Articles | Volume 6, issue 1
https://doi.org/10.5194/soil-6-35-2020
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
https://doi.org/10.5194/soil-6-35-2020
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Review article
| Highlight paper
| 
06 Feb 2020
Review article | Highlight paper |
| 06 Feb 2020
| 06 Feb 2020
Machine learning and soil sciences: a review aided by machine learning tools
José Padarian
CORRESPONDING AUTHOR
Sydney Institute of Agriculture & School of Life and Environmental Sciences, University of Sydney, Sydney, New South Wales, Australia
Budiman Minasny
Sydney Institute of Agriculture & School of Life and Environmental Sciences, University of Sydney, Sydney, New South Wales, Australia
Alex B. McBratney
Sydney Institute of Agriculture & School of Life and Environmental Sciences, University of Sydney, Sydney, New South Wales, Australia
Related authors
Marliana Tri Widyastuti, Budiman Minasny, José Padarian, Federico Maggi, Matt Aitkenhead, Amélie Beucher, John Connolly, Dian Fiantis, Darren Kidd, Yuxin Ma, Fraser Macfarlane, Ciaran Robb, Rudiyanto, Budi Indra Setiawan, and Muh Taufik
Earth Syst. Sci. Data Discuss., https://doi.org/10.5194/essd-2024-333, https://doi.org/10.5194/essd-2024-333, 2024
Revised manuscript under review for ESSD
Short summary
Short summary
PEATGRIDS, the first dataset containing maps of global peat thickness and carbon stock at 1 km resolution. The dataset has been publicly available at Zenodo to support further analyses and modelling of peatlands across the globe. This work employed the random forest machine learning model to provide spatially explicit peat carbon stock at pixel basis.
Marliana Tri Widyastuti, José Padarian, Budiman Minasny, Mathew Webb, Muh Taufik, and Darren Kidd
EGUsphere, https://doi.org/10.5194/egusphere-2024-2253, https://doi.org/10.5194/egusphere-2024-2253, 2024
Short summary
Short summary
This work aims to predict soil water content across a large region at fine spatial and temporal resolution (80 m grids, daily) to support agricultural management. It covers modelling assessment to predict multilevel soil moisture spatially via deep learning method. We address the challenge of mapping soil moisture at field scale resolution for Tasmania and perform the optimal model for near-real-time monitoring. This contributes to the deep learning method's applicability in soil science.
José Padarian, Budiman Minasny, Alex B. McBratney, and Pete Smith
SOIL Discuss., https://doi.org/10.5194/soil-2021-73, https://doi.org/10.5194/soil-2021-73, 2021
Manuscript not accepted for further review
Short summary
Short summary
Soil organic carbon sequestration is considered an attractive technology to partially mitigate climate change. Here, we show how the SOC storage potential varies globally. The estimated additional SOC storage potential in the topsoil of global croplands (29–67 Pg C) equates to only 2 to 5 years of emissions offsetting and 32 % of agriculture's 92 Pg historical carbon debt. Since SOC is temperature-dependent, this potential is likely to reduce by 18 % by 2040 due to climate change.
José Padarian, Alex B. McBratney, and Budiman Minasny
SOIL, 6, 389–397, https://doi.org/10.5194/soil-6-389-2020, https://doi.org/10.5194/soil-6-389-2020, 2020
Short summary
Short summary
In this paper we introduce the use of game theory to interpret a digital soil mapping (DSM) model to understand the contribution of environmental factors to the prediction of soil organic carbon (SOC) in Chile. The analysis corroborated that the SOC model is capturing sensible relationships between SOC and climatic and topographical factors. We were able to represent them spatially (map) addressing the limitations of the current interpretation of models in DSM.
José Padarian and Alex B. McBratney
SOIL, 6, 89–94, https://doi.org/10.5194/soil-6-89-2020, https://doi.org/10.5194/soil-6-89-2020, 2020
Short summary
Short summary
Data sharing and collaboration are critical to solving large-scale problems. The prevailing soil data-sharing model is of a centralized nature and, consequently, results in the participants ceding control and governance over their data to the lead party. Here we explore the use of a distributed ledger (blockchain) to solve the aforementioned issues. We also describe the potential use case of developing a global soil spectral library between multiple, international institutions.
José Padarian and Ignacio Fuentes
SOIL, 5, 177–187, https://doi.org/10.5194/soil-5-177-2019, https://doi.org/10.5194/soil-5-177-2019, 2019
Short summary
Short summary
A large amount of descriptive information is available in geosciences. Considering the advances in natural language it is possible to
rescuethis information and transform it into a numerical form (embeddings). We used 280764 full-text scientific articles to train a language model capable of generating such embeddings. Our domain-specific embeddings (GeoVec) outperformed general domain embedding tasks such as analogies, relatedness, and categorisation, and can be used in novel applications.
José Padarian, Budiman Minasny, and Alex B. McBratney
SOIL, 5, 79–89, https://doi.org/10.5194/soil-5-79-2019, https://doi.org/10.5194/soil-5-79-2019, 2019
Short summary
Short summary
Digital soil mapping has been widely used as a cost-effective method for generating soil maps. DSM models are usually calibrated using point observations and rarely incorporate contextual information of the landscape. Here, we use convolutional neural networks to incorporate spatial context. We used as input a 3-D stack of covariate images to simultaneously predict organic carbon content at multiple depths. In this study, our model reduced the error by 30 % compared with conventional techniques.
Marliana Tri Widyastuti, Budiman Minasny, José Padarian, Federico Maggi, Matt Aitkenhead, Amélie Beucher, John Connolly, Dian Fiantis, Darren Kidd, Yuxin Ma, Fraser Macfarlane, Ciaran Robb, Rudiyanto, Budi Indra Setiawan, and Muh Taufik
Earth Syst. Sci. Data Discuss., https://doi.org/10.5194/essd-2024-333, https://doi.org/10.5194/essd-2024-333, 2024
Revised manuscript under review for ESSD
Short summary
Short summary
PEATGRIDS, the first dataset containing maps of global peat thickness and carbon stock at 1 km resolution. The dataset has been publicly available at Zenodo to support further analyses and modelling of peatlands across the globe. This work employed the random forest machine learning model to provide spatially explicit peat carbon stock at pixel basis.
Marliana Tri Widyastuti, José Padarian, Budiman Minasny, Mathew Webb, Muh Taufik, and Darren Kidd
EGUsphere, https://doi.org/10.5194/egusphere-2024-2253, https://doi.org/10.5194/egusphere-2024-2253, 2024
Short summary
Short summary
This work aims to predict soil water content across a large region at fine spatial and temporal resolution (80 m grids, daily) to support agricultural management. It covers modelling assessment to predict multilevel soil moisture spatially via deep learning method. We address the challenge of mapping soil moisture at field scale resolution for Tasmania and perform the optimal model for near-real-time monitoring. This contributes to the deep learning method's applicability in soil science.
Tobias Karl David Weber, Lutz Weihermüller, Attila Nemes, Michel Bechtold, Aurore Degré, Efstathios Diamantopoulos, Simone Fatichi, Vilim Filipović, Surya Gupta, Tobias L. Hohenbrink, Daniel R. Hirmas, Conrad Jackisch, Quirijn de Jong van Lier, John Koestel, Peter Lehmann, Toby R. Marthews, Budiman Minasny, Holger Pagel, Martine van der Ploeg, Shahab Aldin Shojaeezadeh, Simon Fiil Svane, Brigitta Szabó, Harry Vereecken, Anne Verhoef, Michael Young, Yijian Zeng, Yonggen Zhang, and Sara Bonetti
Hydrol. Earth Syst. Sci., 28, 3391–3433, https://doi.org/10.5194/hess-28-3391-2024, https://doi.org/10.5194/hess-28-3391-2024, 2024
Short summary
Short summary
Pedotransfer functions (PTFs) are used to predict parameters of models describing the hydraulic properties of soils. The appropriateness of these predictions critically relies on the nature of the datasets for training the PTFs and the physical comprehensiveness of the models. This roadmap paper is addressed to PTF developers and users and critically reflects the utility and future of PTFs. To this end, we present a manifesto aiming at a paradigm shift in PTF research.
