Abstract
Groundwater–surface water interactions strongly influence streamflow, water quality, and ecosystem processes, yet many temperature-based methods assume stable streambed conditions and may become unreliable when scour and deposition alter sensor burial depth. This dissertation presents a flexible temperature-based framework that estimates streambed elevation, effective thermal diffusivity, and groundwater–surface water exchange fluxes under dynamic bed conditions using vertical temperature profiles. The framework combines signal processing and parameter estimation within a modular architecture that facilitates future methodological developments and has been implemented in a graphical user interface for practical application. A new parameter-estimation approach, MLEnZ, extends the Maximum Likelihood Estimator (MLEn) framework to jointly estimate streambed elevation changes, thermal diffusivity, and exchange fluxes through time. The dissertation also introduces a methodology for scaling point estimates to the reach scale by spatially integrating fluxes along the stream corridor to quantify exchanged volumes and characterize longitudinal exchange patterns. Results demonstrate that the framework can identify spatially heterogeneous exchange conditions and produce reach-scale estimates consistent with independent hydrological observations, extending the applicability of temperature-based methods across a wider range of environmental conditions and spatial scales.