Isolating parameter sensitivity in reach scale transient storage modeling. (March 2016)

• Record Type: Journal Article
• Authors: Schmadel, Noah M.; Neilson, Bethany T.; Heavilin, Justin E.; Wörman, Anders
• Is Part Of: Advances in water resources. Volume 89(2016)
• Page Start: 24
• Page End: 31
• Publication Date: 2016-03
• Subjects: Transient storage -- Stream hydraulics -- Parameter sensitivity -- Statistical moments; Hydrology -- Periodicals; Hydrodynamics -- Periodicals; Hydraulic engineering -- Periodicals; 551.48
• DOI: 10.1016/j.advwatres.2015.12.020
• Languages: English
• ISSNs: 0309-1708

Abstract

Highlights: Statistical moments are sensitive to hydraulic parameters. Setting hydraulic parameters isolates transient storage parameter sensitivity. A reach-average hydraulic representation is sufficient to isolate sensitivity. Abstract: Parameter sensitivity analyses, although necessary to assess identifiability, may not lead to an increased understanding or accurate representation of transient storage processes when associated parameter sensitivities are muted. Reducing the number of uncertain calibration parameters through field-based measurements may allow for more realistic representations and improved predictive capabilities of reach scale stream solute transport. Using a two-zone transient storage model, we examined the spatial detail necessary to set parameters describing hydraulic characteristics and isolate the sensitivity of the parameters associated with transient storage processes. We represented uncertain parameter distributions as triangular fuzzy numbers and used closed form statistical moment solutions to express parameter sensitivity thus avoiding copious model simulations. These solutions also allowed for the direct incorporation of different levels of spatial information regarding hydraulic characteristics. To establish a baseline for comparison, we performed a sensitivity analysis considering all model parameters as uncertain. Next, we set hydraulic parameters as the reach averages, leaving the transient storage parameters as uncertain, and repeated the analysis. Lastly, we incorporated high resolution hydraulic information assessed from aerial imagery to examine whether more spatial detail was necessary to isolate the sensitivity of transient storage parameters. We found that a reach-average hydraulic representation, as opposed to using detailed spatial information, was sufficient to highlight transient storage parameter sensitivity and provide more information regarding the potential identifiability of these parameters.