Frisa Irawan Ginting, Rudiyanto Rudiyanto, Fatchurahman, Ramisah Mohd Shah, Norhidayah Che Soh, Sunny Goh Eng Giap, Dian Fiantis, Budi Indra Setiawan, Sam Schiller, Aaron Davitt, and Budiman Minasny
Earth Syst. Sci. Data Discuss., https://doi.org/10.5194/essd-2024-90, https://doi.org/10.5194/essd-2024-90, 2024
Preprint withdrawn
Short summary
Short summary
This study is the first to map rice cropping intensity and the harvested area across Southeast Asia at a spatial resolution of 10 m (SEA-Rice-Ci10). We have developed a geospatial inventory of paddy rice parcels and rice cropping intensity by integrating Sentinel-1 and 2 time-series data in a framework called LUCK-PALM, based on local phenological expert interpretation. According to our best knowledge, it is the finest-resolution and most accurate database of paddy rice in Southeast Asia.
Wartini Ng, Budiman Minasny, Alex McBratney, Patrice de Caritat, and John Wilford
Earth Syst. Sci. Data, 15, 2465–2482, https://doi.org/10.5194/essd-15-2465-2023, https://doi.org/10.5194/essd-15-2465-2023, 2023
Short summary
Short summary
With a higher demand for lithium (Li), a better understanding of its concentration and spatial distribution is important to delineate potential anomalous areas. This study uses a framework that combines data from recent geochemical surveys and relevant environmental factors to predict and map Li content across Australia. The map shows high Li concentration around existing mines and other potentially anomalous Li areas. The same mapping principles can potentially be applied to other elements.
Mercedes Román Dobarco, Alexandre M. J-C. Wadoux, Brendan Malone, Budiman Minasny, Alex B. McBratney, and Ross Searle
Biogeosciences, 20, 1559–1586, https://doi.org/10.5194/bg-20-1559-2023, https://doi.org/10.5194/bg-20-1559-2023, 2023
Short summary
Short summary
Soil organic carbon (SOC) is of a heterogeneous nature and varies in chemistry, stabilisation mechanisms, and persistence in soil. In this study we mapped the stocks of SOC fractions with different characteristics and turnover rates (presumably PyOC >= MAOC > POC) across Australia, combining spectroscopy and digital soil mapping. The SOC stocks (0–30 cm) were estimated as 13 Pg MAOC, 2 Pg POC, and 5 Pg PyOC.
José Padarian, Budiman Minasny, Alex B. McBratney, and Pete Smith
SOIL Discuss., https://doi.org/10.5194/soil-2021-73, https://doi.org/10.5194/soil-2021-73, 2021
Manuscript not accepted for further review
Short summary
Short summary
Soil organic carbon sequestration is considered an attractive technology to partially mitigate climate change. Here, we show how the SOC storage potential varies globally. The estimated additional SOC storage potential in the topsoil of global croplands (29–67 Pg C) equates to only 2 to 5 years of emissions offsetting and 32 % of agriculture's 92 Pg historical carbon debt. Since SOC is temperature-dependent, this potential is likely to reduce by 18 % by 2040 due to climate change.
Edward J. Jones, Patrick Filippi, Rémi Wittig, Mario Fajardo, Vanessa Pino, and Alex B. McBratney
SOIL, 7, 33–46, https://doi.org/10.5194/soil-7-33-2021, https://doi.org/10.5194/soil-7-33-2021, 2021
Short summary
Short summary
Soil physical health is integral to maintaining functional agro-ecosystems. A novel method of assessing soil physical condition using a smartphone app has been developed – SLAKES. In this study the SLAKES app was used to investigate aggregate stability in a mixed agricultural landscape. Cropping areas were found to have significantly poorer physical health than similar soils under pasture. Results were mapped across the landscape to identify problem areas and pinpoint remediation efforts.
Wartini Ng, Budiman Minasny, Wanderson de Sousa Mendes, and José Alexandre Melo Demattê
SOIL, 6, 565–578, https://doi.org/10.5194/soil-6-565-2020, https://doi.org/10.5194/soil-6-565-2020, 2020
Short summary
Short summary
The number of samples utilised to create predictive models affected model performance. This research compares the number of samples needed by a deep learning model to outperform the traditional machine learning models using visible near-infrared spectroscopy data for soil properties predictions. The deep learning model was found to outperform machine learning models when the sample size was above 2000.
José Padarian, Alex B. McBratney, and Budiman Minasny
SOIL, 6, 389–397, https://doi.org/10.5194/soil-6-389-2020, https://doi.org/10.5194/soil-6-389-2020, 2020
Short summary
Short summary
In this paper we introduce the use of game theory to interpret a digital soil mapping (DSM) model to understand the contribution of environmental factors to the prediction of soil organic carbon (SOC) in Chile. The analysis corroborated that the SOC model is capturing sensible relationships between SOC and climatic and topographical factors. We were able to represent them spatially (map) addressing the limitations of the current interpretation of models in DSM.
Yosra Ellili-Bargaoui, Brendan Philip Malone, Didier Michot, Budiman Minasny, Sébastien Vincent, Christian Walter, and Blandine Lemercier
SOIL, 6, 371–388, https://doi.org/10.5194/soil-6-371-2020, https://doi.org/10.5194/soil-6-371-2020, 2020
Sanjeewani Nimalka Somarathna Pallegedara Dewage, Budiman Minasny, and Brendan Malone
SOIL, 6, 359–369, https://doi.org/10.5194/soil-6-359-2020, https://doi.org/10.5194/soil-6-359-2020, 2020
Short summary
Short summary
Most soil management activities are implemented at farm scale, yet digital soil maps are commonly available at regional/national scales. This study proposes Bayesian area-to-point kriging to downscale regional-/national-scale soil property maps to farm scale. A regional soil carbon map with a resolution of 100 m (block support) was disaggregated to 10 m (point support) information for a farm in northern NSW, Australia. Results are presented with the uncertainty of the downscaling process.
José Padarian and Alex B. McBratney
SOIL, 6, 89–94, https://doi.org/10.5194/soil-6-89-2020, https://doi.org/10.5194/soil-6-89-2020, 2020
Short summary
Short summary
Data sharing and collaboration are critical to solving large-scale problems. The prevailing soil data-sharing model is of a centralized nature and, consequently, results in the participants ceding control and governance over their data to the lead party. Here we explore the use of a distributed ledger (blockchain) to solve the aforementioned issues. We also describe the potential use case of developing a global soil spectral library between multiple, international institutions.
José Padarian and Ignacio Fuentes
SOIL, 5, 177–187, https://doi.org/10.5194/soil-5-177-2019, https://doi.org/10.5194/soil-5-177-2019, 2019
Short summary
Short summary
A large amount of descriptive information is available in geosciences. Considering the advances in natural language it is possible to
rescuethis information and transform it into a numerical form (embeddings). We used 280764 full-text scientific articles to train a language model capable of generating such embeddings. Our domain-specific embeddings (GeoVec) outperformed general domain embedding tasks such as analogies, relatedness, and categorisation, and can be used in novel applications.
Alexandre M. J.-C. Wadoux, José Padarian, and Budiman Minasny
SOIL, 5, 107–119, https://doi.org/10.5194/soil-5-107-2019, https://doi.org/10.5194/soil-5-107-2019, 2019
José Padarian, Budiman Minasny, and Alex B. McBratney
SOIL, 5, 79–89, https://doi.org/10.5194/soil-5-79-2019, https://doi.org/10.5194/soil-5-79-2019, 2019
Short summary
Short summary
Digital soil mapping has been widely used as a cost-effective method for generating soil maps. DSM models are usually calibrated using point observations and rarely incorporate contextual information of the landscape. Here, we use convolutional neural networks to incorporate spatial context. We used as input a 3-D stack of covariate images to simultaneously predict organic carbon content at multiple depths. In this study, our model reduced the error by 30 % compared with conventional techniques.
Edward J. Jones and Alex B. McBratney
SOIL Discuss., https://doi.org/10.5194/soil-2018-12, https://doi.org/10.5194/soil-2018-12, 2018
Revised manuscript has not been submitted
Short summary
Short summary
Variable soil moisture content is one of the main factors limiting field application of visible near-infrared spectroscopy. External parameter orthogonalisation of soil spectra was found to conserve intrinsic soil information under variable moisture conditions. k-means clustering of treated spectra yielded similar classifications under in situ, field moist (laboratory) and air-dried condition. Homogeneous spectral response zones were identified that corresponded with field observed horizons.
