A study on numerical problems of a coupled modelling of regional water balance and anthropogenic landcover change
W. ZHANG (TU Delft - Civil Engineering & Geosciences)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
This thesis describes the approaches applied to attempt to solve the numerical problems of the regional atmospheric model incorporated in the coupled modelling of regional water balance and anthropogenic land cover change in Amazon basin. For computational efficiency, the previous atmospheric model is evaluated at monthly scale. In order to cope with the numerical instability, nearest neighbouring averaging interpolation is iteratively performed to smooth the solutions as a transitional approach. Therefore a subsequent study is conducted to investigate the origin of the numerical instability and whether there are feasible measures to fix the numerical problem of the modelling. Chapter 1 serves as an introduction, which briefly introduces the background and research question — Whether there is any possible remedy that can solve the numerical instability of the monthly-timestep regional water balance model and obtain convergent solutions? Chapter 2 contains 7 sections, each of which gives statement of the specific problem, the experimental method applied, the corresponding results and related discussions. It has been concluded from the series of experiments that — a. adding diffusion terms makes no sense; b. applying smaller fractions of wind helps alleviate instability but the applied monthly timestep length seems to make the model paradoxic and inherently not convergent; c. instability is not really relevant with the iterative method; d. after correcting a dormant error in previous research, the results gets no better; e. the model may be so oversimplified that cannot reflect the reality; f . wind and monstrous timestep length are the keys to the problem, especially the latter is more problematic. Chapter 3 summarizes the discussions and conclusions from chapter 2 and proposes several recommendations for future research such as trade-offs between model complexity and efficiency, a heuristic way of making wind endogenous and reconsideration of the model architecture.