While I've enjoyed collaborating on multiple topics, I have had four main research fields to-date:
1) Global-scale geospatial data analysis to answer old questions in geomorphology and sedimentology using new tools.
2) Numerical modeling using my own cellular model to study how river and floodplain landscapes change over large areas and long timescales.
3) Drone-based lidar to study how river landscapes change over small areas and short timescales.
4) Sedimentology and stratigraphy to study the deposits left behind by ancient river and delta landscapes.
I find excuses to incorporate field work into 3) and 4). Scroll down for more detail on each!
1) Geospatial data analysis
We are in the middle of a modern scientific revolution. For the first time in history, spatial data are common, comprehensive, and abundant. Each year new satellites are launched to collect exciting novel observations, new tools are developed to convert them into natural properties, and new explanations are found for how the world works. My postdoc at Caltech has focused on harnessing and harmonizing these global-scale datasets to distill new answers for formerly unanswerable questions.
I have a few of these projects in the pipeline, but the first was just published in Geology. In it, we took a new look at the classic source-to-sink framework. It's a conceptual model for understanding landscapes and stratigraphic units by focusing on sediment transport. It classifies areas into those that produce sediment via erosion (sources), places that transport sediment (bypass), and those that sequester sediment to build the rock record (sinks). The source to sink diagram is one of the first figures you will see in introductory Earth Science textbooks (above, drawn in 1-dimension (A) and 2.5-dimensions (B)). The source to sink framework explains what parts of the modern Earth end up in the rock record, and it defines some of the largest fluxes and pools in the global carbon cycle. Yet, it had never been tested over the whole world surface. What is the relative abundance of sources, bypass zones, and sinks? Even more simply: Does the world look like a source-to-sink diagram? Why or why not?
You can see the global map above, with some inset images (i - iii). You can see a zoomable online copy at my website:
https://sourcetosink.mapsof.rocks
Overall, the story is that about half of Earth's non-Antarctic land is a sediment source, one-quarter is a bypass zone, and one-quarter is a sediment sink. Including oceans and Antarctica changes the ratio to 13% source, 5% bypass, 82% sink. Some places in the world really do look like source-to-sink diagrams. Many are passive margins (i), like the eastern coast of North America... not coincidentally, the same place that source-to-sink studies were born. Other places don't look like source-to-sink diagrams, though. Look at the world's major deserts, or formerly glaciated regions. We found that anywhere that lacks major rivers tends not to resemble source-to-sink diagrams. We argue that while the source-to-sink framework is useful, it needs a fourth category for "inactive" regions before it can be used to explain the entire world.
https://sourcetosink.mapsof.rocks
Overall, the story is that about half of Earth's non-Antarctic land is a sediment source, one-quarter is a bypass zone, and one-quarter is a sediment sink. Including oceans and Antarctica changes the ratio to 13% source, 5% bypass, 82% sink. Some places in the world really do look like source-to-sink diagrams. Many are passive margins (i), like the eastern coast of North America... not coincidentally, the same place that source-to-sink studies were born. Other places don't look like source-to-sink diagrams, though. Look at the world's major deserts, or formerly glaciated regions. We found that anywhere that lacks major rivers tends not to resemble source-to-sink diagrams. We argue that while the source-to-sink framework is useful, it needs a fourth category for "inactive" regions before it can be used to explain the entire world.
|
As always, I found a way to bring the research back towards rivers. To the left you can see results from a global analysis of ~10 million km of rivers on Earth, where we split reaches up based on whether they were in areas of erosion, bypass, or deposition. Contrary to some previous thought, we showed that the largest river reaches in the world will disproportionately be preserved in the rock record. We also quantified the cutoff point for how large they have to be: rivers with stream order [a proxy for river size] ≤ 3 are disproportionately less likely to end up in the rock record, while those ≥ 3 are more likely.
|
Publication (feel free to contact me if you can't access it):
- Martin, H.K. and Lamb, M.P., (2025). The unexpected global distribution of sediment sources and sinks: Geology, in press. DOI: 10.1130/G53289.1.
