<?xml version="1.0" encoding="UTF-8"?><?xml-stylesheet type="text/xsl" href="static/style.xsl"?><OAI-PMH xmlns="http://www.openarchives.org/OAI/2.0/" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.openarchives.org/OAI/2.0/ http://www.openarchives.org/OAI/2.0/OAI-PMH.xsd"><responseDate>2026-11-21T06:01:59Z</responseDate><request verb="GetRecord" identifier="oai:macsphere.mcmaster.ca:11375/33921" metadataPrefix="dim">https://prod-ms-be.lib.mcmaster.ca/server/oai/request</request><GetRecord><record><header><identifier>oai:macsphere.mcmaster.ca:11375/33921</identifier><datestamp>2026-11-13T08:13:04Z</datestamp><setSpec>com_11375_271</setSpec><setSpec>col_11375_272</setSpec></header><metadata><dim:dim xmlns:dim="http://www.dspace.org/xmlns/dspace/dim" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:doc="http://www.lyncode.com/xoai" xsi:schemaLocation="http://www.dspace.org/xmlns/dspace/dim http://www.dspace.org/schema/dim.xsd">
   <dim:field mdschema="dc" element="contributor" qualifier="advisor">Kevlahan, Nicholas</dim:field>
   <dim:field mdschema="dc" element="contributor" qualifier="author">Moazzami, Hamidreza</dim:field>
   <dim:field mdschema="dc" element="date" qualifier="accessioned">2026-11-12T15:54:31Z</dim:field>
   <dim:field mdschema="dc" element="date" qualifier="issued">2026</dim:field>
   <dim:field mdschema="dc" element="identifier" qualifier="uri">https://hdl.handle.net/11375/33921</dim:field>
   <dim:field mdschema="dc" element="identifier" qualifier="uri">https://doi.org/10.71548/25690</dim:field>
   <dim:field mdschema="dc" element="description" qualifier="abstract">Variational data assimilation provides a framework for estimating the solution of dynamical systems by combining model predictions with observational and background information. However, the high computational cost of four-dimensional variational data assimilation (4D-Var), particularly the repeated solution of Hessian-based systems, limits its application. This thesis develops innovative approaches to reduce the computational cost of 4D-Var based on adaptive wavelet methods and Fourier neural operators (FNOs).

First, an adaptive wavelet multigrid method is developed for solving the 4D-Var Hessian system to estimate the optimal initial condition from sparse observations distributed in space and time. By exploiting the natural multiscale structure and ellipticity of the Hessian induced by the background term, the method achieves a relative error of O(10⁻³) with respect to the true initial condition for the linear advection equation while using only 35% of the available wavelet coefficients. Secondly, an adaptive wavelet method is proposed for incremental 4D-Var. It uses adaptive grids obtained from the nonlinear trajectory for tangent linear model (TLM) integrations. For the viscous Burgers equation, this method provides significant computational savings while maintaining faithful reconstructions. With an adaptive wavelet tolerance of 0.1% it gives a TLM speed-up of 7 times and a 13% reduction in total assimilation time. Thirdly, an FNO model is developed to learn the nonlinear mapping associated with the optimality conditions of 4D-Var to provide efficient initializations for conventional solvers. For linear advection, this hybrid FNO–4D-Var approach reduces average relative error by 62% and conjugate gradient iterations by 17%. For the viscous Burgers equation, it achieves higher accuracy in 72% of test cases and fewer iterations in 75% of cases. Finally, the applicability of incremental 4D-Var is investigated for the chaotic Kuramoto–Sivashinsky equation, revealing the importance of appropriate observation term weighting for accurate trajectory reconstruction.

The results of this thesis demonstrate the potential of hybrid 4D-Var methods employing adaptive wavelets and FNOs to reduce computational cost while preserving the robustness of standard variational data assimilation.</dim:field>
   <dim:field mdschema="dc" element="description" qualifier="degreetype">Thesis</dim:field>
   <dim:field mdschema="dc" element="description" qualifier="degree">Doctor of Philosophy (PhD)</dim:field>
   <dim:field mdschema="dc" element="title">ADAPTIVE WAVELETS AND FOURIER NEURAL OPERATORS FOR ACCELERATING VARIATIONAL DATA ASSIMILATION</dim:field>
   <dim:field mdschema="dc" element="type" lang="en">Thesis</dim:field>
   <dim:field mdschema="others" element="access-status">open.access</dim:field>
</dim:dim></metadata></record></GetRecord></OAI-PMH>