Lead Investigators:

Thomas Harmon (UC Merced), Jenny Jay (UCLA), Deborah Estrin (UCLA)

Overview

Describing flow and constituent transport within river systems requires estimates of mass transfer across the rivers surface water – groundwater interface. For any particular stretch of river at any given time the question may be asked; does groundwater feed into the river or does the river recharge groundwater supplies? Calculation of mass flux across the interface is typically made more difficult by a lack of data describing groundwater movement within the vicinity of the river. The goal of this project is improve upon existing estimates of mass fluxes across the surface water – groundwater boundary of the San Joaquin River.

Approach

Calculation of the mass flux is based on observations of river stage, water table elevations, and soil samples within the surrounding groundwater system.  State and local agencies provide river stage data through a series of gauging stations located along the San Joaquin River.  Groundwater monitoring is achieved by groundwater data collection deployments along the rivers edge.  A deployment consists of installing a cluster of piezometers (referred to here as pressure javelins) and chemical observation wells (referred to as chemical javelins) to monitor spatiotemporal patterns in hydraulic pressure specific dissolved chemicals (nitrate and ammonium). 

A pressure javelin is constructed from 1.25 inch PVC tubing approximately 2.75 m in length.  The tubing is slotted at its base to allow water entry and instrumented with a pressure transducer to measure the pressure head or water level within the tube.  Figure 1 gives a schematic drawing of the pressure javelin installation and those variables associated with groundwater monitoring.  Where  is the length of javelin ,  is the spatial location of the javelins top, and  is the pressure head recorded by the pressure transducer.  The hydraulic gradient is calculated as
$17+29i\in \mathbb{C}$                                                                                       (1)
where  is hydraulic head and  is the principle direction of flow.  Using Darcy’s Law the specific discharge, , is expressed as
                                                                                                                            (2)
where  is the hydraulic conductivity determined from laboratory tests of the soil samples collected at the field site.  The mass flux passing through the surface water – groundwater interface is then calculated by multiplying the specific discharge by the product of the surface area of the boundary and the water density or constituent concentration.

Chemical javelins are similar to the pressure javelins, but house ion selective electrodes for nitrate and ammonium detection (DirectIon Electrodes, Sentek Ltd., UK). 

Figure 1:  Schematic drawing of Pressure Javelin setup with groundwater feeding into the river.

Systems/Experiments

In February 2007 a total of eight pressure and twenty-two chemical javelins were deployed along the banks of the San Joaquin River approximately 300 m downstream of the confluence of the San Joaquin and Merced Rivers (Figure 2A).  The chemical javelins included a mix of dataloggers (Onset Computers, U12 Industrial Logger) and wireless data acquisition systems (Crossbow MDA300 interface board equipped with a Mica2 Mote transceiver).  Each Pressure Javelin was instrumented with a HOBO water level data logger shown in Figure 2B.  Nitrate and ammonium sensors were calibrated and installed in the javelins near the pressure javelins.  The entire network of javelins extended roughly 100 m along the bank of the river and approximately 150 m upland through the riparian zone. 

Figure 2:  (A) Pressure and chemical javelin distributed along the San Joaquin River bank instrumented with (B) water level sensors and nitrate or ammonium sensors.

Accomplishments

Water table elevations collected from 02/17/07 1:00 AM to 02/19/2007 6:00 AM are shown in Figure 3.  As expected the water table within the field site is relatively flat with an approximate elevation difference of 165 cm.  The laminar movement of groundwater results in small changes in water table elevations over large spatial areas.  A downward trend is also observed during the duration of the experiment with groundwater response almost identical for each of the pressure transducers.

Figure 3:  Elevation of the water table over time for pressure javelins deployed along the San Joaquin river bank.

A filled-contour plot of water table elevations is shown in Figure 4.  The contour plot was constructed from time averaged elevation values interpolated over the spatial domain of the field site.  A downward gradient is observed with groundwater moving towards the river.  Pressure Javelin P4 shows the lowest water table elevation while P6 shows the highest.

Figure 4:  A filled-contour plot of water table elevations averaged over 02/17/07 1:00 AM to 02/19/2007 6:00 AM.

