Much has been pontificated about the source of the Verde River and the recent USGS regional model of groundwater-streamflow interaction. A simple fact is regional models do not always represent headwater springs and Quaternary sediments that are above (younger than) the modeled aquifers like the basin fill of the Big and Little Chino aquifers.
In the case of the upper Verde watershed, many of the once perennial/intermittent tributary streams were and are hydraulically perched above the basin fill aquifers where groundwater is withdrawn by deep wells. It is this deeper groundwater that is modeled above the Paulden gage by the USGS and not the mountain front springs and relatively small alluvial aquifers along stream channels.
These streams were depleted long before deep wells were used in the watershed. Early settlers diverted water for irrigation using low rock dams and shallow wells with centrifugal pumps all along tributary streams like Granite Creek, Walnut Creek, Williamson Valley Creek and Big Chino Creek. Based on original Federal Land Survey plats and field notes, more than 8,000 acres of land along headwater streams were cultivated and irrigated by settlers. This amount of water applied to this acreage corresponds to an early depletion of base flow in the Verde River even before construction of Watson Lake dam.
Other uses of streamflow and springflow by settlers included the powering of grist mills on Granite Creek a short distance above Prescott. In 1867 the Bowers brothers operated a mill that ground about one ton of corn meal per day. Another grist mill was operated at Del Rio Springs in 1871 when springflow was at least 150 miners inches (3.8 cfs) as measured by Federal Land Surveyors.
Del Rio Springs was a landmark known to pioneers as a permanent and plentiful source of water. It is located just beyond the northern limit of the artesian area. In 1863 Whipple Barracks was located at this site and then moved to Prescott. In 1904 the City of Prescott built a pumping station at Del Rio Springs and pumped water from a shallow well for a distance of 21 miles into the city through an 8-inch pipeline. The operation was discontinued in 1926.
A small but significant consideration gleaned from history and grist mills and arrastras is that to be successful, the quantity of flow and the natural uniformity of flow are important, particularly in the summer time. The success of the mill implies natural storage of sufficient size to yield a sufficient uniform quantity of water to power the mill at nearly all times. Therefore, the powering of a grist mill and arrastras on Granite Creek a short distance above Prescott implies a good-steady base runoff before human impacts like diversions and storage.
In the mid to late 1800s a few hundred acres in the vicinity of Prescott was watered by diverted flow from Willow and Granite Creek. Competition for streamflow and shallow subsurface flow in the stream sediment was fierce by the late 1800s. As far back as 1899 and 1900 Prescott was accused of "taking" water used by others. Lawsuits had started and been settled in court as Judge Sloan, for example, had decided against John Duke and in favor of Prescott where Duke claimed Prescott was diverting his water along Granite Creek.
A few months ago the USGS presented the regional groundwater model to the public at a large gathering in Camp Verde. They discussed how the model showed a decrease of about 7 cfs in the base flow of the upper Verde River as a result of withdrawal of water from the Big and Little Chino aquifers. This newspaper reported that Doubting Thomases picked apart every aspect of the model and have even questioned the integrity of USGS employees and river preservation leaders. Some doubted there has been a 7 cfs drop in base flow.
The truth is the USGS was modeling only part of the base flow above the Paulden gage - the part associated with the deep and large basin fill aquifers and other aquifers. The USGS on p. 44 of the model (Pool, D.R., Blasch, K.W., Callegary, J.B., Leake, S.A., and Graser, L.F., 2011, Regional groundwater-flow model of the Redwall-Muav, Coconino, and alluvial basin aquifer systems of northern and central Arizona: U.S. Geological Survey Scientific Investigations Report 2010-5180, 101 p.) clearly says the Quaternary alluvium located above the basin fill of the upper Verde watershed was not simulated.
From my perspective, there was little need for such simulation to satisfy the purpose of the model to simulated hydrologic effects of future groundwater use. The USGS has been clear that important factors are distance between the pumping well(s) and the Verde River and aquifer diffusivity in the upper Verde watershed.
So how much was the total base runoff of the Verde River for virgin conditions before human diversion from the perennial/intermittent tributary streams and stream sediment? Three independent estimates can be made using published USBR information on virgin flow at the mouth of the Verde River, cultivated land shown on Federal Land Survey plats and field notes, typically of the 1870s, and the measured channel width and depth for base flow made by the Federal Land Surveyors. These estimates suggest the natural and ordinary base runoff at the Verde River near Paulden gage was about 50-60 cfs or more than double the present base flow that is in the low to mid 20s. Obviously this is considerably more than the 7 cfs change in base flow from aquifer depletion using deep wells.
It seems that explicit simulation of the shallow aquifer could be a useful addition to the USGS model because this is where most of the action is in the GW flow system. An obvious limiting aspect of adding a layer to the USGS model is always the lack of information about, for example, interaction of the old diversion ditches with the shallow groundwater system.
However, there is always a lack of information for groundwater models. It goes with the territory and one gags when reading the word "estimate" more than 330 times in the 100 page model report of the USGS. So even with the fear of certain local transitory political officials, the USGS is the best to perform such modeling of our precious water resource.
Win Hjalmarson is a retired USGS hydrologist living in Cottonwood.