Results of 2002 and 2003
Muddy Creek Flow and Sediment Assessment

by Linzy Browning, Research Assistant
James W. Bauder, Soil and Water Quality Specialisty
Kim Hershberger, Research Associate
Holly Sessoms, Research Associate

Complete Background and Design of Muddy Creek Flow and Sediment Assessment Project

Brief Background
Irrigation within Muddy Creek’s watershed has resulted in significant alteration of its hydrology. Each year, coincident with commencement of irrigation in GID, Muddy Creek’s flow and sediment volumes measured at its confluence with the Sun River increase substantially. In 1996, Muddy Creek and its receiving stream, the Sun River , were both placed on Montana ’s 303(d) list, largely because of flow and sediment loads exceeding limits defined in the TMDL process. Muddy Creek’s inclusion on this list resulted in a need for comprehensive, quantified, reliable data that could be used to develop flow and sediment TMDLs. This study monitored flow and sediment in Muddy Creek’s 5 irrigation-fed tributaries and in its 40 km (25-mi.) reach from Power to Vaughn. Information gathered was then used to partition stream flow and sediment among their sources, assess volumes of water contributed by farm field sources and operational spillage, identify relationships between flow volume and sediment concentration, and assess the degree of fluctuation within Muddy Creek and its tributaries.

Project Objectives
The flow and sediment assessment of Muddy Creek had four main objectives.

• Identify and quantify flow contributions to Muddy Creek from irrigation project operational spillage and farm field sources.
• Quantify amounts of sediment contributed by each irrigation-fed tributary.
• Quantify the volume of sediment generated within Muddy Creek proper and the amounts and sources of sediment at Vaughn (the confluence of Muddy Creek with the Sun River).
• Attempt to define relationships between flow volume and sediment load in Muddy Creek and its tributaries.

2002 Results
A primary objective of the Muddy Creek assessment was to partition stream flow and sediment during the irrigation season among their possible sources. Figure 1 presents the complete 2002 and 2003 irrigation season data sets, which consists of cumulative flow volumes and sediment loads measured at each gauging station. With this information, flow volumes and sediment loads contributed by the reach above Power, irrigation-fed tributaries, and Muddy Creek proper were calculated (Table 1) . Although MC#1 empties into Muddy Creek below Gordon, for simplicity, its flow and sediment contributions were included with the other tributaries.

The gauging station at Power did not function for most of the 2002 irrigation season. Consequently, flow at Power was determined by subtracting flows from the tributaries above Gordon from flow at Gordon. This eliminated the opportunity to calculate sediment and seepage contributions between Power and Gordon and probably attributed more flow to the reach above Power than it actually contributed.

2002 flow and sediment data confirmed and quantified what had previously been hypothesized and observed qualitatively. Forty-nine percent of water delivered to the Sun River by Muddy Creek coincident with the irrigation season originated within the GID tributaries, while the majority of sediment, 72%, originated within Muddy Creek proper in the 20 km lower reach between Gordon and Vaughn (Table 1).

Partitioning of Flow
Knowing that nearly half the water flowing down Muddy Creek originated within GID, the next step was to further investigate percentages attributable to operational spillage and farm field sources. MC#1, Tank Coulee, and Spring Coulee all received operational spillage from GID in their upper reaches, so their upper gauging stations primarily measured spillage, while stations downstream measured all sources combined. Differences in flow between lower and upper stations were assumed to be flow contributed by farm field sources, which included seepage and field-end return flow. Figure 2 illustrates flow volumes attributed to operational spillage and farm field sources in the five tributaries during 2002. In aggregate, seventy percent of tributary flows were attributed to farm field sources.

Figure 2. Flow volumes contributed to irrigation fed tributaries (ac.-ft.) by farm field sources and operational spillage. Farm field sources include irrigation return flow and seepage.

Flow x Sediment Relationship
Assessing correlations between increased flow volumes and sediment concentrations was another objective of the Muddy Creek study. Figure 3 displays attempts at defining relationships between flow volume and sediment concentration in Muddy Creek and its monitored tributaries during the 2002 irrigation season. Flow in Muddy Creek proper ranged from 150 to 500 cfs, while tributary flow volumes remained below 100 cfs. Sediment data for flows below this threshold displayed a grouped pattern, while sediment data showed weak trends in relation to flow volume for flows greater than 100 cfs. This was consistent with historic data, which suggested a threshold of approximately 80 cfs, below which there was no identifiable relationship between flow volume and sediment concentration.

Figure 3. 2002 relationships between flow (cfs) and sediment concentration (mg/L) in Muddy Creek and its tributaries.

As can be seen in Figure 3, even at flow values greater than 80 cfs, the relationship between flow volume and sediment concentration, at both the Gordon and Vaughn stations, is poorly defined and inverse in 2002. This inverse relationship was attributed to the fact that flows greater than 200 cfs were generally consequences of single precipitation events in the basin up gradient of the Power gauging station, which is not a significant source of sediment. However, in the 2002 data shown in Figure 3, sediment concentrations at Vaughn are generally greater than sediment concentrations at Gordon at comparable flow rates. This circumstance substantiates the hypothesis that a measurable portion of sediment at the Vaughn station is a consequence of flow from GID causing erosion below the Gordon station (in Muddy Creek’s 12 mi. lower reach) and not necessarily sediment sourced above the Gordon station or within the GID boundaries.

