Abstracts

Characterizing Hydrologic Vulnerabilities under Climate Uncertainties using Physical Process-based and Machine Learning Models for Water Systems Management

by Saiful Haque Rahat

Inefficient water systems management is likely when climate change risks are poorly understood. To better understand these risks, decision-makers require credible future climate information to prepare for a range of scenarios from climate vulnerabilities. Keeping that in mind, this doctoral dissertation tries to characterize hydrologic vulnerabilities under climate uncertainties for water quantitative events (e.g., flood, drought) and river water quality in general for long-term risk management. Chapter 1 focuses on establishing an explicit link to thermodynamic and dynamic pathways in the climate that has been missing from current water systems operational policies prescribed by regulation manuals within the US. The study contributes an extended version of an existing weather regime (WR)–based stochastic weather generator (SWG) that allows (1) hourly simulation, (2) over the entire year, and (3) a corrected representation of extremes for a range of climate scenarios. Chapter 2 highlights the need for developing a generalizable water quality assessment technique under future climate uncertainties. In this regard, the study develops a machine learning-based Long Short Term Memory (LSTM) model that reconciles discontinuous water quality data to gridded hydro climatic data, identifies causal factors, and provides effective water quality prediction under climate and human-induced land-use changes. Chapter 3 focuses on utilizing water quality tools developed in Chapter 2 and performs a water quality stress test that relates historical climate and human-behavioral conditions to river water quality outcomes, which will subsequently enable a climate change risk assessment for riparian communities vulnerable to river water contamination. Chapter 4 provides an outline of a future study that focuses on developing a spatially and temporarily comprehensive water quality database to provide an opportunity for the scientific community to do a long-term climate impact analysis for the water quality of river systems in general.