Our 31 model runs spanned the discharge range established from analysis of historical stream flow records. Discharges fell between 78.5 cms, 1 % flow excedence during the wet climatic scenario, and 0.4 cms, 99% flow excedence during the dry climatic scenario (Table 2.1). We have organized and displayed representative model output in both map views (Figure 3.1) and as contour plots of bivariate distributions of depth and velocity (Figure 3.2). Hotlinks between these two figures facilitate comparisons between the display formats.

Modeling output can be viewed and interpreted in three ways. Depictions of depth, velocity, and Froude number provide direct, un-interpreted information about the modeled hydraulic characteristics of flow at each discharge (Figure 3.1 and Figure 3.2). These data are useful for analyzing the discharge dependence of hydraulic conditions and for comparing hydraulic conditions produced by equivalent flow durations of the two climatic scenarios.

The second way data can viewed is through the filter of the habitat classification scheme (Figure 3.1, Figure 3.2, Figure 3.3). This interpretation is useful for analyzing the biological significance of different hydraulic conditions, however it is important to recognize the dependence of habitat conclusions on the strengths and weaknesses of the habitat classification scheme itself. In our case, the habitat classification system is based on field observations and biological sampling at low flows (<2.5 cms) and there is greater uncertainty about the habitat classification extrapolated to higher discharge hydraulics. To emphasize this uncertainty, we have identified habitat areas that are based on habitat classifications extrapolated beyond the range of field biological sampling (Figure 3.1 and Figure 3.3).

Data can also viewed as a comparison between habitat availability during the two hydrologic scenarios.  We calculated habitat-duration curves (Figure 3.4), measures of the persistence of different habitat areas during the wet and dry climatic scenarios.  These curves along with calculations of the average daily habitat areas available during the two hydrologic scenarios (Figure 3.5) are a measure of the sensitivity of habitat distribution to changing hydrologic regimes.

Hydraulic Observations

Map views of model output illustrate the overall drop in flow depths and velocities with declining discharge and the topographic control on flow hydraulics. At all discharges, the highest velocities are concentrated in the shallowest portions of the reach near cross sections 4 and 13 (Figure 2.5) where channel gradient is steepest and where riffles were noted in field observations (Figure 1.4; Figure 1.5). Maximum velocities were reached in the upstream riffle and ranged between 2.5 meters/second (m/s) for a discharge of 78.5 cms, and 0.8 m/s for a discharge of 0.4 cms. Maximum flow depths were concentrated in the middle portion of the reach where a scour hole has formed along the bedrock bluff (Figure 1.4; Figure 1.5). Depths in this pool ranged up to 4.8 m at the highest modeled discharge and 3 m at the lowest modeled discharge.

Contour plots of bivariate distributions of depth and velocity illustrate the discharge dependence of hydraulic variability. At higher discharges, isolines of channel area are smooth and spread out (Figure 3.2), indicating that the discharges create a wide range of depths and velocities across the channel. Greater flow depths and velocities along the channel center line accommodate the higher discharges, while channel perimeters remain zones of slower, shallower flow. This lack of hydraulic homogeneity is reflected in the depth and velocity combination that covers the largest area of the channel, at a discharge of 78.5 cms the most common depth and velocity combination of ~2.1 m and ~1.4 m/s occurs over only 200 m² of the channel. In contrast, for discharges less than 0.6 cms the most common depth and velocity combination of ~0.6 m and 0.1 m/s occurs over 700 m² of the channel. Overall, contours of channel area become progressively more concentrated at lower discharges, indicating greater homogeneity of depths and velocities.

Map views also show that, as discharge declines, Froude numbers shift in opposite directions in low and high gradient portions of the reach. In steep channel segments (riffles), declining discharge leads to a more rapid decrease in depth than in velocity and Froude numbers initially rise. In contrast, in low-gradient sections there is a more rapid decline in velocity than in depth and Froude numbers drop. This causes a bifurcation of Froude numbers with decreasing discharge. Intermediate Froude numbers (0.2-0.4) dominate flow conditions at higher discharges, while at lower discharges the channel becomes more clearly divided into low and high Froude number zones.

Habitat Observations

One of most significant results of the habitat classification is its demonstration of the non-linear relationship between discharge and habitat area (Figure 3.3). As discharge decreases, total channel area decreases so that there is 52% less in-channel habitat between the discharge extremes of 78.5 and 0.4 cms. This overall loss of total channel area does not translate into a gradual decline in the area of all habitat types. Instead, the bulk of this area loss is created by a reduction in channel area dominated by intermediate (0.2-0.4) Froude number hydraulics. Much of the higher depths and velocities which produce these intermediate Froude number conditions are outside of the range of field biological sampling, therefore low flow based habitat classification shows the loss of these channel conditions as a drop in unclassified habitat (Figure 3.3A). When the extrapolated habitat classification system is used, the habitat inventory (Figure 3.3B) shows that the bulk of habitat loss is accommodated by a reduction in the area of races. Both habitat classification systems show that the area of all other habitats, riffles, edgewaters, glides, and pools, remains fairly constant for the higher discharges between 78.5 and 25 cms.

The bulk of changes in habitat areas takes place at discharges less than 12.8 cms. At these lower flows, declines in flow depths and velocities leads to an increase in the area of edgewaters and a decrease in races and riffles. Pool and glide response is more complex. Initially, the area of pools and glides increases, as velocities decline faster than flow depths in low gradient channel zones. But after peaking at a discharge of 4.1 cms, pool area begins to decline as depths become shallower than the threshold of 0.7 m, the estimated minimum depth at which water provides sufficient cover in Ratcliff Ford pools. The ratio of pool to glide habitat changes depending on the classification system used for inventory. The extrapolated classification system categorizes some of the conditions beyond field sampling as glides, therefore, they are more prevalent than pools at higher discharges (Figure 3.3B). But as Froude numbers drop under lower discharge conditions, pool area surpasses glide area. Only when depths begin to drop below the threshold depth for pools, does glide area again begin to rise.

Overall, the habitat inventory shows that pools and glides dominate channel area when discharge is less than 10 cms, conditions which prevail 90% of the year under both wet and dry climatic scenarios. This result is in good agreement with habitat inventories made using field observations and photogrametric methods. For a discharge of 3.3 cms, McKenney (1997) measured similar habitat areas for the Ratcliff Ford reach.

Climatic Implications

The habitat inventories show that one of the greatest differences created by changing hydrologic regimes is in the constancy of flows and the permanence of habitat areas. In general, dry years are much less variable than wet years on the Jacks Fork (Figure 2.2C). For example, the inter-quartile range of daily mean discharges (Q₇₅-Q₂₅) during the wet scenario is five times larger than that of the dry scenario. This means that although total habitat area is at a minimum under dry climatic conditions, the aerial extent of these habitats are relatively stable; during 90% of the year discharge and habitat fluctuations are limited to less than 2 cms (Figure 2.4A). In the wet scenario however, not only is the range of daily mean flows greater but there is also greater variability in low flows, 90% of the flows in the wet scenario span the broader range of 2-8 cms (Figure 2.4A). This implies greater fluctuation in habitat area as discharge rises and falls more often.

Sensitivity of habitat units to the climatic scenarios also can be analyzed in terms of duration and area of habitats. For the wet (1928) and dry (1956) year scenarios, the area of habitat associated with each daily mean discharge was calculated from the model results. The cumulative distributions of habitat areas (Figure 3.4 A-E) show the percent time during the year for which the area of habitat is equal or exceeded by the indicated value. Because some habitat areas can increase and decrease with increasing discharge (Figure 3.3B), habitat duration curves differ markedly from flow duration curves.