Related subject area
Soil and methods
Spatial prediction of organic carbon in German agricultural topsoil using machine learning algorithms
On the benefits of clustering approaches in digital soil mapping: an application example concerning soil texture regionalization
An open Soil Structure Library based on X-ray CT data
Identification of thermal signature and quantification of charcoal in soil using differential scanning calorimetry and benzene polycarboxylic acid (BPCA) markers
Estimating soil fungal abundance and diversity at a macroecological scale with deep learning spectrotransfer functions
An underground, wireless, open-source, low-cost system for monitoring oxygen, temperature, and soil moisture
Estimation of soil properties with mid-infrared soil spectroscopy across yam production landscapes in West Africa
The central African soil spectral library: a new soil infrared repository and a geographical prediction analysis
Developing the Swiss mid-infrared soil spectral library for local estimation and monitoring
Predicting the spatial distribution of soil organic carbon stock in Swedish forests using a group of covariates and site-specific data
Improved calibration of the Green–Ampt infiltration module in the EROSION-2D/3D model using a rainfall-runoff experiment database
Quantifying soil carbon in temperate peatlands using a mid-IR soil spectral library
Are researchers following best storage practices for measuring soil biochemical properties?
Quantifying and correcting for pre-assay CO2 loss in short-term carbon mineralization assays
The influence of training sample size on the accuracy of deep learning models for the prediction of soil properties with near-infrared spectroscopy data
Game theory interpretation of digital soil mapping convolutional neural networks
Comparing three approaches of spatial disaggregation of legacy soil maps based on the Disaggregation and Harmonisation of Soil Map Units Through Resampled Classification Trees (DSMART) algorithm
Oblique geographic coordinates as covariates for digital soil mapping
Development of pedotransfer functions for water retention in tropical mountain soil landscapes: spotlight on parameter tuning in machine learning
The 15N gas-flux method to determine N2 flux: a comparison of different tracer addition approaches
A new model for intra- and inter-institutional soil data sharing
Identification of new microbial functional standards for soil quality assessment
Identifying and quantifying geogenic organic carbon in soils – the case of graphite
Error propagation in spectrometric functions of soil organic carbon
Word embeddings for application in geosciences: development, evaluation, and examples of soil-related concepts
Soil lacquer peel do-it-yourself: simply capturing beauty
Multi-source data integration for soil mapping using deep learning
Using deep learning for digital soil mapping
No silver bullet for digital soil mapping: country-specific soil organic carbon estimates across Latin America
Separation of soil respiration: a site-specific comparison of partition methods
Proximal sensing for soil carbon accounting
Evaluation of digital soil mapping approaches with large sets of environmental covariates
Planning spatial sampling of the soil from an uncertain reconnaissance variogram
Mapping of soil properties at high resolution in Switzerland using boosted geoadditive models
Quantitative imaging of the 3-D distribution of cation adsorption sites in undisturbed soil
Decision support for the selection of reference sites using 137Cs as a soil erosion tracer
Soil organic carbon stocks are systematically overestimated by misuse of the parameters bulk density and rock fragment content
The added value of biomarker analysis to the genesis of plaggic Anthrosols; the identification of stable fillings used for the production of plaggic manure
Synchrotron microtomographic quantification of geometrical soil pore characteristics affected by compaction
Pedotransfer functions for Irish soils – estimation of bulk density (ρb) per horizon type
Assessing the performance of a plastic optical fibre turbidity sensor for measuring post-fire erosion from plot to catchment scale
Passive soil heating using an inexpensive infrared mirror design – a proof of concept
The application of terrestrial laser scanner and SfM photogrammetry in measuring erosion and deposition processes in two opposite slopes in a humid badlands area (central Spanish Pyrenees)
Soil surface roughness: comparing old and new measuring methods and application in a soil erosion model
Comparison of spatial association approaches for landscape mapping of soil organic carbon stocks
Eddy covariance for quantifying trace gas fluxes from soils
Ali Sakhaee, Anika Gebauer, Mareike Ließ, and Axel Don
SOIL, 8, 587–604, https://doi.org/10.5194/soil-8-587-2022, https://doi.org/10.5194/soil-8-587-2022, 2022
Short summary
Short summary
As soil carbon has become a key component of climate-smart agriculture, the demand for high-resolution maps has increased drastically. Meanwhile, machine learning algorithms are becoming more widely used and are opening up new solutions in soil mapping. This paper shows which algorithms perform best, how soil inventory data can be most efficiently used for digital soil mapping, and the different available options and methods to derive high-resolution soil carbon data at the large regional scale.
István Dunkl and Mareike Ließ
SOIL, 8, 541–558, https://doi.org/10.5194/soil-8-541-2022, https://doi.org/10.5194/soil-8-541-2022, 2022
Short summary
Short summary
Digital soil mapping (DSM) allows us to regionalize soil properties by relating them to environmental covariates with the help of an empirical model. Legacy soil data provide a valuable basis to generate high-resolution soil maps with DSM. We studied the usefulness of data-clustering methods to tackle potential sampling bias in legacy soil data while applying DSM for soil texture regionalization. Clustering has proved to be useful in various steps of the DSM process.
Ulrich Weller, Lukas Albrecht, Steffen Schlüter, and Hans-Jörg Vogel
SOIL, 8, 507–515, https://doi.org/10.5194/soil-8-507-2022, https://doi.org/10.5194/soil-8-507-2022, 2022
Short summary
Short summary
Soil structure is of central importance for soil functions. It is, however, ill defined. With the increasing availability of X-ray CT scanners, more and more soils are scanned and an undisturbed image of the soil's structure is produced. Often, a qualitative description is all that is derived from these images. We provide now a web-based Soil Structure Library where these images can be evaluated in a standardized quantitative way and can be compared to a world-wide data set.
Brieuc Hardy, Nils Borchard, and Jens Leifeld
SOIL, 8, 451–466, https://doi.org/10.5194/soil-8-451-2022, https://doi.org/10.5194/soil-8-451-2022, 2022
Short summary
Short summary
Soil amendment with artificial black carbon (BC; biomass transformed by incomplete combustion) has the potential to mitigate climate change. Nevertheless, the accurate quantification of BC in soil remains a critical issue. Here, we successfully used dynamic thermal analysis (DTA) to quantify centennial BC in soil. We demonstrate that DTA is largely under-exploited despite providing rapid and low-cost quantitative information over the range of soil organic matter.
Yuanyuan Yang, Zefang Shen, Andrew Bissett, and Raphael A. Viscarra Rossel
SOIL, 8, 223–235, https://doi.org/10.5194/soil-8-223-2022, https://doi.org/10.5194/soil-8-223-2022, 2022
Short summary
Short summary
We present a new method to estimate the relative abundance of the dominant phyla and diversity of fungi in Australian soil. It uses state-of-the-art machine learning with publicly available data on soil and environmental proxies for edaphic, climatic, biotic and topographic factors, and visible–near infrared wavelengths. The estimates could serve to supplement the more expensive molecular approaches towards a better understanding of soil fungal abundance and diversity in agronomy and ecology.
Elad Levintal, Yonatan Ganot, Gail Taylor, Peter Freer-Smith, Kosana Suvocarev, and Helen E. Dahlke
SOIL, 8, 85–97, https://doi.org/10.5194/soil-8-85-2022, https://doi.org/10.5194/soil-8-85-2022, 2022
Short summary
Short summary
Do-it-yourself hardware is a new approach for improving measurement resolution in research. Here we present a new low-cost, wireless underground sensor network for soil monitoring. All data logging, power, and communication component cost is USD 150, much cheaper than other available commercial solutions. We provide the complete building guide to reduce any technical barriers, which we hope will allow easier reproducibility and open new environmental monitoring applications.
