Lagrangian modeling of advection-diffusion transport in open channel flow

B.H. Devkota, Jorg Imberger

    Research output: Contribution to journalArticlepeer-review

    25 Citations (Web of Science)

    Abstract

    A fully Lagrangian method is presented for the accurate simulation of advectiondiffusion transport in both steady and unsteady open channel flows. Numerical results arepresented for Gaussian tracer distributions, top hat tracer distributions, and steep tracerfronts (step function) profiles in a uniform flow and are compared against analyticalsolutions and against the results obtained with an Eulerian Quadratic UpstreamInterpolation for Convective Kinematics with Estimated Streaming Terms (QUICKEST)method. It is demonstrated that the Lagrangian scheme can totally eliminate numericaldiffusion and oscillations including those normally observed in steep frontal regions in mostEulerian schemes. In steady, uniform flow, the scheme allowed a large time step to beused and provides exact solutions for a wide range of Courant numbers (results are presentedfor Cr from 0 to 20) and for an entire range of grid Peclet numbers (from 0 to infinity). Thesesimulation results for a uniform flow were extended to flows due to simple waves,solitary waves, and undular bores; again, the scheme produced excellent results. It is shownthat the simple Lagrangian model is an efficient and accurate tool to predict transport inadvection dominated systems, and when it is coupled with the new Lagrangian river model,the dynamic river model, it can be an ideal model for long-term and large-scale simulation oftransport for water quality in river systems and for transport with steep frontal regions
    Original languageEnglish
    Pages (from-to)Article number W12406, 14pp
    JournalWater Resources Research
    Volume45
    Issue numberW12406
    DOIs
    Publication statusPublished - 2009

    Fingerprint

    Dive into the research topics of 'Lagrangian modeling of advection-diffusion transport in open channel flow'. Together they form a unique fingerprint.

    Cite this