Estimating the fill thickness and bedrock topography in intermontane valleys using artificial neural networks

Abstract

Thick sedimentary fills in intermontane valleys are common in formerly glaciated mountain ranges but difficult to quantify. Yet knowledge of the fill thickness distribution could help to estimate sediment budgets of mountain belts and to decipher the role of stored material in modulating sediment flux from the orogen to the foreland. Here we present a new approach to estimate valley fill thickness and bedrock topography based on the geometric properties of a landscape using artificial neural networks. We test the potential of this approach following a four-tiered procedure. First, experiments with synthetic, idealized landscapes show that increasing variability in surface slopes requires successively more complex network configurations. Second, in experiments with artificially filled natural landscapes, we find that fill volumes can be estimated with an error below 20%. Third, in natural examples with valley fill surfaces that have steeply inclined slopes, such as the Unteraar and the Rhône Glaciers in the Swiss Alps, for example, the average deviation of cross-sectional area between the measured and the modeled valley fill is 26% and 27%, respectively. Finally, application of the method to the Rhône Valley, an overdeepened glacial valley in the Swiss Alps, yields a total estimated sediment volume of 97 ± 11 km 3 and an average deviation of crosssectional area between measurements and model estimates of 21.5%. Our new method allows for rapid assessment of sediment volumes in intermontane valleys while eliminating most of the subjectivity that is typically inherent in other methods where bedrock reconstructions are based on digital elevation models.

Glacial isostatic uplift of the European Alps

“…The isostatic adjustment to deglaciation was likely attenuated by the postglacial accumulation of sediments in these valleys8. We used an artificial neural network (ANN) algorithm11 to estimate the sediment thickness within all Alpine valleys (see ‘Methods’ section; Fig. 3a).…”

“…Based on the assumption of geometric similarity between the exposed and the buried parts of the landscape, we used an ANN algorithm11 and a 90- m-resolution digital elevation model (DEM) to explicitly estimate the depth to bedrock for grid cells that include valley fill. We expect geometric similarity of the bedrock surface, because the entire landscape was subject to glacial erosion before deposition of the valley fill.…”

“…Some of these processes, such as lithospheric delamination, manifest themselves on timescales of ∼10 6 –10 7 years, whereas others, such as postglacial rebound, occur relatively rapidly (∼10 3 years). New approaches to modelling orogen-scale sediment storage11, glaciation12 and spatial variations in EET (ref. 13) provide new constraints for estimating the contribution of glacial isostatic adjustment (GIA) to present-day uplift rates in the European Alps.…”

Toward improved prediction of the bedrock depth underneath hillslopes: Bayesian inference of the bottom‐up control hypothesis using high‐resolution topographic data

“…Topographic variables (terrain and landform), bedrock properties (geology and geochemistry), and climatologic characteristics (radiation, precipitation, and temperature) have all been used as predictors of the regolith depth in regression models [ DeRose et al ., ; Boer et al ., ; Ziadat , ; Wilford and Thomas , ; Yang et al ., ]. Other regression‐type methods published in the geomorphologic literature include the use of artificial neural networks [ Zhou and Wu , ; Mey et al ., ], principal component analysis, and maximum likelihood classification [ Boer et al ., ; Ziadat , ], canonical correspondence analyses [ Odeh et al ., ], support vector machines [ Sitharam et al ., ], and generalized additive models and random forests [ Tesfa et al ., ; Shafique et al ., ]. These latter two methods use secondary data of land cover and other soil attributes derived from remote sensing products.…”

Long‐Term Storage and Age‐Biased Export of Fluvial Organic Carbon: Field Evidence From West Iceland

“…To estimate the areal extent of sediment fill in the river catchment (as in Blöthe & Korup, ; Mey et al, ), we analyzed 200 evenly spaced valley cross sections. For this analysis, we split the valley into an upper and lower portion based on the dominant orientation of the valley axis and analyzed an equal number of cross sections in each region.…”