Philipp Baumann, Juhwan Lee, Emmanuel Frossard, Laurie Paule Schönholzer, Lucien Diby, Valérie Kouamé Hgaza, Delwende Innocent Kiba, Andrew Sila, Keith Sheperd, and Johan Six
SOIL, 7, 717–731, https://doi.org/10.5194/soil-7-717-2021, https://doi.org/10.5194/soil-7-717-2021, 2021
Short summary
Short summary
This work delivers openly accessible and validated calibrations for diagnosing 26 soil properties based on mid-infrared spectroscopy. These were developed for four regions in Burkina Faso and Côte d'Ivoire, including 80 fields of smallholder farmers. The models can help to site-specifically and cost-efficiently monitor soil quality and fertility constraints to ameliorate soils and yields of yam or other staple crops in the four regions between the humid forest and the northern Guinean savanna.
Laura Summerauer, Philipp Baumann, Leonardo Ramirez-Lopez, Matti Barthel, Marijn Bauters, Benjamin Bukombe, Mario Reichenbach, Pascal Boeckx, Elizabeth Kearsley, Kristof Van Oost, Bernard Vanlauwe, Dieudonné Chiragaga, Aimé Bisimwa Heri-Kazi, Pieter Moonen, Andrew Sila, Keith Shepherd, Basile Bazirake Mujinya, Eric Van Ranst, Geert Baert, Sebastian Doetterl, and Johan Six
SOIL, 7, 693–715, https://doi.org/10.5194/soil-7-693-2021, https://doi.org/10.5194/soil-7-693-2021, 2021
Short summary
Short summary
We present a soil mid-infrared library with over 1800 samples from central Africa in order to facilitate soil analyses of this highly understudied yet critical area. Together with an existing continental library, we demonstrate a regional analysis and geographical extrapolation to predict total carbon and nitrogen. Our results show accurate predictions and highlight the value that the data contribute to existing libraries. Our library is openly available for public use and for expansion.
Philipp Baumann, Anatol Helfenstein, Andreas Gubler, Armin Keller, Reto Giulio Meuli, Daniel Wächter, Juhwan Lee, Raphael Viscarra Rossel, and Johan Six
SOIL, 7, 525–546, https://doi.org/10.5194/soil-7-525-2021, https://doi.org/10.5194/soil-7-525-2021, 2021
Short summary
Short summary
We developed the Swiss mid-infrared spectral library and a statistical model collection across 4374 soil samples with reference measurements of 16 properties. Our library incorporates soil from 1094 grid locations and 71 long-term monitoring sites. This work confirms once again that nationwide spectral libraries with diverse soils can reliably feed information to a fast chemical diagnosis. Our data-driven reduction of the library has the potential to accurately monitor carbon at the plot scale.
Kpade O. L. Hounkpatin, Johan Stendahl, Mattias Lundblad, and Erik Karltun
SOIL, 7, 377–398, https://doi.org/10.5194/soil-7-377-2021, https://doi.org/10.5194/soil-7-377-2021, 2021
Short summary
Short summary
Forests store large amounts of carbon in soils. Implementing suitable measures to improve the sink potential of forest soils would require accurate data on the carbon stored in forest soils and a better understanding of the factors affecting this storage. This study showed that the prediction of soil carbon stock in Swedish forest soils can increase in accuracy when one divides a big region into smaller areas in combination with information collected locally and derived from satellites.
Hana Beitlerová, Jonas Lenz, Jan Devátý, Martin Mistr, Jiří Kapička, Arno Buchholz, Ilona Gerndtová, and Anne Routschek
SOIL, 7, 241–253, https://doi.org/10.5194/soil-7-241-2021, https://doi.org/10.5194/soil-7-241-2021, 2021
Short summary
Short summary
This study presents transfer functions for a calibration parameter of the Green–Ampt infiltration module of the EROSION-2D/3D model, which are significantly improving the model performance compared to the current state. The relationships found between calibration parameters and soil parameters however put the Green–Ampt implementation in the model and the state-of-the-art parametrization method in question. A new direction of the infiltration module development is proposed.
Anatol Helfenstein, Philipp Baumann, Raphael Viscarra Rossel, Andreas Gubler, Stefan Oechslin, and Johan Six
SOIL, 7, 193–215, https://doi.org/10.5194/soil-7-193-2021, https://doi.org/10.5194/soil-7-193-2021, 2021
Short summary
Short summary
In this study, we show that a soil spectral library (SSL) can be used to predict soil carbon at new and very different locations. The importance of this finding is that it requires less time-consuming lab work than calibrating a new model for every local application, while still remaining similar to or more accurate than local models. Furthermore, we show that this method even works for predicting (drained) peat soils, using a SSL with mostly mineral soils containing much less soil carbon.
Jennifer M. Rhymes, Irene Cordero, Mathilde Chomel, Jocelyn M. Lavallee, Angela L. Straathof, Deborah Ashworth, Holly Langridge, Marina Semchenko, Franciska T. de Vries, David Johnson, and Richard D. Bardgett
SOIL, 7, 95–106, https://doi.org/10.5194/soil-7-95-2021, https://doi.org/10.5194/soil-7-95-2021, 2021
Matthew A. Belanger, Carmella Vizza, G. Philip Robertson, and Sarah S. Roley
SOIL, 7, 47–52, https://doi.org/10.5194/soil-7-47-2021, https://doi.org/10.5194/soil-7-47-2021, 2021
Short summary
Short summary
Soil health is often assessed by re-wetting a dry soil and measuring CO2 production, but the potential bias introduced by soils of different moisture contents is unclear. Our study found that wetter soil tended to lose more carbon during drying than drier soil, thus affecting soil health interpretations. We developed a correction factor to account for initial soil moisture effects, which future studies may benefit from adapting for their soil.
Wartini Ng, Budiman Minasny, Wanderson de Sousa Mendes, and José Alexandre Melo Demattê
SOIL, 6, 565–578, https://doi.org/10.5194/soil-6-565-2020, https://doi.org/10.5194/soil-6-565-2020, 2020
Short summary
Short summary
The number of samples utilised to create predictive models affected model performance. This research compares the number of samples needed by a deep learning model to outperform the traditional machine learning models using visible near-infrared spectroscopy data for soil properties predictions. The deep learning model was found to outperform machine learning models when the sample size was above 2000.
José Padarian, Alex B. McBratney, and Budiman Minasny
SOIL, 6, 389–397, https://doi.org/10.5194/soil-6-389-2020, https://doi.org/10.5194/soil-6-389-2020, 2020
Short summary
Short summary
In this paper we introduce the use of game theory to interpret a digital soil mapping (DSM) model to understand the contribution of environmental factors to the prediction of soil organic carbon (SOC) in Chile. The analysis corroborated that the SOC model is capturing sensible relationships between SOC and climatic and topographical factors. We were able to represent them spatially (map) addressing the limitations of the current interpretation of models in DSM.
Yosra Ellili-Bargaoui, Brendan Philip Malone, Didier Michot, Budiman Minasny, Sébastien Vincent, Christian Walter, and Blandine Lemercier
SOIL, 6, 371–388, https://doi.org/10.5194/soil-6-371-2020, https://doi.org/10.5194/soil-6-371-2020, 2020
Anders Bjørn Møller, Amélie Marie Beucher, Nastaran Pouladi, and Mogens Humlekrog Greve
SOIL, 6, 269–289, https://doi.org/10.5194/soil-6-269-2020, https://doi.org/10.5194/soil-6-269-2020, 2020
Short summary
Short summary
Decision trees have become a widely adapted tool for mapping soil properties in geographic space. However, it is problematic to implement spatial relationships in the models. We present a new method which uses geographic coordinates along several axes tilted at oblique angles in the models. We test this method on four spatial datasets. The results show that the new method is at least as accurate as other proposed alternatives, has a computational advantage and is flexible and interpretable.
Anika Gebauer, Monja Ellinger, Victor M. Brito Gomez, and Mareike Ließ
SOIL, 6, 215–229, https://doi.org/10.5194/soil-6-215-2020, https://doi.org/10.5194/soil-6-215-2020, 2020
Short summary
Short summary
Pedotransfer functions (PTFs) for soil water retention were developed for two tropical soil landscapes using machine learning. The models corresponding to these PTFs had to be adjusted by tuning their parameters. The standard tuning approach was compared to mathematical optimization. The latter resulted in much better model performance. The PTFs derived are of particular importance for soil process and hydrological models.