2) Numerical modeling
|
Much of the sedimentary rock record around the world is made up of deposits from river systems in slowly subsiding mountain-front regions. These systems redistribute sediment through river avulsions, where rivers suddenly change course on landscapes and take up new pathways on their former floodplains. I became interested in what determines when, where, and why these avulsions occur, and how that affects landscapes. To the right, you'll see some sketches I did as part of planning; I enjoy drawing elements of my models.
|
|
|
|
I developed a model ("RiverWalk") in MATLAB that pairs a 1-D diffusion model (to describe how sediment moves along a river) to a 2-D cellular model (to describe how avulsions pathfind and what happens to the rest of the floodplain). As it turns out, the abandoned channels left behind by previous avulsions have an important role to play in determining when, where, and why future avulsions occur! They also help shape landscapes over long timescales: we found a way to explain why only some rivers form fluvial fans, which is a decades-old scientific conundrum. For more details, see the manuscripts below.
|
Publications (feel free to contact me if you can't access any):
First-authored:
Co-authored:
First-authored:
- Martin, H.K. and Edmonds, D.A., (2023). Avulsion dynamics determine fluvial fan morphology in a cellular model: Geology, v. 51(8): 796–800. DOI: 10.1130/G51138.1.
- Martin, H.K. and Edmonds, D.A., (2022). The push and pull of abandoned channels: How floodplain processes and healing affect avulsion dynamics and alluvial landscape evolution in foreland basins: Earth Surface Dynamics, v.10(3). DOI: 10.5194/esurf-10-555-2022.
- GITHub code repository: https://doi.org/10.5281/zenodo.5576789.
- GITHub code repository: https://doi.org/10.5281/zenodo.5576789.
Co-authored:
- Valenza, J., Edmonds, D., Martin, H., Sifuentes, C., and Toby, S., (2024). Stratigraphic architecture of fluvial fans shaped by downstream changes in avulsion style: Sedimentology, DOI: 10.1111/sed.13217.
- Jeffery Valenza, a former PhD candidate at Indiana University, demonstrated stratigraphic changes in avulsion style in different locations on fluvial fans using a combination of stratigraphy fieldwork and remote sensing. I contributed a modeling effort that demonstrates that these patterns of change also appear in simulations, where they can be analyzed more-quantitatively. This was the final chapter of his PhD and it is out at Sedimentology. Congrats Jeff!
- Sifuentes, C., Martin, H.K., Straub, K.M., Hajek, E.A., and Edmonds, D.A., (2025). Floodplain topography and avulsion pathfinding control stratigraphic architecture in a numerical model of a fluvial fan: Journal of Sedimentary Research, v. 95(1), p. 209-222. DOI: 10.2110/jsr.2024.045.
- Caitlin Sifuentes, a former MSc student at Indiana University, and I built a stratigraphic module for the avulsion model that works to translate the surface simulation of RiverWalk into the subsurface rock record. She found some exciting emergent behaviors that show up re: compensation, and published her MSc thesis at the Journal of Sedimentary Research. Congrats Caiti!
3) Drone-based lidar
For geomorphologists, many of our questions begin or end with mapping the surface of the earth. Lidar (like radar but with lasers) allows us to make very accurate and precise maps even through vegetation! While most folks collect lidar from ground-based tripods or aircraft, our lab is fairly unique in that we have a drone-mounted lidar set-up, allowing us to rapidly deploy wherever and whenever we need.
|
|
|
|
So far I've applied drone-based lidar to answer two research questions. The first: what are the effects of sudden, catastrophic dam failures on downstream landscapes? We collected scans three weeks, three months, and one year after the 2020 failure of two dams in mid-Michigan. The second question is: how and why do meandering rivers tend to maintain width as they move across landscapes. To answer this, we've collected 20 drone-based lidar scans of a point-bar and cutbank near Worthington, IN over the last four years. Combined with bathymetry and field work, we have a wholistic dataset for river migration.
|
Publications (feel free to contact me if you can't access any):
First-authored:
Co-authored:
First-authored:
- Martin, H.K., Edmonds, D.A., and Lewis, Q.L., 2024. Four years of meandering captured by drone-based lidar surveys reveal lack of width maintenance on the White River, Indiana, USA: Journal of Geophysical Research: Earth Surface, v. 129(6). DOI: 10.1029/2023JF007574.
- Martin, H.K., Edmonds, D.A., Yanites, B.J., and Niemi, N.A., 2024. Quantifying landscape change following catastrophic dam failures in Edenville and Sanford, Michigan, USA: Earth Surface Processes and Landforms, p. 1-12. DOI: 10.1002/esp.5855.