2003 Results
The functioning stage level recorder at Power during the 2003 irrigation season measured flow at Power at approximately 40 cfs, significantly less than originally thought. Flow increased from less than 20 cfs to approximately 40 cfs coincident with irrigation in GID, indicating that seepage, return flow and/or operational spillage upstream of the project boundary was impacting flow volume within the project area. 2003 data, presented in Figure 1, also revealed that sediment concentration tripled between Power and Vaughn, increasing from 0.21 to 0.74 T./ac.-ft. Concentration of sediment in Muddy Creek at Power most likely reflects sediment concentration under natural base flow conditions, while concentration at Vaughn reflects consequences of substantial flow increase below the GID project.

Table 2 displays flow and sediment load in the reach of Muddy Creek upstream of Power, the area within the GID discharge boundary, and the reach from Gordon to Vaughn during the 2003 irrigation season. As in 2002, the largest portion of flow at Vaughn, 67%, was attributable to irrigation-related sources. The majority of sediment at Vaughn, 58%, originated within Muddy Creek’s channel in its 12 mi. lower reach between Gordon and Vaughn.

Partitioning of Flow
Table 3 partitions flow volumes contributed to Muddy Creek by its tributaries and its reach above Power among base flow, GID-reported operational spills, and farm field sources, which include field spills, irrigation return flow, and seepage. Table 3 also includes total sediment contributions from these sources. Base flow was estimated from perennial flows, and farm field sources were calculated by subtracting base flow and operational spills from total gauged flow. Operational spill records from GID in 2003 revealed that MC#2 and MC#3 received significant volumes of spillage.

Flow x Sediment Relationship
Flow volume and sediment concentration were significantly correlated in Muddy Creek and its tributaries during 2003. Figure 4 displays these positive linear relationships. Lower flows at Power and in MC#1, Tank Coulee, and Spring Coulee led to more positive correlations between flow volume and sediment concentration compared to higher flow volumes at Gordon and Vaughn. While Figure 1 indicates that low flow volume streams had little overall erosion, steeper trend line slopes in the lower volume parts of Muddy Creek and its tributaries suggest that additional factors influence erosive characteristics of these streams.

Figure 4. 2003 relationships between flow (cfs) and sediment concentration (mg/L) in Muddy Creek at Power, Gordon, and Vaughn and in Spring Coulee, Tank Coulee, and Muddy Creek #1. For tributaries, data from upper, middle, and lower stations are combined.

Average sediment concentration in Muddy Creek at Vaughn nearly doubled from 2002 to 2003, increasing from 0.39 to 0.74 T./ac.-ft. Although the increased sediment load in 2003 was somewhat unexpected, it likely resulted from differences in rainfall and stream level fluctuation during the two years. In 2002, 9.03 in. of precipitation fell in the GID area during the irrigation season in comparison to 0.235 in. received during the 2003 irrigation season. Substantial rainfall during the 2002 irrigation season reduced demand for irrigation water deliveries and consequent operational spillages, while lack of rainfall in 2003 increased demand for irrigation water, as indicated by the fact that total GID diversions increased by 49%, from 46,247 ac.-ft. in 2002 to 68,999 ac.-ft. in 2003. GID reported spills totaled 6,591 ac.-ft. in 2002 and 9,016 ac.-ft. in 2003, a 37% increase.

Zaimes et al (2004) found that the majority of stream bank erosion in a second order stream, 60-80%, occurred in just a few days a year, during sudden high flow events. In Muddy Creek, erratic cycles of low flow followed by large volumes of spillage into the 5 monitored tributaries simulate the consequence of major rainfall events in lower Muddy Creek, which cause water level fluctuation and subsequent scouring and stream bank loss. This effect is more pronounced in drought years like 2003, when spillage more dramatically influences flows and water levels in depleted streams. Table 4 displays average daily fluctuation (average difference between daily minimum and maximum flows) at Power, Gordon, and Vaughn and within each Muddy Creek tributary during 2003.

Comparison of 20-year average, 2002, and 2003 flow and sediment data and flow x sediment relationships presented in Figures 3 and 4 clearly demonstrates diverse short and long-term relationships between flow volume and sediment in Muddy Creek. In fact, a relationship between flow and sediment in Muddy Creek clearly is not consistent over time, primarily because of intervening effects of irrigation-related flow contributions and differences in stream response to flows related to operational spillage and precipitation events.

Conclusions
Original estimates attributed 43% of total flow in Muddy Creek to natural sources, including base flow and spring seepage, 18% to farm field sources, and 39% to operational spillage. In 2003, base flow accounted for 23% of total gauged flow, operational spillage for 23%, and farm field sources for 55%. In short, initial estimates attributed more flow to natural sources and spillage and less to farm field sources than was measured during 2003, a result that probably would not be seen in a more normal rainfall year in which greater base flow, less irrigation demand, and therefore less operational spillage, return flow and seepage, would elevate the portion of flow attributable to natural sources. It thus becomes obvious that in the case of sourcing irrigation return flow and sediment, effects of climatic and return flow conditions on stream dynamics need to be taken into consideration.

This study substantiated the hypothesis that irrigation seepage, return flow, and spillage increase irrigation season flow volumes and sediment concentrations in Muddy Creek. Although increased flows alter stream hydrology, increased flow volume alone should not prevent stream bank stabilization and establishment of vegetation. However, it appears that flow volume fluctuation has prevented such stabilization from occurring.