value of spatially and temporally dense temperature measurements and the need for non-isothermal models of shallow subsurface biogeochemical reaction networks to accurately capture biogeochemical reaction rates in the subsurface and to appropriately estimate saturated and unsaturated hydraulic parameters.