What I have learned from my model
From GeoMod
There is a significant amount of water flowing into the river from overland sources. It appears to be more stable in flow rate along the length of the river than the aquifer discharge rate. This at first seems counter-intuitive.
By USGS gauging data: The flow rate increases dramatically and then drops to near 0 cfs halfway across Shelby County. It then increases in rate again further downriver. The only way for this to be possible is if halfway down the river, a massive amount of water is being lost into an aquifer. Since it is not likely that the same aquifer can be both losing and gaining, there must be a separate aquifer unit taking water from the river. At some point this aquifer must terminate since there is another drastic gain in flow rate even further downstream. This means at least two, possibly three separate aquifer units that the Loosahatchie crosses over in Shelby County.
By Model data: The model shows that there is a massive amount of water discharged for the first quarter length of the river. There is then a negative flow rate where water is being lost to the aquifer. The aquifer rate then increases again in the last quarter stretch of the river. The overland flow rate appears to be relatively constant throughout the stretch of the river.
By Concentration Plots: The plots require a bit of interpretation. The peak in overland concentration near the halfway point may not mean a jump in overland flow rate. Since the volume of water at this point is also dropping drastically, the loss of river volume will raise the concentration of these components while input volume remains constant. The aquifer components will drop at this point since no new water is being supplied to the river from the aquifer.
What this tells us is that there are three separate aquifer units separated by thin confining layers. There is a significant amount of water entering the river from storm drains and rain runoff. It seems to be, on average, fairly constant. It may therefore be possible that the spike predicted in previous models at the western edge of the Memphis Sand outcrop may not be an aquifer discharge spike, but instead, an overland point source of overland flow. The problem could be better analyzed with more sampling locations to add higher resolution to the model. It may also be beneficial to try and travel along the Loosahatchie and identify al possible sources of overland flow, such as tributaries and storm drains. What has been conclusively determined is that multiple aquifer units exist in the area and that a significant portion of the water in the Loosahatchie River is derived from overland sources.
