| Summary: | Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Civil and Environmental Engineering, 2008. This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. Includes bibliographical references. This thesis consists of four separate studies which explore the use of random walk particle tracking (RWPT) in simulating environmental mass transport. Two of the studies also consider the efficacy and marine impact of ocean CO2 sequestration by direct injection. The first study compares RWPT to two other simple Lagrangian techniques (forward and backward Gaussian puff tracking) to simulate mixing beyond the near field of a pollutant discharge. RWPT is found to be more accurate, but also more computationally expensive, thus motivating hybrid approaches where Lagrangian calculations transition to Eulerian schemes in the far field. The second study considers 1D RWPT when strong gradients in ambient diffusivity exist. For step profiles, the work of past investigators is unified and extended, and the Thomson et al. (1997) particle reflection approach is recommended. For piecewise linear profiles, a novel and efficient particle reflection with probability translation approach is proposed. The third study implements RWPT to emulate the tracer transport of an ocean general circulation model (OGCM) using the OGCM's flow and diffusivity fields. A high level of agreement between RWPT and OGCM results is achieved. Particle reflection with probability translation successfully handled sharply varying vertical diffusivities. However, precisely mimicking OGCM calculations proved difficult due to complications in specifying the subgrid scale variation of isopycnal slope and diffusivity in steeply sloped or convectively unstable regions, and in accurately implementing the Gent-McWilliams eddy-induced transport. (cont.) Further development is recommended to resolve spurious upwelling occurring mainly in the Southern Ocean. The utility ...
|
|---|