This project titled Innovative Methods for Characterizing Natural Riparian Discharge, focused on the analysis of the hydroecology of meadow systems in the northern Sierra Nevada of California. Land-use practices, particularly grazing and logging, have caused stream incision that results in a loss of hydrologic function in these meadow systems. To counteract these trends, landscape-scale restoration has been initiated in the watershed by a local nonprofit group, the Feather River Coordinated Resource Management Group FRCRM. We developed remote sensing techniques to help characterize, monitor, and model the hydrologic processes that occur in both the degraded and restored systems. We employed a combination of field work, high resolution remote sensing, and numerical modeling to help understand the spatially variable processes of baseflow and evapotranspiration. Since intensive monitoring can easily double the cost of restoration projects while still inadequately capturing the spatial variability, the objective of this research is to develop a transferable, cost-effective methodology for assessing restoration initiatives.
It is demonstrated that very high resolution, (~20cm), airborne thermal remote sensing can be used to quantify both hyporheic exchange and baseflow contribution to streams. During warm summer months, particularly during the afternoon, stream temperature is significantly warmer than the groundwater temperature, which is similar to the mean annual air temperature. Therefore, inflowing groundwater tends to have a cooling effect in stream reaches that are strongly gaining. Hyporheic fluxes buffer stream temperature fluctuations resulting in lower daily maximum stream temperature, higher daily minimum stream temperature, and delayed occurrence of these temperature extremes. Stream temperature is simulated using an energy budget model approach and a one-dimensional model of heat transport in the stream. Spatially distributed baseflow and hyporheic exchange are determined by varying these quantities until agreement (root mean square difference = 1.1 ºC) is found between the simulated stream temperature and both the remotely sensed profiles of stream temperature (0.6 ºC accuracy) and the recorded in situ temperature histories (0.1 ºC accuracy).
This is the first time baseflow and hyporheic exchange have been quantified using remotely sensed temperatures. For small watersheds with significant baseflow components, this technique provides a means to evaluate the spatial distribution of baseflow to an extent not practical or possible with commonly available methods. This technique showed both increased baseflow and hyporheic exchange in streams running through a restored subreach during late spring. This finding supports the hypothesis that restoration improves fish habitat in the stream by creating a more hospitable temperature regime.
The theoretical framework for an ET mapping algorithm is presented along with application of this methodology in two restored and two degraded meadow systems. The algorithm takes advantage of high resolution (FLIR) thermography data and local weather station data to partition the measured available heat flux into the sensible and latent heat components of energy budget in a spatially distributed manner. The sensible heat flux is driven by the difference between the land surface temperature and the air temperature; therefore, FLIR imagery and weather station data can be used to estimate this component. The latent heat flux is directly related to ET and can be calculated as the residual of the energy budget.
The method performed with an estimated accuracy of ±10% for this application and holds potential as a routine monitoring technique for riparian systems. For the first time, the practitioners conducting the restoration (the Feather River Coordinated Resource Management Group) were given a quantitative estimate of changes to the evapotranspiration component of the water budget as a result of restoration efforts. As a result of restoration, this component nearly doubles from about 1.5-4 mm/day xeric vegetation, which dominates degraded meadows (dryland grasses and sagebrush) to 5-6.5 mm/day for healthy wet meadow vegetation (sedges and rushes). Very high resolution data sets (1 m) such as this one will be critical for hydroecological studies by elucidating patterns in vegetative water use and serving as a basis for evaluation of the effectiveness of riparian restoration.
Nearly three quarters of a century ago, White  introduced a method for estimating ET based on analysis of diurnal water table fluctuations induced by transpirative uptake by riparian vegetation. Since that time, the method has been used sporadically, with no consensus regarding the reliability of this approach. We present a rigorous analysis of the White method using variably saturated groundwater flow modeling. Results suggest that the method is reliable for coarse-grained sediments, but can be expected to fail miserably when applied in areas with fine grained soils. The method fails because less water is released from storage as the water table drops than would be expected using a traditional definition of specific yield since, 1) in fine grained soils water is released slowly from the vadose zone to the saturated zone and 2) when the water table is very shallow, the vadose zone is too thin to release the volume of water expected using the traditional specific yield model for a given change in the position of the water table. To address these two issues, a method was introduced to estimate the readily available specific yield, which is appropriate for time scales associated with diurnal fluctuations. In addition, a distinction between total ET and ETG (the component of ET from groundwater) was introduced because only the latter is detected with this method.
This is a very practical and inexpensive method that can be easily implemented to quantify ETG in environments with phreatophytic vegetation and a shallow water table. The method results in a continuous record of ETG over the growing season for the meadow vegetation.
We present a modeling framework to understand and predict the hydrologic and ecologic differences between pristine, degraded, and restored meadow systems. Based on observed well hydrographs, the concept of a threshold vegetation hydrograph is introduced to distinguish between mesic and xeric vegetation communities using time-dependent, depth-to-water requirements. Three-dimensional, variably-saturated groundwater flow is simulated in an archetypical meadow system under conditions ranging from pristine to degraded to restored. Vegetation patterning is predicted by modeling the depth to the water table and applying the vegetation threshold hydrograph. Through an iterative process, we simulate feedback mechanisms between changing vegetation communities, evapotranspiration, and groundwater flow.
This is apparently the first demonstrated approach in which the groundwater flow system in a riparian meadow system has been modeled and coupled to a vegetation model to predict vegetation patterning. It is a potentially useful approach for designing and improving the success of riparian restoration efforts in semi-arid environments and groundwater dependent ecosystems. The methodology also holds potential for helping to understand and predict feedback mechanisms between the hydrology, vegetation communities, and atmospheric water exchanges as driven by global climate change.