Scaling frequency of channel‐forming flows in snowmelt‐dominated streams

Abstract

1] The scaling properties of channel-forming flows are investigated using a regional flow frequency model developed for snowmelt-dominated streams in Colorado. The model is derived from analyses of daily flow records at 32 gauging stations where we have independent measurements of the bankfull discharge. The study sites are located in alpine/subalpine basins with drainage areas ranging from 4 to 3700 km 2 . The frequency distribution of daily flows at these locations can be reproduced with a broken power law (BPL) function described by two free parameters. Both parameters are strongly correlated with drainage area, and based on these correlations, a regional model capable of predicting the frequency of daily flows above the mean annual flow was formulated.The applicability of the model was tested using daily flow records from 32 similar-size basins in Idaho. The frequency distributions of daily flows in snowmelt-dominated streams in Colorado and Idaho with highly predictable hydrographs (i.e., 1 year autocorrelation above 0.7) are well fitted by the BPL function. According to the model, the frequency of flows greater than bankfull decreases downstream from about 15 d/yr in headwater reaches to about 6 d/yr in downstream reaches. These results imply that the basin response to precipitation and runoff is nonlinear. This multiscaling behavior can be physically interpreted as the result of scale-dependent variations in runoff and sediment supply, which influence downstream trends in the bankfull channel geometry and intensity of sediment transport.Citation: Segura, C., and J. Pitlick (2010), Scaling frequency of channel-forming flows in snowmelt-dominated streams, Water Resour. Res., 46, W06524,

Effective discharge in Rocky Mountain headwater streams

“…Q brk also increases with basin area A (Fig. 3b) as was shown by Segura and Pitlick (2010) who describe F Q = f(Q) and Q brk by broken power law functions.…”

“…Differences in Q eff by a few flow bins are attributable to irregularities in the flow frequency data. Results obtained for Q eff in this study are also expected to be unchanged if F Q = f(Q) is expressed by a broken power law function (Segura and Pitlick, 2010).…”

“…Flow frequency curves F Q = f(Q) are similar among high elevation Rocky Mountain streams with snowmelt regimes (see also Torizzo and Pitlick, 2004;Segura and Pitlick, 2010). The majority of flows (on average 60-80%) fall into the first discharge bin, and approximately 5-10% into the second.…”

“…Using 35 bins as opposed to 20 or 50 is a compromise between broad flow classes that result from 20 bins, which might better represent high flows, and narrow flow classes that result from 50 bins and might define the Q eff more tightly. Using 35 rather than 50 bins (Segura and Pitlick, 2010) also helps minimizing the number of empty flow bins that may occur adjacent to bins containing the largest flows. The number of mean daily flows per bin were counted and expressed as a percentage of N. For a bin with no data, its zero percentage was averaged with the percentage flow frequency of a neighboring non-zero bin, and that percentage value was assigned to both bins.…”

Estimation of Daily Streamflow of Southeastern Coastal Plain Watersheds by Combining Estimated Magnitude and Sequence

“…The selection was achieved using a greedyheuristic search process, which searches for a global optimal solution by finding locally optimal solutions. The selection of DA is consistent with results of other studies that demonstrated the relevance of DA as a major hydrologic scaling parameter, especially for peak and annual flows (e.g., Gupta et al, 1994;Vogel and Sankarasubramanian, 2000;Ogden and Dawdy, 2003;Furey and Gupta, 2005;Segura and Pitlick, 2010). The drainage area, which controls the scale of exchange and interaction between climatic and watershed processes, was selected for all 19 percentiles; the HSI, which is a measure of the control of water movement through the soil matrix (drainage potential), was selected for 10 of the 19 percentiles; and the RF1, which is a measure of extreme climatic events, was selected for 7 of the 19 percentiles.…”