Montana State University Land Resources and Environmental Sciences Water Quality HomePage

A TMDL Approach to Muddy Creek

By Holly N. Sessoms and James W. Bauder
Research Associate and Professor, respectively

Background

Muddy Creek is located in North Central Montana, near Great Falls. It is a tributary of the Sun River and has been a topic of debate and concern for the past forty plus years. Muddy Creek, as indicated by its name, has always been a turbid creek due to naturally erosive conditions. During early settlement times, it was considered an ephemeral stream and flowed only in response to precipitation events. Since implementation of irrigation projects in the 1920s, stream flows in Muddy Creek have greatly increased and become widely fluctuating. These increased flows have contributed to substantial erosion of stream banks and to greatly increased sediment loads delivered to the Sun River. On average, eight acres of bottomland along the banks of Muddy Creek are washed away each year. Muddy Creek experienced its greatest erosion and sediment loads in the 1960s, with more than 200,000 tons of sediment delivered to the Sun River annually. Since that time, substantial management and stream bank improvements have been made and sediment loads have decreased to less than 50,000 tons per year. However, Muddy Creek and the Sun River below Muddy Creek are still considered impaired by the Montana Department of Environmental Quality (DEQ) and are listed on the state's 303(d) list as "high priority" streams for Total Maximum Daily Load (TMDL) development. Muddy Creek and the Sun River below Muddy Creek are both unable to meet their designated uses, greatly due to excessive sediment loads. The ongoing problem of excessive flows and sediment in Muddy Creek is partially attributable to the fact that the specific source of excessive flows (other than irrigation) is unknown. The Bureau of Reclamation has requested that the Montana State University Extension Service, along with the help of the Sun River Watershed Association and Natural Resource Conservation Service (NRCS), define and quantify these contributing flows.

Setting

Muddy Creek is 42 miles long and encompasses a 314 square mile watershed. Within the Muddy Creek watershed lies the Greenfields Irrigation District (GID), a Bureau of Reclamation project situated just west of Muddy Creek. It is considered the major contributor of excess flows to the creek. GID consists of 80,000 irrigated acres, 50,000 of which are in the Muddy Creek drainage. The area has been intensely cropped and major irrigated crops include malt barley, alfalfa, and canola. GID sits atop a terrace underlain by fifteen to twenty feet of gravel. This gravel layer is underlain by an impermeable bedrock layer and is apparent from seeps caused by subsurface return flows. Soils on the bench land are generally shallow and have a low available water capacity. Soils along Muddy Creek are mostly alluvial clays and silty clay soils, and also exhibit poor drainage. This combination of excessive flows and naturally erosive conditions has caused Muddy Creek between Gordon and Vaughn to become severely incised with steep, raw banks and numerous headcuts.
Figure 1. Muddy Creek Watershed. All monitoried tributaries (MC#1, MC#2, MC#3, Tank Coulee, and Spring Coulee) are shown, along with monitoring stations at Power, Gordon, and Vaughn.

Approach

The ultimate goal when considering Muddy Creek is to reduce sediment loads to the Sun River. Because of the causal relationship between flow and sediment, flows within Muddy Creek must be reduced in order to reduce sediment loads and a maximum allowable flow must be identified. This idea of identifying maximum allowable loads can be considered a TMDL approach to a watershed. TMDL is defined by the MT DEQ is "the total amount of pollutant, per day, that a water body may receive from any source (point, non-point, or natural background) without exceeding state water quality standards." In the case of Muddy Creek, inflows must first be identified, and then quantified. Excess flows in Muddy Creek, or those not induced by natural causes, can be categorized as irrigation seepage and irrigation spillage. The natural, low flow of Muddy Creek is referred to as base flow and natural contributions to base flow are considered spring seepage. Figure 1 illustrates these flows relative to one another.

Figure 1. Comparison of estimated flow contributions to twenty-year average flows for Muddy Creek at Vaughn.

