For over 100 years the rate of bedrock disintegration to erodable soil had been assumed to depend on the overlying thickness of the soil mantle. This assumption underlies some of the fundamental principals governing how hillslopes erode and therefore evolve under active processes of erosion. Importantly, the maximum rate of soil production defines the maximum erosion rate under which a landscape could remain soil mantled: for the soil to persist it has to be produced at a rate equal to or greater than the rate of erosion.
Here I propose the term "soil production function" for the quantitative relationship between the rate of bedrock conversion to soil and the overlying soil thickness. Soil production rates were thought to decline with increasing mantle thickness to follow the decreasing effectiveness of mechanical processes such as freeze-thaw or biogenic disturbance. There were no data quantifying this relationship. This project presents the first data for the rates of soil production as functions of soil depth across a wide variety of geologic, climatic, and tectonic settings and develops two new, independent, field-based methods to quantify the soil production function.
The first method uses topographic curvature as a surrogate for soil production under very specific field conditions. Field measurements of soil thickness and curvature measurements from high-resolution topographic surveys lead to determining the form of the soil production function.
The second method uses concentrations of the in situ produced cosmogenic radionuclides, 26Al and 10Be, extracted from quartz in bedrock at the base of the soil column. Measurements of the radionuclide concentrations enable determination of soil production rates under the full range of soil depths to fully define the soil production function.
Soil production functions were quantified in my dissertation research from two field sites in northern California, one in the Oregon Coast Range, and two in southeastern Australia and the implications of the soil production functions on landscape evolution for each of the field sites were explored. Radionuclide analyses of stream sediments from each of the field sites are used to determine catchment-averaged erosion rates for comparison with the soil production rates. Soil production and average erosion rates from an additional site in the San Gabriel Mountains of southern California are presented as an upper end on the range of erosion rates from soil-mantled landscapes.
These data are some of the only field-based data on soil production, landscape lowering, and average erosion rates around. They therefore provide important parameters for landscape evolution models as well as a first-order comparison for further research that will quantify contemporary erosion rates as a function of land-use and climate changes.
I am working with John Chappell and Nigel Spooner at the Research School of Earth Sciences at Australian National University in Canberra, Australia to develop a new technique for measuring the rates and processes of hillslope erosion. Soil on hilly landscapes is produced and transported by a variety of processes (e.g. bioturbation, soil creep, solifluction, slope wash, and landsliding). Understanding the mechanisms controlling hillslope erosion is crucial for assessing the impact of timber harvesting, grazing, and other land uses. Additionally, the importance of quantifying these processes for drainage basin modeling is not new and there are techniques to measure the contemporary or long-term erosion rates. Measurements of in situ produced cosmogenic radionuclide concentrations were used in the above studies to determine rates for bedrock erosion, basin-averaged erosion, and soil production.
As powerful as the radionuclide method is, it will not necessarily enable the mode of sediment transport or short-term erosion rates to be determined. Optically Stimulated Luminescence (OSL) has recently become the established successor to thermoluminescence (TL) as the means for measuring the time since the trapped charge population in crystalline minerals was reset. OSL measures the last time of exposure to sunlight of mineral grains, is now being developed for use on single quartz grains, and offers a unique tool for determining erosion rates. We use OSL on soil profiles at different positions across landscapes where the above research has defined the soil production function with cosmogenic nuclides. The results have so far provided the only depth profiles of soil mixing times and, when combined with the soil production function, residence times with an application of a cutting edge research tool. Indeed, at this point, the technique of measuring the luminescence of single grains of quartz is developing in parallel with our geomorphic applications of the technique. Results from these depth profiles as well as hillslope transects across the full range of field sites will lead toward defining field-based sediment transport functions. In combination with the soil-production function, such transport laws form the basis of process geomorphology.
I am also pushing ahead with my graduate student, Jim Kaste, on developing measurements of short-lived isotopes in soil and stream sediments to get at new constraints on hillslope erosion rates. More on this later, but in brief we're measuring 7Be, 210Pb, 137Cs, and a host of other short-lived isotopes and using analyses similar to the modeling done with my OSL results to understand how surficial processes are actually operating.
With Douglas Burbank (Penn State), Anna Barros (Harvard), Ann Blythe (USC), Stephen Fisher (Film), Kip Hodges (MIT), Neil Humphrey (U of Wyoming), Jerome Lave (Grenoble), Jaakko Putkonen (U of Washington))
This collaborative project is a natural path for my research that builds on my Master's research project in Nepal that sought to quantify erosion rates from forested and agricultural watersheds as well as the above stated use of cosmogenic nuclides and OSL to get at erosion rates. One of the most provocative (yet largely untested) recent hypotheses concerning orogenic evolution is that regional variations in climate strongly influence spatial variations in the style and magnitude of deformation across an actively deforming orogen. Recent progress in quantifying rates of both tectonic and geomorphic processes and in modeling surface and lithospheric processes sets the stage for an integrated, quantitative, field- and model-based investigation of the interactions and feedbacks between geomorphic, climatic, and tectonic processes.