Dominika Lewicka-Szczebak and Reinhard Well
SOIL, 6, 145–152, https://doi.org/10.5194/soil-6-145-2020, https://doi.org/10.5194/soil-6-145-2020, 2020
Short summary
Short summary
This study aimed at comparison of various experimental strategies for incubating soil samples to determine the N2 flux. Such experiments require addition of isotope tracer, i.e. nitrogen fertilizer enriched in heavy nitrogen isotopes (15N). Here we compared the impact of soil homogenization and mixing with the tracer and tracer injection to the intact soil cores. The results are well comparable: both techniques would provide similar conclusions on the magnitude of N2 flux.
José Padarian and Alex B. McBratney
SOIL, 6, 89–94, https://doi.org/10.5194/soil-6-89-2020, https://doi.org/10.5194/soil-6-89-2020, 2020
Short summary
Short summary
Data sharing and collaboration are critical to solving large-scale problems. The prevailing soil data-sharing model is of a centralized nature and, consequently, results in the participants ceding control and governance over their data to the lead party. Here we explore the use of a distributed ledger (blockchain) to solve the aforementioned issues. We also describe the potential use case of developing a global soil spectral library between multiple, international institutions.
Sören Thiele-Bruhn, Michael Schloter, Berndt-Michael Wilke, Lee A. Beaudette, Fabrice Martin-Laurent, Nathalie Cheviron, Christian Mougin, and Jörg Römbke
SOIL, 6, 17–34, https://doi.org/10.5194/soil-6-17-2020, https://doi.org/10.5194/soil-6-17-2020, 2020
Short summary
Short summary
Soil quality depends on the functioning of soil microbiota. Only a few standardized methods are available to assess this as well as adverse effects of human activities. So we need to identify promising additional methods that target soil microbial function. Discussed are (i) molecular methods using qPCR for new endpoints, e.g. in N and P cycling and greenhouse gas emissions, (ii) techniques for fungal enzyme activities, and (iii) field methods on carbon turnover such as the litter bag test.
Jeroen H. T. Zethof, Martin Leue, Cordula Vogel, Shane W. Stoner, and Karsten Kalbitz
SOIL, 5, 383–398, https://doi.org/10.5194/soil-5-383-2019, https://doi.org/10.5194/soil-5-383-2019, 2019
Short summary
Short summary
A widely overlooked source of carbon (C) in the soil environment is organic C of geogenic origin, e.g. graphite. Appropriate methods are not available to quantify graphite and to differentiate it from other organic and inorganic C sources in soils. Therefore, we examined Fourier transform infrared spectroscopy, thermogravimetric analysis and the smart combustion method for their ability to identify and quantify graphitic C in soils. The smart combustion method showed the most promising results.
Monja Ellinger, Ines Merbach, Ulrike Werban, and Mareike Ließ
SOIL, 5, 275–288, https://doi.org/10.5194/soil-5-275-2019, https://doi.org/10.5194/soil-5-275-2019, 2019
Short summary
Short summary
Vis–NIR spectrometry is often applied to capture soil organic carbon (SOC). This study addresses the impact of the involved data and modelling aspects on SOC precision with a focus on the propagation of input data uncertainties. It emphasizes the necessity of transparent documentation of the measurement protocol and the model building and validation procedure. Particularly, when Vis–NIR spectrometry is used for soil monitoring, the aspect of uncertainty propagation becomes essential.
José Padarian and Ignacio Fuentes
SOIL, 5, 177–187, https://doi.org/10.5194/soil-5-177-2019, https://doi.org/10.5194/soil-5-177-2019, 2019
Short summary
Short summary
A large amount of descriptive information is available in geosciences. Considering the advances in natural language it is possible to
rescuethis information and transform it into a numerical form (embeddings). We used 280764 full-text scientific articles to train a language model capable of generating such embeddings. Our domain-specific embeddings (GeoVec) outperformed general domain embedding tasks such as analogies, relatedness, and categorisation, and can be used in novel applications.
Cathelijne R. Stoof, Jasper H. J. Candel, Laszlo A. G. M. van der Wal, and Gert Peek
SOIL, 5, 159–175, https://doi.org/10.5194/soil-5-159-2019, https://doi.org/10.5194/soil-5-159-2019, 2019
Short summary
Short summary
Teaching and outreach of soils is often done with real-life snapshots of soils and sediments in lacquer or glue peels. While it may seem hard, anyone can make such a peel. Illustrated with handmade drawings and an instructional video, we explain how to capture soils in peels using readily available materials. A new twist to old methods makes this safer, simpler, and more successful, and thus a true DIY (do-it-yourself) activity, highlighting the value and beauty of the ground below our feet.
Alexandre M. J.-C. Wadoux, José Padarian, and Budiman Minasny
SOIL, 5, 107–119, https://doi.org/10.5194/soil-5-107-2019, https://doi.org/10.5194/soil-5-107-2019, 2019
José Padarian, Budiman Minasny, and Alex B. McBratney
SOIL, 5, 79–89, https://doi.org/10.5194/soil-5-79-2019, https://doi.org/10.5194/soil-5-79-2019, 2019
Short summary
Short summary
Digital soil mapping has been widely used as a cost-effective method for generating soil maps. DSM models are usually calibrated using point observations and rarely incorporate contextual information of the landscape. Here, we use convolutional neural networks to incorporate spatial context. We used as input a 3-D stack of covariate images to simultaneously predict organic carbon content at multiple depths. In this study, our model reduced the error by 30 % compared with conventional techniques.
Mario Guevara, Guillermo Federico Olmedo, Emma Stell, Yusuf Yigini, Yameli Aguilar Duarte, Carlos Arellano Hernández, Gloria E. Arévalo, Carlos Eduardo Arroyo-Cruz, Adriana Bolivar, Sally Bunning, Nelson Bustamante Cañas, Carlos Omar Cruz-Gaistardo, Fabian Davila, Martin Dell Acqua, Arnulfo Encina, Hernán Figueredo Tacona, Fernando Fontes, José Antonio Hernández Herrera, Alejandro Roberto Ibelles Navarro, Veronica Loayza, Alexandra M. Manueles, Fernando Mendoza Jara, Carolina Olivera, Rodrigo Osorio Hermosilla, Gonzalo Pereira, Pablo Prieto, Iván Alexis Ramos, Juan Carlos Rey Brina, Rafael Rivera, Javier Rodríguez-Rodríguez, Ronald Roopnarine, Albán Rosales Ibarra, Kenset Amaury Rosales Riveiro, Guillermo Andrés Schulz, Adrian Spence, Gustavo M. Vasques, Ronald R. Vargas, and Rodrigo Vargas
SOIL, 4, 173–193, https://doi.org/10.5194/soil-4-173-2018, https://doi.org/10.5194/soil-4-173-2018, 2018
Short summary
Short summary
We provide a reproducible multi-modeling approach for SOC mapping across Latin America on a country-specific basis as required by the Global Soil Partnership of the United Nations. We identify key prediction factors for SOC across each country. We compare and test different methods to generate spatially explicit predictions of SOC and conclude that there is no best method on a quantifiable basis.
Louis-Pierre Comeau, Derrick Y. F. Lai, Jane Jinglan Cui, and Jenny Farmer
SOIL, 4, 141–152, https://doi.org/10.5194/soil-4-141-2018, https://doi.org/10.5194/soil-4-141-2018, 2018
Short summary
Short summary
To date, there are still many uncertainties and unknowns regarding the soil respiration partitioning procedures. This study compared the suitability and accuracy of five different respiration partitioning methods. A qualitative evaluation table of the partition methods with five performance parameters was produced. Overall, no systematically superior or inferior partition method was found and the combination of two or more methods optimizes assessment reliability.
Jacqueline R. England and Raphael A. Viscarra Rossel
SOIL, 4, 101–122, https://doi.org/10.5194/soil-4-101-2018, https://doi.org/10.5194/soil-4-101-2018, 2018
Short summary
Short summary
Proximal sensing can be used for soil C accounting, but the methods need to be standardized and procedural guidelines developed to ensure proficient measurement and accurate reporting. This is particularly important if there are financial incentives for landholders to adopt practices to sequester C. We review sensing for C accounting and discuss the requirements for the development of new soil C accounting methods based on sensing, including requirements for reporting, auditing and verification.
