Journal Paper Digests

Reading time ~4 minutes

Journal Paper Digests 2019 #12

  • Physically Based Model for Extracting Dual-Permeability Parameters Using Non-Newtonian Fluids
  • Scale Issues in Soil Hydrology
  • Strength matters: Resisting erosion across upland landscapes
  • Using deep learning for multivariate mapping of soil with quantified uncertainty

Physically Based Model for Extracting Dual-Permeability Parameters Using Non-Newtonian Fluids

Authors: Basset, CN; Abou Najm, MR; Ammar, A; Stewart, RD; Hauswirth, SC; Saad, G

Source: VADOSE ZONE JOURNAL, 18 (1):80172-80172; JUL 11 2019

Abstract: Dual-permeability models simulate flow and transport within soils characterized by preferential (macro) and matrix (micro) pore domains, with each exhibiting distinct hydraulic properties. The lack of suitable methods for determining appropriate and physically based model parameters remains a major challenge to applying these models. Here, we present a method that characterizes dual-permeability model parameters using experimental results of saturated flows from water and a non-Newtonian fluid. We present two submodels that solve for the effective pore sizes of micropores and macropores, with macropores represented either with cylindrical (for biological pores) or planar (for shrinkage cracks and fissures) pore geometries. The model also determines the percentage contribution (wi) of the representative macro-and micropores to water flow. We applied the model to experimental soil samples complemented with capillary tubes simulating the macropores and showed its ability to deri ve the bimodal pore size distributions in dual-domain soils using only two fluids. As such, we present this method of characterization of dual structures for improved modeling of nonuniform preferential flow and transport in macroporous soils.

Scale Issues in Soil Hydrology

Authors: Vogel, HJ

Source: VADOSE ZONE JOURNAL, 18 (1):90001-90001; JUL 11 2019

Abstract: Soil hydrology is a key control for the functioning of the terrestrial environment. Many environmental issues that we need to tackle today are directly linked to soil water dynamics. This includes agricultural production and food security, nutrient cycling and carbon storage, prevention of soil degradation and erosion, and last but not least, clean water resources and flood protection. However, these problems need to be addressed at the scales of fields, regions, and landscapes, while soil water dynamics and soil hydraulic properties are well understood and typically measured at much smaller scales-the comfort zone of soil physics. An obvious problem is how to link these vastly different scales and how to profit from small-scale understanding to improve our capability to predict what is going on at the large scale. In this update, this problem is discussed based on insights gained during the last decades. As a synthesis, a two-step scaling approach is proposed for modeling so il water dynamics from local to landscape scales where the scale of the soil profile is the stepping stone.

Strength matters: Resisting erosion across upland landscapes

Authors: Heimsath, AM; Whipple, KX

Source: EARTH SURFACE PROCESSES AND LANDFORMS, 44 (9):1748-1754; JUL 2019

Abstract: Soil-covered upland landscapes comprise a critical part of the habitable world and our understanding of their evolution as a function of different climatic, tectonic, and geologic regimes is important across a wide range of disciplines. Soil production and transport play essential roles in controlling the spatial variation of soil depth and therefore hillslope hydrological processes, distribution of vegetation, and soil biological activity. Field-based confirmation of the hypothesized relationship between soil thickness and soil production is relatively recent, however, and here we quantify a direct, material strength-based influence on variable soil production across landscapes. We report clear empirical linkages between the shear strength of the parent material (its erodibility) and the overlying soil thickness. Specifically, we use a cone penetrometer and a shear vane to determine saprolite resistance to shear. We find that saprolite shear strength increases systematically with overlying soil thickness across three very different field sites where we previously quantified soil production rates. At these sites, soil production rates, determined from in situ produced beryllium-10 (Be-10) and aluminum-26 (Al-26), decrease with overlying soil thickness and we therefore infer that the efficiency of soil production must decrease with increasing parent material shear strength. We use our field-based data to help explain the linkages between biogenic processes, chemical weathering, hillslope hydrology, and the evolution of the Earth’s surface.

Using deep learning for multivariate mapping of soil with quantified uncertainty

Authors: Wadoux, AMJC

Source: GEODERMA, 351 59-70; OCT 1 2019

Abstract: Digital soil mapping (DSM) techniques are widely employed to generate soil maps. Soil properties are typically predicted individually, while ignoring the interrelation between them. Models for predicting multiple properties exist, but they are computationally demanding and often fail to provide accurate description of the associated uncertainty. In this paper a convolutional neural network (CNN) model is described to predict several soil properties with quantified uncertainty. CNN has the advantage that it incorporates spatial contextual information of environmental covariates surrounding an observation. A single CNN model can be trained to predict multiple soil properties simultaneously. I further propose a two-step approach to estimate the uncertainty of the prediction for mapping using a neural network model. The methodology is tested mapping six soil properties on the French metropolitan territory using measurements from the LUCAS dataset and a large set of environmental covariates portraying the factors of soil formation. Results indicate that the multivariate CNN model produces accurate maps as shown by the coefficient of determination and concordance correlation coefficient, compared to a conventional machine learning technique. For this country extent mapping, the maps predicted by CNN have a detailed pattern with significant spatial variation. Evaluation of the uncertainty maps using the median of the standardized squared prediction error and accuracy plots suggests that the uncertainty was accurately quantified, albeit slightly underestimated. The tests conducted using different window size of input covariates to predict the soil properties indicate that CNN benefits from using local contextual information in a radius of 4.5 km. I conclude that CNN is an effective model to predict several soil properties and that the associated uncertainty can be accurately quantified with the proposed approach.

Journal Paper Digests

## Journal Paper Digests 2020 #19* Operational Bayesian GLS Regression for Regional Hydrologic Analyses* The Role of Fire in the Coevolut...… Continue reading

Journal Paper Digests

Published on November 07, 2020

Journal Paper Digests

Published on November 03, 2020