Co-authored:
- Romain, W.F., Herrmann, E.W., Martin, H.K., Barefoot, E.A., and Scott, S., (2025). High resolution lidar drone imagery assessment of the Rattlesnake Mound Complex at Cahokia, Illinois, USA: Midcontinental Journal of Archaeology, v. 50(1), p. 31-62. DOI: 10.5406/23274271.50.1.02.
- For this study, I collaborated with two geoarchaeologists from Indiana University to study the Cahokia Mounds World Heritage & State Historic Site in what is now southwest Illinois, USA. Cahokia was the largest pre-Columbian city north of Mexico. It was occupied ~1050-1350 CE and at its peak had a larger population than London. We used drone-based lidar to measure the shape of mound landscapes, including in a densely treed area that had previously not been well-investigated. Our archaeology colleagues combined these modern data with historic records to quantify erosion from early European looting and investigate alignments between mounds and astronomical bodies, giving us a hint at the region's cultural beliefs and burial practices.
4) Sedimentology & stratigraphy
Under the surface of the earth, the rocks are lain down in layers, in a language that tells stories of landscapes never-before-seen by humans. Learning to read this language is the study of stratigraphy: just how did each layer get laid down, in which order, and what does that tell us about what the earth looked like back then?
During my MSc under Dr. Stephen Hubbard at the University of Calgary, I spent years in a deeply collaborative environment. As data, we primarily used drill-cores (tubes of sediment or rock from hundreds of meters underground that are brought up to the surface intact, like an apple corer) and wireline logs (geophysical tools that scan the inside of well boreholes after drilling is done). I interpreted 45 drill cores (>3 km) and ~4,500 wireline logs (>200 km) to develop a stratigraphic framework (a sort of guide to interpreting the rocks) for an area >6,500 km². I also mapped cross-cutting river valleys, deltas, channel belts, and flooding surfaces for this whole area: 18 distinct time-surfaces, in all, for one ~115 million year old formation. I then interpreted these units to come up with a depositional timeline and an interpretation of the ancient paleoenvironment at the time the rocks were deposited.
Publications (feel free to contact me if you can't access any):
First-authored:
Co-authored:
First-authored:
- Martin, H.K., Hubbard, S.H., Hagstrom, C.A., Horner, S.C., and Durkin, P.R., (2019). Planform Recognition and Implications of a Cretaceous-age Continental-scale River Avulsion Node in the Western Interior Basin, Alberta, Canada: Journal of Sedimentary Research, v. 89(7), p. 610-628. DOI: 10.2110/jsr.2019.37.
Co-authored:
- Peng, Y., Hagstrom, C.A., Horner, S.C., Leckie, D.A., Martin, H.K., Durkin, P.R., and Hubbard, S.M., 2024. Early Cretaceous evolution of the McMurray Formation: A review towards a better understanding of the paleo-depositional system: Earth-Science Reviews, v. 252, 104740. DOI:10.1016/ j.earscirev.2024.104740.
- Hagstrom, C.A., Hubbard, S.M., Horner, S.C., Martin, H.K., and Peng, Y., 2023. Comparison of the morphology, facies, and reservoir quality of valley fills in the southern Athabasca Oil Sands Region, Alberta, Canada: AAPG Bulletin v. 107(4), p. 553-591. DOI: 10.1306/10242219118.
- Peng, Y., Hagstrom, C.A., Horner, S.C., Hodgson, C., Martin, H.K., Leckie, D.A., Pedersen, P.K., and Hubbard, S.M., (2022). Low-accommodation foreland basin response to long-term transgression: A record of change from continental-fluvial and marginal-marine to open-marine sequences over 60,000 km2 in the western Canada foreland basin: Marine and Petroleum Geology, v. 139. DOI: 10.1016/j.marpetgeo.2022.105583.
- Horner, S., Hubbard, S., Martin, H.K., Hagstrom, C., and Leckie, D., (2019). The impact of Aptian glacio‐eustasy on the stratigraphic architecture of the Athabasca Oil Sands, Alberta, Canada: Sedimentology, v. 66(5), p. 1600-1642. DOI: 10.1111/sed.12545.
- Horner, S.C., Hubbard, S.M., Martin, H.K., and Hagstrom, C.A., (2019). Reconstructing basin-scale drainage dynamics with regional subsurface mapping and channel-bar scaling, Aptian, Western Canada Foreland Basin: Sedimentary Geology, v. 385, p. 26-44. DOI: 10.1016/j.sedgeo.2019.03.012.
Harrison K. Martin
Copyright © 2025