Because irrigation seepage continues to influence flow through November, the hydrograph year starts in December when flows have returned to normal base flows. Base flow throughout the irrigation-influenced season (May 1 - November 30) is considered the average low flow value, or 41 cubic feet per second (cfs). Spring seepage was estimated by extending the slope of flows from May 1 up to a peak flow value on June 15. Flow remains at peak values until June 15, and then begins to return at a constant rate to base flow on September 15. Irrigation seepage was determined by extending the slope of flows backwards from November 30. Irrigation seepage is expected to reflect total flows at a ten day delayed rate. For this reason, irrigation seepage is estimated to peak on August 10, about ten days after total flow peaks. Irrigation seepage induced flows begin on May 10, approximately 10 days after irrigation begins. Total seepage represents the total flow expected from both spring seepage and irrigation seepage. Irrigation spillage is expected to be the greatest contributor to excess flows, and is comprised of all other flow. These projected flows (other than base flows) are all best estimates, as there is no existing data for individual contributors.

Determining flow was just the first step in determining sediment contributions. By using USGS sediment data collected at Muddy Creek at Vaughn and Muddy Creek at Gordon stations, a linear relationship between flow and sediment loads was identified. To be most accurate, regression analyses was done for both sites for flows above and below 80 cfs, as illustrated in Figures 2 and 3.

Figure 2. Sediment loads (tons/day) for flows < 80 cfs for the Muddy Cr. at Vaughn and Muddy Creek at Gordon USGS stations.

Figure 3. Sediment loads (tons/day) for flows 80 cfs for Muddy Cr. at Vaughn and Muddy Creek at Gordon USGS stations.

These relationships then allowed us to determine daily and yearly sediment loads for base flow, spring seepage, irrigation seepage, and irrigation spillage. Sediment loads, acre-feet of flow, and the percentage of the average total for Muddy Creek at Vaughn are summarized in Table 1a. Total sediment load and acre-feet of water produced above the Gordon station is summarized in Table 1b.

Table 1a. Summary of estimated flows and sediment contributions for Muddy Creek at Vaughn from the identified sources: base flow, spring seepage, irrigation seepage, and irrigation spillage.
Source
Estimated Volume of Flow (acre feet)
% of Total Volume
Estimated Sediment Load (tons/year)
% of Total Load
Base flow
29,464

33

11,890

24

Spring Seepage

8,415

9

5,278

11

Irrigation Seepage

16,131

18

9,547

20

Irrigation Spillage

34,937

39

22,117

45

Total Contributions (20 yr. average)

88,947

100

48,831

100

Table 1b. Summary of estimated flow volumes and estimated sediment loads for Muddy Creek at Gordon
Source
Estimated Volume of Flow (acre feet)
Estimated Sediment Load (tons/year)
Difference in Volumes between sites (acre feet)
Difference in Sediment Load between sites (tons/year)
Total Contributions (20 yr. Average)

71,840

13,174

17,107

35,657

As expected, the majority of flow is produced above the Gordon station (Figure 4), and the majority of sediment is produced between the Gordon station and Vaughn station (Figure 5).

Figure 4. Portion of total flow contribution (acre feet) from above Muddy Creek at Gordon vs. portion of total flow contribution from between Muddy Creek at Gordon and Vaughn.

Figure 5. Portion of total sediment load contributed to Muddy Creek at Gordon vs. portion of total sediment load contributed to Muddy Creek between Gordon and Vaughn.

Estimating these flows and sediment loads now gives us some direction when "heading to the field" and collecting actual data.

So now what…

Now that the number crunching has been done, the fun part begins. In addition to the two USGS gaging stations, the Sun River Watershed Association and Bureau of Reclamation has established eight new monitoring stations along Muddy Creek and its tributaries. These stations are located at the head and bottom of major tributaries to Muddy Creek and are known as MC1, MC2, MC3, Tank Coulee, top of Tank Coulee, Spring Coulee, top of Spring Coulee, and top of MC1. At each of the stations, sediment samples and stream flow measurements will be collected. This will be done on a weekly basis during the irrigation season. Two of the monitoring stations are located at canal spillage sites, and one is located at a canal diversion site. By monitoring these sites, and the bottom of the tributaries, we should be able to distinguish between seepage and spillage. By monitoring several tributaries from GID, we should also be able to identify which areas of GID are contributing the largest amount of water. Sediment data collected at each site will enable us to determine the source of sediment loads, and whether the excessive loads are due to upland erosion or erosion within the channel itself. By targeting major contributors of flow and sediment, we will then be able to focus our efforts on those sore spots and identify achievable goals within the Muddy Creek watershed.

If you are interested or have any questions contact Holly N. Sessoms.

Home Page

Questions/Comments: waterquality@montana.edu