Table 1 summarizes the hydraulic gradients calculated between each pressure javelin pair.  Gradient values range from a minimum of 0.0001 (relatively flat) between P6 and P3 to a maximum of 0.0254 between P4 and P7 (relatively pronounced).

Table 1:  Hydraulic gradients calculated between pressure javelins.

With respect to the chemical javelins, the nitrate and ammonium sensors demonstrated excellent stability, with 20 of the 22 sensors behaving predictably over the course of the 5-day deployment.  Two sensors failed roughly in the middle of the experiment (cause(s) unknown at the time of this report).  A sampling of the results for nitrate and ammonium are plotted in Figure 5 below. 

Figure 5:  Response of two vertically separated chemical sensors deployed in adjacent javelins probing the subsurface within 1m of the San Joaquin River; nitrate (top) which records decreasing voltage with increasing nitrate concentrations; ammonium (bottom graph) records an increasing voltage with increasing ammonium concentrations (note:  investigator-introduced nitrate and ammonium spikes are marked by the large decrease/increase during the afternoon of 2/15).

The plots suggest a diurnal cycle with a maximum for ammonium and accompanying minimum for nitrate occurring simultaneously each day from roughly 10 AM to 3 PM.  We hypothesize that this cycle may be tied to photosynthetic activity in the river.  While the javelins were clearly deployed in the subsurface adjacent to the river, these particular javelins may have tapped into the river hyporheic zone, which is the name for sediments along a river bottom in which reactive transport process can be tightly coupled to stream dynamics.  It may be the case that these javelins are experiencing effects of photosynthetic activity in the stream, where diurnal cycles such as these have recently been reported.  However, this hypothesis will require further testing due to the complex hydraulics occurring in this domain.

The large nitrate and ammonium pulses introduced to the sensors via tubing inside the javelin were an attempt to test the calibration of the sensors in situ.  Unfortunately, the pulses failed to subside in a timely manner, suggesting that flow in the vicinity was insufficient to adequately flush the pulses from the javelin’s sensor housing.  It is interesting to note, however, that the diurnal cycling phenomena continued to occur despite the artificially elevated nitrate and ammonium concentrations in the vicinity of the sensors.  This result suggests that the nitrate and ammonium transformation may have been occurring in the sediments themselves (as opposed to be transported from the stream).  If this is the case, then the reason for the diurnal cycling is less clear at this point.

Future Directions

The javelin component of the river observation project will continue to examine groundwater-surface water interactions in the San Joaquin River test bed.  Specific tasks in the future will include:

• Developing chemical and groundwater flux estimates from further observations in the field.  A spatial survey of the land surface and rivers bank may help to better understand the influence of local topography on water table elevations. 
• Test the hypothesis regarding redox active nitrogen species and whether or not they are transforming in the hyporheic zone (or simply well-connected with river dynamics)
• Using the NIMS RD and AQ systems to perform mass balances on river segments in order to compare loss (gain) of river flow and nitrogen species to javelin-based flux estimates determined along the reach

People

Faculty:

• Thomas C. Harmon (UC Merced Engineering)
• William Kaiser (UCLA Electrical Engineering)
• Mark Hansen (UCLA Statistics)

Researchers:

• Jason C. Fisher (UC Merced Environmental Systems)
• Alexander A. Rat'ko (UC Merced Environmental Systems)

Staff:

• Yeung Lam (UCLA Electrical Engineering)

Graduate Students:

• Sandra Villamizar; (Merced Environmental Systems Program)
• Henry Pai (Merced Environmental Systems Program)
• Amarjeet Singh (UCLA Electrical Engineering)
• Michael Stealey (UCLA Electrical Engineering)
• Nithya Ramanathan (UCLA Computer Science)

Undergraduates:

• Chris Butler (UC Merced Environmental Engineering)

External Research Partnerships

We are in the second year of a 2-year research project funded in part by the UC Salinity Drainage Program and the California Department of Water Resources (Prop 204) to examine wetland drainage timing effects on moist soil plant production in Central California managed wetlands. Our collaborators on this project are the Grasslands Water District, Dr. Nigel Quinn (Lawrence Berkeley National Lab and UC Merced), and the California Department of Fish and Game.