This project developed tools for managing groundwater quality and quantity in the municipal well recharge area for Stevens Point, Whiting, and Plover. The recharge area includes parts of the three municipalities and parts of the Towns of Hull, Plover, Stockton, and Sharon. The tools consisted of models for simulating groundwater flow, groundwater movement, and nitrate loading. These tools were used to map recharge areas for wellfields, calculate nitrate loadings to groundwater, and estimate future nitrate concentrations.
Several groundwater quantity and quality challenges were revealed. Groundwater pumping is reducing streamflows in the Plover and Little Plover Rivers. Using year 2005 projected pumping rates, groundwater withdrawals will deplete about 10% of the baseflow to the Plover River, and possibly more than 40% of the flow to the Little Plover. Since the Plover is a larger stream, it might be able to sustain this impact without losing its value for recreation and wildlife. However, the loss of 40% of Little Plover baseflow would certainly severely impact its ecology, and possibly dry some stretches. Even lower-than-projected pumping rates will likely impact the Little Plover. Additional work needs to be done to quantify stream impacts from groundwater withdrawals.
Nitrate concentrations have been rising in the recharge area for over 30 years. The tools developed in this study were used to estimate what nitrate levels will eventually be reached in individual municipal recharge areas under current land uses and management practices. Estimates were made for both conventional and "Best Management Practice" (BMP) agriculture. The tools predict that nitrate concentrations will continue to increase. Predicted steady-state nitrate-N concentrations at the municipal wellfields are Stevens Point Main - 4/3 mg/L (4 = conventional agriculture, 3 = BMP); Stevens Point #5 - 22/16; Whiting - 38/26; Plover 26/19. Most of the nitrate is from agriculture. Even eliminating all nonagricultural nitrate loading to groundwater in general makes only a small improvement in groundwater quality. Agricultural best management practices are able to reduce nitrate loadings by about 30%, assuming 100% farmer adoption, but this is not sufficient to improve groundwater very much over conventional practices.
The tools developed for this project are intended to be used to manage groundwater in the future. The models can be updated as new information becomes available. They can be used to predict future groundwater levels, to help determine where a particular pollutant is originating, or to determine where a pollutant will travel from a contamination source.
| |
Figure 1. Location of Stevens Point - Whiting - Plover area in Wisconsin |
The aquifer in the Stevens Point - Whiting - Plover (SWP) area is an important local resource, providing domestic water to 40,000 residents of the villages and city, as well as to many rural residents. The aquifer also supplies water for crop irrigation, industry, and livestock. Groundwater discharging from the aquifer feeds streams and wetlands, and thereby also supports fish and wildlife.
Groundwater quality in the SWP area has been especially impacted by nitrate and pesticides. The nitrate standard (10 mg/L as nitrate-N) is exceeded by both the Whiting and Plover wells and perhaps 25% or more of rural domestic wells. Long term trends indicate that nitrate concentrations continue to increase (Fig. 2 and 3).
 |
Figure 2. Nitrate levels in the Village of Whiting municipal well |
| Figure 3. Nitrate levels in the Little Plover River |
 |
A recent sampling effort in Portage County found that 45% of tested wells contained atrazine residues, but additional information is needed to characterize the extent of pesticide pollution. Groundwater quantity problems may become more prevalent in the future. As more water is diverted from the aquifer for consumptive uses, less discharges to streams. Decisions may be needed in the future as to whether some surface waters should be impacted to supply water for other uses.
Until now, only conjecture and back-of-the-envelope type calculations have been available to try to predict the future. The tools developed in this study allow a more quantitative approach.
 |
Groundwater in the area originates as precipitation which infiltrates
to the water table. About 10" of precipitation recharges groundwater per
year. Groundwater generally flows to the west and southwest, where it
discharges into the Plover, Little Plover, or Wisconsin Rivers (Fig. 4),
or to production wells (Fig. 5). Figure 4. Water table elevations for the SWP area |
The USGS groundwater flow computer code "MODFLOW" was used to simulate groundwater flow in the SWP area (Fig. 6 and 7). The flow model illustrated that projected pumping will reduce flows in local streams. Using year 2005 projected pumping rates, the model predicts that nonagricultural pumping will reduce Little Plover River baseflow by perhaps more than 40%. (Agricultural pumping currently reduces flows by an estimated 10%.) This will substantially impact the stream, and possibly dry up some stretches. Even present pumping rates are likely to impact the Little Plover. Additional work needs to be done to quantify Little Plover impacts. Pumping by the Stevens Point wells are projected to reduce flow in the Plover River by about 10% compared to nonpumping conditions.
 Figure 6. MODFLOW model design |
 Figure 7. MODFLOW water table simulation |
"Particle-tracking models" track the movement of groundwater in a simulated flow system. They are equivalent to tracking the movement of a tiny particle injected into groundwater. With the particle tracking model, we can track groundwater forward in time as it moves downgradient of a given location, or backward in time to where the groundwater originated. The particle tracking model can be used to help locate the source of pollution to a well, or where polluted groundwater will travel from a particular pollution source. We used the USGS MODPATH particle tracking computer code to delineate recharge areas and time of travel for groundwater to the municipal wells.
 |
To delineate the recharge area of the municipal wells, we used the
particle tracking model to place particles around each of the wells (plus
a number of industrial wells) and tracked them upgradient to where they
originated. We used year 2005 projected pumping conditions. The computer
code also calculates time-of-travel zones around the pumping wells. The
result is Fig. 8, which shows the zones of contribution for the municipal
wells. Note that the Whiting recharge area is disproportionately large
because its well is co-located with a number of industrial
wells. |
 |
Figure 9. Land uses in the 30 year time-of-travel
zones Predicted nitrate concentrations at the municipal wells were calculated using a mass-balance approach for the land-uses and land management practices in the 30 year time-of-travel (Fig. 9). The 30 year TOT was used because it is workable in size, is representative of the entire zone of contribution, and comprises about 2/3 of the total zone of contribution. The mass balance approach depends on an assumption that denitrification does not occur in the area's aquifer or soils. No evidence of denitrification in the area has been found, with the possible exception of part |
Most of the nitrate entering the recharge area originates from agricultural land-uses. The agricultural component includes cropland, legumes, and manure.
