Skip to main content

The Belmont Hydrology and Fine Sediment Lab: Research

Department of Watershed Sciences


My research group conducts a wide range of studies within the fields of hydrology, geomorphology and climate adaptation sciences including:

  • Hydrologic modeling and development of watershed water budgets
  • Modeling, measurement and analysis of landscape erosion, sediment transport, and deposition
  • Development of sediment budgets
  • Geochemical ‘sediment fingerprinting’ and analysis of cosmogenic nuclides to study Earth-surface processes and long-term erosion rates
  • Topographic analysis for feature extraction, geomorphic change detection and watershed classification
  • Sediment as a water quality metric in streams and rivers

A sediment budget is a quantitative accounting of sediment sources and sinks distributed throughout a watershed for a specified period of time. A suite of new research tools allow us to more easily and precisely date landforms, determine sediment provenance, and measure small changes on spatially extensive landforms. These tools show potential to revolutionize predictive watershed geomorphology by providing specificity in determining the locations, rates, and mechanisms of erosion, transport, and storage of sediment. A combination of new and traditional methods, extrapolated to the landscape scale in a geomorphically-sensitive manner, provides the first viable pathway toward reliable sediment yield forecasting and therefore more effective targeting of conservation efforts.

To develop sediment budgets my research group uses multiple, semi-redundant techniques, including aerial lidar analyses, repeat terrestrial lidar scans, geochemical fingerprinting, sediment dating techniques (cosmogenic nuclides, radiocarbon, and/or optically stimulated luminescence), air photo analyses, field surveys and mapping, and water and sediment gaging. Above is an example of a sediment budget that we could develop using such techniques.


Understanding where sediment comes from in a watershed and the pathways it travels during transport are intriguing basic science questions that have big implications for watershed management. Numerous techniques have been developed to identify sources from a sample of suspended sediment collected from a river. The common basis for these techniques is that different sources impart a slightly different mix of naturally-occurring geochemical tracers onto the sediment, which can be thought of as a geochemical ‘fingerprint’.

For example, the graph above shows tracer concentrations for two sediment sources, landslides (green dots), which contribute sediment that has low concentrations of both tracers, and hillslope soils (brown dots), which contribute sediment that has high concentrations of both tracers. Suspended sediment samples (red dots) collected from the river indicate the relative proportion of sediment derived from each source. Using radiogenic tracers with very different half-lives, such that one decays during floodplain storage and the other does not, allows us to determine the amount of sediment exchange that occurs between the channel and floodplains. In the figure, group A would be indicative of landslides contributing most sediment. Group B would be indicative of soil erosion contributing most sediment. Group C is indicative of soil erosion as the ultimate source, but a significant amount of channel-floodplain exchange having occurred, which dilutes the concentration of the short-lived tracer. For more information, see Belmont et al. (2011) Environmental Science and Technology.


The resolution of digital topography maps has increased 1000-fold over the past decade (from SRTM data with 90 m resolution to lidar data with sub-meter resolution in some cases). The enhanced resolution provides an extraordinary opportunity for extracting important information from topography to answer both basic and applied questions in geomorphology. For example, the two images below show an analysis of local relief (elevation differential) for the exact same part of a landscape using a conventional 30 meter DEM (30 m on each side of each pixel; left image) versus a 3 m DEM (right image). The additional detail included in the 3 m DEM allows for an entirely new range of questions that can be studied in this landscape.

30 m DEM


3 m DEM


High resolution topography data allows us to study the morphology and evolution of landforms. The image to the left was generated by digitally ‘flooding’ this river valley. Note the meandering mainstem channel (3.5-4 m water depth) as well as many older channels carved into the floodplains. The patchiness of the topography influences the distribution of vegetation as well as erosion and deposition of sediment.


High resolution topography data also allows us to examine the how topography is organized in a watershed, which helps us understand the basic processes by which physical processes sculpt landscapes. Gangodagamage, Belmont, and Foufoula (2011) used an ensemble statistics approach to generate the plot to the left, which shows how upstream contributing area changes with distance from a drainage divide in two basins in northern California. Three distinct regions are observed (A, B, and C) as differences in the scaling exponents (numbers shown). For example, Region B (which translates to the lower portions of hillslopes) are most convergent, drainage area increases as a function of distance taken to the power of 3. The scaling exponent of ~1.85 in Region C (fluvial channels) is consistent with a general rule theorized by John Hack in 1957. We are just beginning to investigate these questions regarding the organization of topography in landscapes and what it can tell us about geomorphic processes and how landscapes evolve over long- and short-timescales.