Ashraf Rateb

I study how water mass is redistributed across basins, regions, and continents, and how the statistical structure of that redistribution changes under climate variability, human influence, and environmental change. This includes long-range memory, nonstationarity, scaling behavior, and trend dynamics, with particular attention to identifying the drivers of observed change while separating genuine dynamics from artifacts introduced by observational aggregation and short records.

I study how droughts, floods, and transitions between wet and dry conditions develop across space and time, with emphasis on their persistence, severity, drivers, and predictability. A central question is how much of this behavior reflects identifiable structure in the climate and hydrologic system, and how much remains limited by internal variability, short records, and uncertainty in the observations themselves.

I use GRACE and GRACE-FO, InSAR, and GNSS to investigate groundwater depletion and recovery, land subsidence and uplift, and the relation between subsurface storage change and surface deformation. Because these are ill-posed inverse problems, the work depends on physically informed regularization and careful treatment of heterogeneous errors and sampling across the different observing systems.

I study how externally forced hydroclimate signals can be separated from internal climate variability, and how teleconnections shape regional hydroclimate variability, extremes, and impacts. The emphasis is on probabilistic frameworks that make structural, parametric, and observational uncertainty explicit rather than reducing complex behavior to a single estimate.

This project examines historical and projected climate extremes across Texas using SMILEs to separate forced change from internal variability, Self-Organizing Maps to characterize circulation regimes, and Bayesian Hidden Markov Models to infer latent hydroclimate states and transition probabilities. The goal is to quantify uncertainty in forms that are directly relevant to reservoir operations and water-resources planning under nonstationarity.

This work develops hybrid statistical-dynamical frameworks to diagnose submonthly Gulf Coast sea-level variability, identify sources of predictability, and evaluate model error against CORA observations and CESM simulations. A central aim is to translate water-level anomalies into spatially explicit flood metrics, including duration, extent, and severity, while carrying uncertainty through the full analysis chain.

This project develops a constrained multi-sensor inversion that combines GNSS loading deformation, InSAR surface displacement, and GRACE terrestrial water storage to monitor groundwater change across Texas aquifer systems. Because each observing system has distinct errors and spatiotemporal sampling characteristics, the framework relies on physically informed regularization and explicit treatment of elastic and inelastic deformation to better constrain storage change from surface motion.

Examines the global dynamics and spatial coupling of terrestrial water storage extremes during 2002–2024, with emphasis on how large-scale climate variability organizes drought and flood conditions across regions.

Analyzes the hydrological and ecological impacts of the Kakhovka dam collapse using multi-sensor remote sensing and causal inference, with emphasis on water storage loss, river dynamics, flooding, and vegetation response.