Loading from cropland (Fig. 10) was calculated using the average crop census for the 30 year TOT and the mass-balance approach of Meissinger and Randall (1991). Predicted nitrate-N loading rates for crops in the SWP area were about 60-130 lb/acre under conventional practices. Best Management Practice agriculture in general was able to reduce this amount by about 30%. Loading from legume forages was assigned 95 lb/acre (conventional) or 0 lb/acre (BMP). These estimates are likely low by 40-80 lb/acre. Manure loading was estimated at 5.2 lb/ton (conventional) and 2.5 lb/ton (BMP).
 |
Figure 10. Nitrate-N loading for selected crops for conventional (CON) and best management practices (BMP) |
Nitrate-N loading from unsewered residential land-uses was assigned as 48 lbs per acre per year for one acre lots, and 12 lbs per acre per year for 4 acre lots. Urban (sewered) residential areas were assigned 8 lbs per acre per year. These amounts are likely somewhat on the high side. Virtually no nitrate comes from land uses such as forest, grassland, airport, etc.
| Conventional ag | BMP ag | ||
| Stevens Point Main | 4 mg/L | 3 mg/L | |
| Stevens Point #5 | 22 | 16 | |
| Whiting | 38 | 26 | |
| Plover | 26 | 19 |
 |
Figure 11. Predicted future steady-state nitrate
concentrations in municipal well recharge areas All predicted concentrations exceed current nitrate levels. Data for the Whiting and Plover wells support the notion that nitrate will increase over time. Whiting has increased from 4.0 mg/L of nitrate-nitrogen in the 1960s to almost 20 mg/L at present. |
Plover wells were about 9 mg/L in 1989, and are currently in the 11-15 range. Upward nitrate trends have also been observed in baseflow in the Little Plover River, which is located between the Whiting and Plover wells. The Stevens Point #5 well is currently around 5 mg/L, considerably less than the predicted 16 to 22. Several factors may account for this discrepancy. Presently, the well is not pumped consistently so that an approximate steady-state scenario may never develop. Second, an inaccuracy in the model may neglect actual induced recharge from the river which would tend to dilute nitrate. The Stevens Point main wellfield has remained fairly consistent with respect to nitrate; current values are generally in the 1.5 to 2.5 mg/L range, close to the predicted concentrations.
The current and predicted low nitrate concentrations at the Stevens Point Main wellfield are due to where the groundwater originates and the land uses in those areas. This wellfield has three distinct recharge zones; west of the Plover River, the river itself, and east of the river. The proportions of groundwater contributed in each zone are roughly 27% west, 64% River, and 9% east. Land uses in the east cause this to be a high nitrate area with a predicted nitrate nitrogen concentration of 28.1 (conventional) or 19.7 mg/L (Best Management Practices) compared to river water (1.2 mg/L) or west water (about 2.5 mg/L).
 Figure 12. Sources of nitrate-N in municipal well recharge areas |
The relative contribution of nitrate loading varies by wellfield (Fig.
12). For Plover, 98% is from agriculture, with about 94% from irrigated
lands.
Whiting is also strongly impacted by agriculture, which accounts for about 90% of nitrate loading. Irrigated cropland, which comprises about 51% of the land use contributes 65-69% of the loading; dryland agriculture contributes 5%, and manure contributes 22 or 15%, depending on conventional or BMP. Unsewered residential use contributes 6-9% of the nitrate. Eliminating all the nonagricultural loading only reduces nitrate concentrations by 2.9 mg/L. The major loadings for Stevens Point #5 is from both agriculture (67-54%) and unsewered residential (31-44%). Water quality in the Stevens Point Main wellfield is largely influenced by the Plover River. The river contributes 64% of the water, at low nitrate concentrations. |
Potential management schemes need to address not only water quality goals at the municipal wells, but also rural water users and surface water goals. Universal use of agricultural BMPs could theoretically reduce nitrate concentrations by about 30%. However, universal adoption would be difficult to achieve; adoption rates are rarely higher than 50% in watershed projects. Even assuming universal adoption, predicted nitrate-N concentrations still lie in the 16 to 26 range in the high nitrate areas. Eliminating all nonagricultural nitrate loading could be accomplished by banning nonagricultural fertilizer use and eliminating septic systems. This approach has potential only in Stevens Point #5 zone-of-contribution. In the other high nitrate areas, nonagricultural loading is only 10% or less of the total, so eliminating this source drops nitrate concentrations only a few parts per million. Any realistic scheme to lower nitrate pollution in groundwater will have to address agricultural inputs.
The modeling procedure was based on a number of assumptions. Sensitivity analyses indicate that the predictions are robust and insensitive to reasonable uncertainties regarding the assumptions. The one assumption with potential to introduce error is that of no denitrification. Field data suggest this assumption is valid for the Whiting, Plover, and Stevens Point #5 wellfields. In the Stevens Point - Main zone of contribution, there may be the potential for denitrifying conditions due to the presence of organic soils. These may cause nitrate concentrations in the East zone of contribution to be substantially lower than predicted.