EXPERIMENT 1

Needless to say we were pretty excited to start this project. I mean, c'mon, we're getting college credit to play in a sandbox!

Hypothesis

As this was our first attempt to use a stream table, we really didn't know what to test, so we decided our initial experiment would be simply: "Let's see what happens if we run water over a box full of sand." We did hypothesize that the discharge flowing over a flat surface would incise a channel.

Figure 1: The stream table set up. A bed of fine-grained quartz sand graded to 0⁰. The cabinet beneath the table contains tools, supplies, and controls for sediment feed and discharge rates. In this photo, the basin is being filled using tubing that can also act as tributaries. Water is pumped from an external catch basin located beneath the table.

Set up (Fig. 1)

Bed fill:	Fine-grained (0.70 mesh) quartz sand
Bed thickness:	3 - 4 cm (estimated)
Bed gradient:	0⁰
Base level:	5 - 6 cm below bed surface (estimated)
Discharge rate:	45 mL/s (estimated)
Sediment feed rate:	NA
Shape of initial channel:	NA
Depth of initial channel:	NA
Width of initial channel:	NA
Discharge stage:	NA
Notes:	· A 2L plastic soda bottle with a hole cut through the side was used to catch the discharge and prevent erosionat the head of the table (Fig. 2). · Base level was set arbitrarily lower than the surface of the sand.
Procedure:	· Discharge was started and allowed to flow uninhibited for the duration of the experiment.

Figure 2: Discharge and sediment feed system with a plastic soda bottle.

Observations

1) Initially, discharge flowed only a few centimeters from its entry point before being absorbed by sand (Fig. 3).

2) With time, a saturation zone developed and expanded until it permeated the entire length and width of the table.

3) Once the sand was saturated, discharge moved across the surface as a laminar sheet.

4) Upon reaching the basin edge, the discharge incised to base level, causing headward erosion (Fig. 4).

5) Assuming a meandering river needs a very low gradient, to counter the headward erosion, base level was raised to be approximately level with the channel mouth (Fig. 5).

Figure 3: Initially, discharge was quickly absorbed by dry sand as it entered the system.

Figure 4: Upon reaching the basin margin, discharge raidly cut to base-level, initiating small channels through headward erosion.

Figure 5: To slow headward erosion, base-level was raised to the elevation of the sand bed surface. This caused an initial, lower-stand delta to cease formation and a new, higher-stand delta to prograde over it. Laminar discharge flowing into the channel formed several smaller tributary channels through continued headward erosion.

Interpretations (Each interpretation is tied by number to an above observation.)

1) Discharge was absorbed, rather than flow across the surface, because of the high porosity and permeability of the sand bed.  Surface flow could not develop until the sand became saturated.

2) Pore space became filled as sand became saturated and the water table rose.

3) Due to uniform saturation throughout the sand bed and because of no initial topography on the bed surface, a channel failed to develop near the head of the table, rather the flow moved as a sheet that fanned out from the discharge entry point.

4) A straight channel did begin to form through headward erosion, beginning where laminar surface flow entered the basin and increased in velocity due to the steeper gradient.

5a) Raising base level did not stop headward erosion but did result in a landward shift in deposition and creation of a flooding surface.

5b) Had we left base level, it is conceivable that continued headward erosion would have extended the channel to the head of the table.  It is uncertain whether or not that channel would ultimately have developed a meandering pattern.



Technical Issues

The table’s base-level control did not function as designed; therefore, part way through the experiment, the original PVC pipe was replaced by a rubber garden hose. A rock was used to hold the hose stationary against the side of the table at the desired base -level height (Fig. 6). Unfortunately, the hose was still not fastened securely, and bumping caused frequent, rapid base level fluctuations. When base level dropped, channels incised into the delta (Figs. 7a, b).

Figure 6: A garden hose was held in place by a rock to try and maintain a constant base level.

Figure 7a: Base level was inadvertently lowered, leading to delta incision.

Figure 7b: Channel incision into the delta plain.