Madlene Nussbaum, Kay Spiess, Andri Baltensweiler, Urs Grob, Armin Keller, Lucie Greiner, Michael E. Schaepman, and Andreas Papritz
SOIL, 4, 1–22, https://doi.org/10.5194/soil-4-1-2018, https://doi.org/10.5194/soil-4-1-2018, 2018
Short summary
Short summary
This paper presents an extensive evaluation of digital soil mapping (DSM) tools. Recently, large sets of environmental covariates (e.g. from analysis of terrain on multiple scales) have become more common for DSM. Many DSM studies, however, only compared DSM methods using less than 30 covariates or tested approaches on few responses. We built DSM models from 300–500 covariates using six approaches that are either popular in DSM or promising for large covariate sets.
R. Murray Lark, Elliott M. Hamilton, Belinda Kaninga, Kakoma K. Maseka, Moola Mutondo, Godfrey M. Sakala, and Michael J. Watts
SOIL, 3, 235–244, https://doi.org/10.5194/soil-3-235-2017, https://doi.org/10.5194/soil-3-235-2017, 2017
Short summary
Short summary
An advantage of geostatistics for mapping soil properties is that, given a statistical model of the variable of interest, we can make a rational decision about how densely to sample so that the map is sufficiently precise. However, uncertainty about the statistical model affects this process. In this paper we show how Bayesian methods can be used to support decision making on sampling with an uncertain model, ensuring that the probability of meeting certain levels of precision is high enough.
Madlene Nussbaum, Lorenz Walthert, Marielle Fraefel, Lucie Greiner, and Andreas Papritz
SOIL, 3, 191–210, https://doi.org/10.5194/soil-3-191-2017, https://doi.org/10.5194/soil-3-191-2017, 2017
Short summary
Short summary
Digital soil mapping (DSM) relates soil property data to environmental data that describe soil-forming factors. With imagery sampled from satellites or terrain analysed at multiple scales, large sets of possible input to DSM are available. We propose a new statistical framework (geoGAM) that selects parsimonious models for DSM and illustrate the application of geoGAM to two study regions. Straightforward interpretation of the modelled effects likely improves end-user acceptance of DSM products.
Hannes Keck, Bjarne W. Strobel, Jon Petter Gustafsson, and John Koestel
SOIL, 3, 177–189, https://doi.org/10.5194/soil-3-177-2017, https://doi.org/10.5194/soil-3-177-2017, 2017
Short summary
Short summary
Several studies have shown that the cation adsorption sites in soils are heterogeneously distributed in space. In many soil system models this knowledge is not included yet. In our study we proposed a new method to map the 3-D distribution of cation adsorption sites in undisturbed soils. The method is based on three-dimensional X-ray scanning with a contrast agent and image analysis. We are convinced that this approach will strongly aid the development of more realistic soil system models.
Laura Arata, Katrin Meusburger, Alexandra Bürge, Markus Zehringer, Michael E. Ketterer, Lionel Mabit, and Christine Alewell
SOIL, 3, 113–122, https://doi.org/10.5194/soil-3-113-2017, https://doi.org/10.5194/soil-3-113-2017, 2017
Christopher Poeplau, Cora Vos, and Axel Don
SOIL, 3, 61–66, https://doi.org/10.5194/soil-3-61-2017, https://doi.org/10.5194/soil-3-61-2017, 2017
Short summary
Short summary
This paper shows that three out of four frequently used methods to calculate soil organic carbon stocks lead to systematic overestimation of those stocks. Stones, which can be assumed to be free of carbon, have to be corrected for in both bulk density and layer thickness. We used data of the German Agricultural Soil Inventory to illustrate the potential bias and suggest a unified and unbiased calculation method for stocks of soil organic carbon, which is the largest terrestrial carbon pool.
Jan M. van Mourik, Thomas V. Wagner, J. Geert de Boer, and Boris Jansen
SOIL, 2, 299–310, https://doi.org/10.5194/soil-2-299-2016, https://doi.org/10.5194/soil-2-299-2016, 2016
Ranjith P. Udawatta, Clark J. Gantzer, Stephen H. Anderson, and Shmuel Assouline
SOIL, 2, 211–220, https://doi.org/10.5194/soil-2-211-2016, https://doi.org/10.5194/soil-2-211-2016, 2016
Short summary
Short summary
Soil compaction degrades soil structure and affects water, heat, and gas exchange as well as root penetration and crop production. The objective of this study was to use X-ray computed microtomography (CMT) techniques to compare differences in geometrical soil pore parameters as influenced by compaction of two different aggregate size classes.
B. Reidy, I. Simo, P. Sills, and R. E. Creamer
SOIL, 2, 25–39, https://doi.org/10.5194/soil-2-25-2016, https://doi.org/10.5194/soil-2-25-2016, 2016
Short summary
Short summary
This study reviews pedotransfer functions from the literature for different soil and horizon types. It uses these formulae to predict bulk density (ρb) per horizon using measured data of other soil properties. These data were compared to known pb per horizon and recalibrated. These calculations were used to fill missing horizon data in the Irish soil database. This allowed the generation of a pb map to 50 cm. These pb data are at horizon level allowing more accurate estimation of C with depth.
J. J. Keizer, M. A. S. Martins, S. A. Prats, L. F. Santos, D. C. S. Vieira, R. Nogueira, and L. Bilro
SOIL, 1, 641–650, https://doi.org/10.5194/soil-1-641-2015, https://doi.org/10.5194/soil-1-641-2015, 2015
Short summary
Short summary
In this study, a novel plastic optical fibre turbidity sensor was exhaustively tested with a large set of runoff samples, mainly from a recently burnt area. The different types of samples from the distinct study sites revealed without exception an increase in normalized light loss with increasing sediment concentrations that agreed (reasonably) well with a power function. Nevertheless, sensor-based predictions of sediment concentration should ideally involve site-specific calibrations.
C. Rasmussen, R. E. Gallery, and J. S. Fehmi
SOIL, 1, 631–639, https://doi.org/10.5194/soil-1-631-2015, https://doi.org/10.5194/soil-1-631-2015, 2015
Short summary
Short summary
There is a need to understand the response of soil systems to predicted climate warming for modeling soil processes. Current experimental methods for soil warming include expensive and difficult to implement active and passive techniques. Here we test a simple, inexpensive in situ passive soil heating approach, based on easy to construct infrared mirrors that do not require automation or enclosures. Results indicated that the infrared mirrors yielded significant heating and drying of soils.
E. Nadal-Romero, J. Revuelto, P. Errea, and J. I. López-Moreno
SOIL, 1, 561–573, https://doi.org/10.5194/soil-1-561-2015, https://doi.org/10.5194/soil-1-561-2015, 2015
Short summary
Short summary
Geomatic techniques have been routinely applied in erosion studies, providing the opportunity to build high-resolution topographic models.The aim of this study is to assess and compare the functioning of terrestrial laser scanner and close range photogrammetry techniques to evaluate erosion and deposition processes in a humid badlands area.
Our results demonstrated that north slopes experienced more intense and faster dynamics than south slopes as well as the highest erosion rates.
L. M. Thomsen, J. E. M. Baartman, R. J. Barneveld, T. Starkloff, and J. Stolte
SOIL, 1, 399–410, https://doi.org/10.5194/soil-1-399-2015, https://doi.org/10.5194/soil-1-399-2015, 2015
B. A. Miller, S. Koszinski, M. Wehrhan, and M. Sommer
SOIL, 1, 217–233, https://doi.org/10.5194/soil-1-217-2015, https://doi.org/10.5194/soil-1-217-2015, 2015
Short summary
Short summary
There are many different strategies for mapping SOC, among which is to model the variables needed to calculate the SOC stock indirectly or to model the SOC stock directly. The purpose of this research was to compare these two approaches for mapping SOC stocks from multiple linear regression models applied at the landscape scale via spatial association. Although the indirect approach had greater spatial variation and higher R2 values, the direct approach had a lower total estimated error.
W. Eugster and L. Merbold
SOIL, 1, 187–205, https://doi.org/10.5194/soil-1-187-2015, https://doi.org/10.5194/soil-1-187-2015, 2015
Short summary
Short summary
The eddy covariance (EC) method has become increasingly popular in soil science. The basic concept of this method and its use in different types of experimental designs in the field are given, and we indicate where progress in advancing and extending the field of applications is made. The greatest strengths of EC measurements in soil science are (1) their uninterrupted continuous measurement of gas concentrations and fluxes and (2) spatial integration over
small-scale heterogeneity in the soil.
