Efi Foufoula-Georgiou

Efi Foufoula-Georgiou’s major research interests are in the area of stochastic modeling of surface hydrologic and geomorphologic processes. Current areas of research include: (1) modeling and estimation of space-time rainfall from multiple sensors, including downscaling and merging of satellite and ground-based observations; and (2) stochastic theories of transport on the earth's surface, including new classes of geomorphic transport laws, river network dynamics under changing forcing, channel morphology, and hydrologic response. All these research topics have the common thread of exploring space-time statistical signatures over a range of scales and relating them to the underlying physical processes. Modeling is pursued using minimal complexity models that explore re-normalization and patterns.

Multi-sensor space-time precipitation for hydrologic and eco-geomorphologic applications
As part of Foufoula’s NASA-funded research, which explores the use of Global Precipitation Measuring (GPM) observations for hydrologic and geomorphologic applications, her group is studying: (1) characterization of the space-time multiscale structure of orographic precipitation and its relation to topography and atmospheric dynamics for the purpose of downscaling remote sensing observations over complex terrain for flood and landslide hazard prediction; (2) new methodologies for merging  multi-sensor and/or multi-model precipitation products of varying uncertainty and at multiple space-time scales. The emphasis is on reproducing the location and frequency of extremes relevant to hazard prediction and control, under current and future climates.

Stochastic theories of geomorphic transport and landscape evolution
Foufoula’s research group explores new methodologies for quantifying the stochastic nature of sediment production and transport in hillslopes, channels, and channel networks using multiscale analysis and dynamical system theory. Her group seeks to understand the relationship between near bed turbulence, river bed morphodynamics, and sediment/tracer transport in rivers using experimental, theoretical, and numerical research. The group is also exploring new classes of geomorphic transport models based on the concept of non-locality which explicitly acknowledges the broad range of space-time scales involved in the transport.  Extensions of the concept of non-locality to formulating new classes of landscape evolution models subject to space-time variable forcing are currently under investigation. 

Hydro-geomorphic feature extraction from LiDAR
Foufoula’s group explores the use of advanced mathematical formalisms for automatic extraction of features of hydrologic/geomorphologic interest (channel heads, channel networks, channel and floodplain morphology, engineering structures, bluffs, terraces, landslide scars, etc.) from high resolution (1 m) LiDAR topography. A recently proposed formalism, packaged under the name GeoNet, is based on two innovations:  (1) a localized nonlinear filtering operation, using nonlinear partial differential equations, which automatically adapts to the terrain and amplifies edges while removing noise, and (2) a global-to-local geodesic optimization operation which robustly extracts continuous paths of transport, based on geomorphologically relevant cost functions. New research is exploring the detailed interrogation of landscapes, from the drainage divides to the basin outlet, for identifying regime transitions and unifying scaling laws depicting basin organization.

Network envirodynamics and scaling
River networks are known to structure most dynamical processes that take place in a river basin, from water, to sediment, to biomass transport. However, how this structuring takes place and how it can be quantified in space and time is a subject of continuous research. Adding to their previous research on the scaling of hydraulic geometry and floods and on how the river network structures bedload sediment, Foufoula’s group is now studying questions of dynamical processes operating on hierarchical trees. The ultimate objective is to develop a generalized conceptual framework for scaling fluxes (water, sediment, biologic activity) at any point of the basin, subject to space-time variable forcing that activates the network intermittently and dynamically.