Assessing the impact of glaciation at the Earth's surface requires simultaneous quantification of the impact of climate variability on past glacier fluctuations and on bedrock erosion. Here we present a new approach for evaluating post-glacial bedrock surface erosion in mountainous environments by combining two different surface exposure dating methods. This approach can be used to estimate how bedrock erosion rates vary spatially and temporally since glacier retreat in an alpine environment.

We apply machine learning techniques to quantify and explain differences between recent palaeoclimates with regards to factors that are important in shaping the Earth's surface. We find that changes in ice cover, near-surface air temperature and rainfall duration create the most distinct differences. We also identify regions particularly prone to changes in rainfall and temperature-controlled erosion, which will help with the interpretation of erosion rates and geological archives.

We use a statistical earthquake generator to investigate the influence of fault activity on river profile development and on the formation of co-seismic knickpoints. We find that the magnitude distribution of knickpoints resulting from a purely seismic fault is homogeneous. Shallow aseismic slip favours knickpoints generated by large-magnitude earthquakes nucleating at depth. Accounting for fault burial by alluvial cover can modulate the topographic expression of earthquakes and fault activity.

Slow-moving landslides are a major hazard but their movement mechanisms during earthquakes and rainstorms are not fully understood. We collected samples from a slow-moving landslide complex in New Zealand and subjected them to a range of porewater pressure and dynamic stress scenarios in a dynamic back-pressured shear box. Our results show how the complex movement patterns, observed in many large slow-moving landslides, may be mobilized by strong earthquakes and significant rain events.

Sand and mud may take many different pathways within a river as they travel from inland to the coast. During the trip, grains may be exposed to daylight, resetting a signal trapped within certain minerals. The signal can be measured in a laboratory to estimate the time since last light exposure. Here, we measure the trapped signal of sand and mud grains from the Mississippi River and its banks. We use this information to infer sediment pathways. Such knowledge is useful for delta management.

Lakes and swales are real landscape features but are generally treated as data errors when calculating water flow across a surface. This is a problem because depressions can store water and fragment drainage networks. Until now, there has been no good generalized approach to calculate which depressions fill and overflow and which do not. We addressed this problem by simulating runoff flow across a landscape, selectively flooding depressions and more realistically connecting lakes and rivers.

The age of formation of sedimentary deposits is often interpreted to record information about past environmental changes. Here, we show that the timing of abandonment of surfaces also provides valuable information. We derive a new set of equations that can be used to estimate when a sedimentary surface was abandoned based on what is known about its activity from surface dating. Estimates of abandonment age can benefit a variety of geomorphic analyses, which we illustrate with a case study.

River deltas are valuable resources that support biodiversity and human habitation. Despite this we do not have a global census of deltas nor do we know the conditions that promote their formation. We surveyed 5399 river mouths greater than 50  m wide and found that 2174 (40  %) create a delta. The conditions that lead to delta formation are high sediment input and low wave and tide conditions. These results can be used to understand how deltas will adapt to environmental changes.

Researchers studying gravel bed rivers almost always require a sample of the bed surface grain size range. These samples are typically expressed as cumulative frequency distributions. We present statistical techniques for generating and plotting confidence intervals for the various size percentiles and for determining whether size percentiles from two samples are statistically different. The techniques are implemented in an R package; a simplified version is implemented in a spreadsheet.

We evaluated the performance of a drone system to reconstruct 3-D topography. We used a direct georeferencing method to make the pictures have precise coordinates, which also improves the survey efficiency. With both consumer-grade and professional-grade camera and drone setups, we obtained centimetric accuracy, which provides a flexible application in topography remote sensing using drones.

Large earthquakes trigger thousands of landslides in the area of their epicentre. For three earthquake cases, we have determined the position of these landslides along hillslopes. These co-seismic landslides tend to cluster at ridge crests and slope toes. We show that crest clustering is specific to seismic triggering. But although co-seismic landslides locate higher in the landscape than rainfall-induced landslides, geological features strongly modulate their position along the hillslopes.

Large wood (LW) pieces are often part of fish habitat restoration projects. We investigated reach-scale changes after the addition of LW that are relevant to juvenile coho salmon. A survivable habitat for juvenile coho was characterized in terms of critical swim speed and bed stability. Model predictions showed that survivable habitat increased by 86–128 % in terms of flow velocity and bed stability. Our findings are applicable to stream restoration efforts throughout the Pacific Northwest.

We develop and test new methods for counting pebble-size distributions in photos of gravel-bed rivers. Our open-source algorithms provide good estimates in complex imagery from high-energy mountain rivers. We discuss methods of river cross-section photo collection and processing into seamless georeferenced imagery. Application of a semi-automated version of the algorithm in small patches can be used as validation data for upscaling to entire survey sites using a fully automated version.

In some settings, landslides trigger valley blockages that impound huge volumes of sediment, often drastically changing river habitat and habitability. In other settings, landslides appear to have little effect on rivers. In this study, we explore what governs the different sensitivity of rivers to blocking from landslide debris. We accomplish this by comparing two sites in California with dramatic differences in blocking from otherwise similar slow-moving landslides.