Cited articles
Ahmad, S., Kalra, A., and Stephen, H.: Estimating soil moisture using remote
sensing data: A machine learning approach, Adv. Water Resour., 33,
69–80, 2010. a
Ahmed, O., Habbani, F. I., Mustafa, A., Mohamed, E., Salih, A., and Seedig, F.:
Quality assessment statistic evaluation of X-ray fluorescence via NIST and
IAEA standard reference materials, World Journal of Nuclear Science and
Technology, 7, 121–128, 2017. a
Arrouays, D., Grundy, M. G., Hartemink, A. E., Hempel, J. W., Heuvelink, G. B.,
Hong, S. Y., Lagacherie, P., Lelyk, G., McBratney, A. B., McKenzie, N., Mendonca-Santos, M. d. L., Minasny, B., Montanarella, L., Odeh, I., Sanchez, P., Thompson, J., and Zhang, G.: GlobalSoilMap: Toward a fine-resolution global grid of soil
properties, in: Advances in agronomy, Elsevier,
125, 93–134, 2014. a
Beguin, J., Fuglstad, G.-A., Mansuy, N., and Paré, D.: Predicting soil
properties in the Canadian boreal forest with limited data: Comparison of
spatial and non-spatial statistical approaches, Geoderma, 306, 195–205,
2017. a
Blanco, C. M. G., Gomez, V. M. B., Crespo, P., and Ließ, M.: Spatial
prediction of soil water retention in a Páramo landscape: Methodological
insight into machine learning using random forest, Geoderma, 316, 100–114,
2018. a
Bondi, G., Creamer, R., Ferrari, A., Fenton, O., and Wall, D.: Using machine
learning to predict soil bulk density on the basis of visual parameters:
Tools for in-field and post-field evaluation, Geoderma, 318, 137–147, 2018. a
Børgesen, C. D. and Schaap, M. G.: Point and parameter pedotransfer
functions for water retention predictions for Danish soils, Geoderma, 127,
154–167, 2005. a
Bui, E. N., Henderson, B. L., and Viergever, K.: Knowledge discovery from
models of soil properties developed through data mining, Ecol.
Modell., 191, 431–446, 2006. a
Butler, B. M., O'Rourke, S. M., and Hillier, S.: Using rule-based regression
models to predict and interpret soil properties from X-ray powder diffraction
data, Geoderma, 329, 43–53, https://doi.org/10.1016/j.geoderma.2018.04.005, 2018. a
Cao, B., Domke, G. M., Russell, M. B., and Walters, B. F.: Spatial modeling of
litter and soil carbon stocks on forest land in the conterminous United
States, Sci. Total Environ., 654, 94–106, 2019. a
Castro-Franco, M., Domenech, M. B., Borda, M. R., and Costa, J.: Spatial
dataset of topsoil texture for the southern Argentine Pampas, Geoderma
Regional, 12, 18–27, 2017. a
Catlett, J.: Mega induction: A test flight, in: Machine Learning Proceedings
1991, Proceedings of the Eighth International Conference, Evanston, Elsevier, Illinois, 596–599, 1991. a
Caubet, M., Dobarco, M. R., Arrouays, D., Minasny, B., and Saby, N. P.:
Merging country, continental and global predictions of soil texture: Lessons
from ensemble modelling in France, Geoderma, 337, 99–110, 2019. a
Chlingaryan, A., Sukkarieh, S., and Whelan, B.: Machine learning approaches
for crop yield prediction and nitrogen status estimation in precision
agriculture: A review, Comput. Electron. Agr., 151,
61–69, 2018. a
Coopersmith, E. J., Minsker, B. S., Wenzel, C. E., and Gilmore, B. J.: Machine
learning assessments of soil drying for agricultural planning, Comput.
Electron. Agr., 104, 93–104, 2014. a
Cortes, C., Jackel, L. D., Solla, S. A., Vapnik, V., and Denker, J. S.:
Learning curves: Asymptotic values and rate of convergence, in: Advances
in Neural Information Processing Systems, 327–334, 1994. a
Costa, J. G., Reigosa, M., Matías, J., and Covelo, E.: Soil Cd, Cr, Cu,
Ni, Pb and Zn sorption and retention models using SVM: variable selection and
competitive model, Sci. Total Environ., 593, 508–522, 2017. a
Dai, F., Zhou, Q., Lv, Z., Wang, X., and Liu, G.: Spatial prediction of soil
organic matter content integrating artificial neural network and ordinary
kriging in Tibetan Plateau, Ecol. Indic., 45, 184–194, 2014. a
Dharumarajan, S., Hegde, R., and Singh, S.: Spatial prediction of major soil
properties using Random Forest techniques-A case study in semi-arid tropics
of South India, Geoderma Regional, 10, 154–162, 2017. a
Dobarco, M. R., Cousin, I., Le Bas, C., and Martin, M. P.: Pedotransfer
functions for predicting available water capacity in French soils, their
applicability domain and associated uncertainty, Geoderma, 336, 81–95,
2019. a
Doherty, M. E. and Balzer, W. K.: Cognitive feedback, in: Advances in
psychology, Elsevier, 54, 163–197, 1988. a
Dybczyński, R., Tugsavul, A., and Suschny, O.: Soil-5, a new IAEA
certified reference material for trace element determinations, Geostandard.
Newslett., 3, 61–87, 1979. a
Fajardo, M., McBratney, A., and Whelan, B.: Fuzzy clustering of Vis–NIR
spectra for the objective recognition of soil morphological horizons in soil
profiles, Geoderma, 263, 244–253, 2016. a
Farfani, H. A., Behnamfar, F., and Fathollahi, A.: Dynamic analysis of
soil-structure interaction using the neural networks and the support vector
machines, Expert Syst. Appl., 42, 8971–8981, 2015. a
Feng, Y., Cui, N., Hao, W., Gao, L., and Gong, D.: Estimation of soil
temperature from meteorological data using different machine learning
models, Geoderma, 338, 67–77, 2019. a
Flynn, T., Rozanov, A., de Clercq, W., Warr, B., and Clarke, C.:
Semi-automatic disaggregation of a national resource inventory into a
farm-scale soil depth class map, Geoderma, 337, 1136–1145, 2019. a
Friedman, J. H.: Greedy function approximation: a gradient boosting machine,
Ann. Stat., 29, 1189–1232, 2001. a
Gal, Y. and Ghahramani, Z.: Dropout as a bayesian approximation: Representing
model uncertainty in deep learning, in: International conference on machine
learning, 1050–1059, 2016. a
Gao, M., Li, H.-Y., Liu, D., Tang, J., Chen, X., Chen, X., Blöschl, G., and
Leung, L. R.: Identifying the dominant controls on macropore flow velocity
in soils: A meta-analysis, J. Hydrol., 567, 590–604, 2018. a
Geissen, V., Kampichler, C., López-de Llergo-Juárez, J., and
Galindo-Acántara, A.: Superficial and subterranean soil erosion in
Tabasco, tropical Mexico: development of a decision tree modeling approach,
Geoderma, 139, 277–287, 2007. a
Gomes, L. C., Faria, R. M., de Souza, E., Veloso, G. V., Schaefer, C. E. G.,
and Fernandes Filho, E. I.: Modelling and mapping soil organic carbon
stocks in Brazil, Geoderma, 340, 337–350, 2019. a
Greifeneder, F., Khamala, E., Sendabo, D., Wagner, W., Zebisch, M., Farah, H.,
and Notarnicola, C.: Detection of soil moisture anomalies based on
Sentinel-1, Phys. Chem. Earth, Pt. A/B/C, 112, 75–82, 2018. a
Grinand, C., Le Maire, G., Vieilledent, G., Razakamanarivo, H., Razafimbelo,
T., and Bernoux, M.: Estimating temporal changes in soil carbon stocks at
ecoregional scale in Madagascar using remote-sensing, Int. J.
