Interactions Between Riparian Evapotranspiration and Streambed Infiltration As Expressed in Shallow Alluvial Aquifer Levels

Abstract:

Rivers in semiarid regions are typically ‘ribbons of green’ through otherwise sparsely vegetated landscapes. As these rivers convey water from sometimes distant sources they lose water to shallow riparian aquifers which frequently support a rich phreatophytic vegetation community. In order to monitor the interactions between streamflow, alluvial aquifers, and riparian vegetation, water levels in 3 monitoring wells near the Rio Grande at Escondida, NM, were recorded beginning in May 2013. Two summers of data reveal a strong diel signal in the water table levels, indicating a clear interaction between vegetation transpiration and aquifer storage. Likewise, comparison with stage data from a USGS gage 25 m upstream of the well transect shows a clear aquifer recharge signal. Prior work at this site during fall 2005 has suggested that streambed infiltration rates vary in a daily cycle, possibly due to temperature controls on water viscosity. In order to evaluate the relative roles of the two potential cycles we also monitored near-bed hydraulic gradients and remotely sensed vegetation indices (VIs). To measure hydraulic gradient we installed an additional piezometer in the bed of the river, screened 3 feet below the sediment surface, and began logging head in December 2013. The Enhanced Vegetation Index (EVI) was calculated from Landsat scenes that had clear views of the nearby riparian forest during the monitoring period. Preliminary results indicate that periods with high EVI and high evaporative demand (as estimated from a nearby weather station) correlate with greater water level cycle amplitude and, presumably, vegetation-aquifer interaction. We did not observe any diel signals in the water table data during periods with low leaf cover, casting some doubt on the importance of the previously reported streambed infiltration rate driven cyclicity, however analysis of the near-bed hydraulic gradient data is ongoing and will be compared with residuals of the EVI-cycle amplitude relationship. Additionally, ongoing soil and pump test analyses are expected to give insight into the specific yield of the aquifer, which may enable ET rate estimation for comparison with VI-based methods.