We will examine these interactions where they are likely to be most clearly expressed: the Nepalese Himalaya. Not only is this the quintessential collisional orogenic belt, but its topographic growth and erosional history have been suggested as key controls on global climatic changes. Our integrated study focuses on a major transverse catchment, stretching from the edge of the Tibetan Plateau to the foreland and traversing some of the highest topography in the world. This transect spans the major structural elements of the Himalaya, as well as monsoon-to-rain shadow climatic conditions. We bring together expertise in process-based geomorphology, glaciology, climatology, structural geology, thermochronology, cosmogenic radionuclide dating, modeling, and documentary film making for a multi-pronged approach intended to evaluate one overarching, but largely untested hypothesis: Rates of erosion vary spatially as a function of climate and this spatial variability in erosion controls the partitioning of deformation within an orogen.
Furthermore, we will collect data to assess the following related, but subsidiary hypotheses: The erosional response to rapid lateral advection of crust across a basement ramp crustal scale fault-bend folding, for example, creates erosion rates that are nearly equal across the entire topographic escarpment of the Himalaya, ranging from 8 km to 1 km in elevation. Above a certain threshold erosion rate, the topography attains a dynamic equilibrium or steady state that is independent of erosion rate. Topographic characteristics (relief, slope angles, normalized river gradients) correlate more strongly with erosion rates than they do with variations in climate or lithology. Despite the broad scope of these hypotheses and the impossibility of resolving all details, we have developed a research strategy that, over a four-year span, will enable us to define the primary characteristics of denudation, rock uplift, climate, and topography across the Himalaya and to calibrate some process-based ărulesä for major erosional agents, such as glaciers, rivers, and landslides.
A key to success will be the integration of data from diverse subdisciplines (climate, geomorphology, tectonics) at the scale both of intensively monitored subcatchments and of the entire trans-Himalayan catchment. Spanning seven subdisciplines in earth and atmospheric sciences, this project brings together researchers from seven US institutions and three governmental agencies in Nepal.
The above project that will quantify erosion and river incision rates on a transect across the Nepal Himalaya drove me to extend the work into Tibet where I've collected preliminary samples for nuclide analyses. My intention is to compare rates from these samples with rates from samples collected across the much more slowly eroding Australian Outback to determine field-based constraints on how bedrock is eroding and the rate at which rivers are incising into very different landscapes. This project fits into a global comparison of erosion rates and how they relate to climate, topography, and related factors that our field continues to be fundamentally interested in. Some specific problems to be tackled here include: 1) what sets the upper bound of bedrock erosion for a given climate and tectonic regime? 2) how rapidly do bedrock escarpments "retreat", or propagate away from a passive margin? 3) how quickly do rivers incise through bedrock and how does that compare with the above? 4) how do these rates compare across different landscapes? 5) what effect do humans have on these rates?
Ronald Amundson (UC Berkeley) and I are testing the hypothesis that natural rates of soil carbon erosion in upland ecosystems may be a significant atmospheric CO2 sink. If the hypothesis is true, then the possible magnitude of the process may represent a significant fraction of the undetermined loss of atmospheric CO2. The scientific community is trying to understand the global C budget so that the effects of anthropogenic CO2 emissions can be accurately determined.
Presently, a significant fraction of fossil fuel-produced CO2 is disappearing into unknown sinks. We hypothesize that soils in upland ecosystems, particularly forests, may be part of this sink. If erosion is removing a fraction of the annual C inputs by plants, and if the eroded C is buried nearby in depositional settings, then soils on uplands may be an unrecognized and on-going sink of atmospheric CO2. For this work we're using a mathematical model, developed from the above work on the soil production function, to calculate soil erosion rates as a function of slope position in upland ecosystems.
We will apply the model to a suite of research sites in western North America and Australia that span a breadth of climate and geology. For nearly all the sites, the rates of soil production (a key model parameter) have been determined by cosmogenic radionuclides in the work summarized above. It is this background work that makes this project feasible. In addition to calculating C loss, we will empirically quantify, via field and laboratory work, the location and rate of local sediment storage at all the sites. This combined research should allow us to formally evaluate our hypothesis that erosional loss of soil C is a significant process in upland ecosystems.
Quantifications of Earth surface topography are essential for modeling the connections between physical and chemical processes of erosion and the shape of the landscape. Enormous investments are made in developing and testing process-based landscape evolution models. These models may never be applied to real topography because of the difficulties in obtaining high-resolution (1÷2 m) topographic data in the form of digital elevation models (DEMs).
In this project, with a computer science colleague, Hany Farid, at Dartmouth, we present a simple methodology to extract the high-resolution three-dimensional topographic surface from photographs taken with a hand-held camera with no constraints imposed on the camera positions or field survey. This technique requires only the selection of corresponding points in three or more photographs. From these corresponding points the unknown camera positions and surface topography are simultaneously estimated. We compare results from surface reconstructions estimated from high-resolution survey data from field sites in the Oregon Coast Range and northern California to verify our technique. Our most rigorous test of the algorithms presented here is from the soil-mantled hillslopes of the Santa Cruz marine terrace sequence.
Results from three unconstrained photographs yield an estimated surface, with errors on the order of 1 m, that compares well with high-resolution GPS survey data and can be used as an input DEM in process-based landscape evolution modeling. Two undergraduate students will be traveling to New Zealand this winter to apply this methodology to the problem of quantifying landslide magnitude in hard-to-reach areas. While the ideas for this project were germinated in Australia and Nepal, it's exciting that the blossoms were harvested in the midst of teaching a full course load last Spring ('02). We are making our algorithm freely available and encourage use of it in the hopes of fostering collaborative efforts.