Appl. Earth Obs., 54, 1–14, 2017. a
Grunwald, S.: Multi-criteria characterization of recent digital soil mapping
and modeling approaches, Geoderma, 152, 195–207, 2009. a
Grunwald, S.: What do we really know about the space–time continuum of
soil-landscapes?, in: Environmental Soil-Landscape Modeling,
CRC Press, 16–49, 2016. a
Grunwald, S., Vasques, G. M., and Rivero, R. G.: Fusion of soil and remote
sensing data to model soil properties, Adv. Agron., Elsevier, 131,
1–109, 2015. a
Han, J., Mao, K., Xu, T., Guo, J., Zuo, Z., and Gao, C.: A soil moisture
estimation framework based on the cart algorithm and its application in
china, J. Hydrol., 563, 65–75, 2018. a
Hanna, A. M., Ural, D., and Saygili, G.: Neural network model for liquefaction
potential in soil deposits using Turkey and Taiwan earthquake data, Soil
Dyn. Earthq. Eng., 27, 521–540, 2007. a
Heggemann, T., Welp, G., Amelung, W., Angst, G., Franz, S. O., Koszinski, S.,
Schmidt, K., and Pätzold, S.: Proximal gamma-ray spectrometry for
site-independent in situ prediction of soil texture on ten heterogeneous
fields in Germany using support vector machines, Soil Till. Res.,
168, 99–109, https://doi.org/10.1016/j.still.2016.10.008, 2017. a
Hutson, M.: Boycott highlights AI's publishing rebellion, Science, 360, p. 699, 2018. a
Ivushkin, K., Bartholomeus, H., Bregt, A. K., Pulatov, A., Bui, E. N., and
Wilford, J.: Soil salinity assessment through satellite thermography for
different irrigated and rainfed crops, Int. J. Appl.
Earth Obs., 68, 230–237, 2018. a
Karandish, F. and Šimŭnek, J.: A field-modeling study for assessing
temporal variations of soil-water-crop interactions under water-saving
irrigation strategies, Agr. Water Manage., 178, 291–303, 2016. a
Khadim, F. K., Su, H., Xu, L., and Tian, J.: Soil salinity mapping in
Everglades National Park using remote sensing techniques and vegetation salt
tolerance, Phys. Chem. Earth, Pt. A/B/C, 110, 31–50, 2019. a
Kheir, R. B., Chorowicz, J., Abdallah, C., and Dhont, D.: Soil and bedrock
distribution estimated from gully form and frequency: A GIS-based
decision-tree model for Lebanon, Geomorphology, 93, 482–492,
https://doi.org/10.1016/j.geomorph.2007.03.010, 2008. a
Koenker, R. and Bassett Jr., G.: Regression quantiles, Econometrica: journal
of the Econometric Society, 33–50, 1978. a
Koestel, J. and Jorda, H.: What determines the strength of preferential
transport in undisturbed soil under steady-state flow?, Geoderma, 217,
144–160, 2014. a
Kohavi, R.: A study of cross-validation and bootstrap for accuracy
estimation and model selection, in: Ijcai,
Montreal, Canada, 14, 1137–1145, 1995. a
Kovačević, M., Bajat, B., and Gajić, B.: Soil type classification
and estimation of soil properties using support vector machines, Geoderma,
154, 340–347, 2010. a
Lacoste, M., Lemercier, B., and Walter, C.: Regional mapping of soil parent
material by machine learning based on point data, Geomorphology, 133,
90–99, https://doi.org/10.1016/j.geomorph.2011.06.026, 2011. a
Leenaars, J. G., Claessens, L., Heuvelink, G. B., Hengl, T., González,
M. R., van Bussel, L. G., Guilpart, N., Yang, H., and Cassman, K. G.:
Mapping rootable depth and root zone plant-available water holding capacity
of the soil of sub-Saharan Africa, Geoderma, 324, 18–36, 2018. a
Liang, Z., Chen, S., Yang, Y., Zhao, R., Shi, Z., and Rossel, R. A. V.:
National digital soil map of organic matter in topsoil and its associated
uncertainty in 1980's China, Geoderma, 335, 47–56, 2019. a
Lin, J.: Divergence measures based on the Shannon entropy, IEEE T. Inform. Theory, 37, 145–151, 1991. a
Liu, S., Yang, Y., Shen, H., Hu, H., Zhao, X., Li, H., Liu, T., and Fang, J.:
No significant changes in topsoil carbon in the grasslands of northern China
between the 1980s and 2000s, Sci. Total Environ., 624,
1478–1487, 2018. a
Lou, Y., Caruana, R., and Gehrke, J.: Intelligible models for classification
and regression, in: Proceedings of the 18th ACM SIGKDD international
conference on Knowledge discovery and data mining, ACM, 150–158, 2012. a
Lu, W., Lu, D., Wang, G., Wu, J., Huang, J., and Li, G.: Examining soil
organic carbon distribution and dynamic change in a hickory plantation region
with Landsat and ancillary data, Catena, 165, 576–589, 2018. a
Ma, Y., Minasny, B., and Wu, C.: Mapping key soil properties to support
agricultural production in Eastern China, Geoderma Regional, 10, 144–153,
2017. a
Ma, Y., Minasny, B., Malone, B. P., and Mcbratney, A. B.: Pedology and digital soil mapping (DSM), Europ. J. Soil Sci., 70, 216–235, 2019. a
Mansuy, N., Thiffault, E., Paré, D., Bernier, P., Guindon, L., Villemaire,
P., Poirier, V., and Beaudoin, A.: Digital mapping of soil properties in
Canadian managed forests at 250 m of resolution using the k-nearest neighbor
method, Geoderma, 235, 59–73, 2014. a
Märker, M., Pelacani, S., and Schröder, B.: A functional entity
approach to predict soil erosion processes in a small Plio-Pleistocene
Mediterranean catchment in Northern Chianti, Italy, Geomorphology, 125,
530–540, 2011. a
Martin, M., Orton, T., Lacarce, E., Meersmans, J., Saby, N., Paroissien, J.,
Jolivet, C., Boulonne, L., and Arrouays, D.: Evaluation of modelling
approaches for predicting the spatial distribution of soil organic carbon
stocks at the national scale, Geoderma, 223, 97–107, 2014. a
Martinez, G., Weltz, M., Pierson, F. B., Spaeth, K. E., and Pachepsky, Y.:
Scale effects on runoff and soil erosion in rangelands: Observations and
estimations with predictors of different availability, Catena, 151,
161–173, 2017. a
Massawe, B. H., Subburayalu, S. K., Kaaya, A. K., Winowiecki, L., and Slater,
B. K.: Mapping numerically classified soil taxa in Kilombero Valley,
Tanzania using machine learning, Geoderma, 311, 143–148, 2018. a
Matthew and Honnibal, M. I.: spaCy 2: Natural language understanding with
Bloom embeddings, convolutional neural networks and incremental parsing,
https://github.com/explosion/spaCy/ (last access: 5 February 2020), 2017. a
McBratney, A., de Gruijter, J., and Bryce, A.: Pedometrics timeline,
Geoderma, 338, 568–575, 2019. a
McCallum, A. K.: MALLET: A Machine Learning for Language Toolkit,
http://mallet.cs.umass.edu (last access: 5 February 2020), 2002. a
Minasny, B. and Flantis, D.: “Helicopter research”: who benefits from
international studies in Indonesia?,
https://theconversation.com/helicopter-research-who-benefits-from-international-studies-in-indonesia-102165
(last access: 29 April 2019), 2018. a
Mjolsness, E. and DeCoste, D.: Machine learning for science: state of the art
and future prospects, Science, 293, 2051–2055, 2001. a
Montavon, G., Samek, W., and Müller, K.-R.: Methods for interpreting and
understanding deep neural networks, Digit. Signal Process., 73, 1–15,
2018. a
Morellos, A., Pantazi, X.-E., Moshou, D., Alexandridis, T., Whetton, R.,
Tziotzios, G., Wiebensohn, J., Bill, R., and Mouazen, A. M.: Machine
learning based prediction of soil total nitrogen, organic carbon and moisture
content by using VIS-NIR spectroscopy, Biosyst. Eng., 152,
104–116, 2016. a
Mutanga, O., Adam, E., and Cho, M. A.: High density biomass estimation for
wetland vegetation using WorldView-2 imagery and random forest regression
algorithm, Int. J. Appl. Earth Obs., 18, 399–406, 2012. a
Naderi-Boldaji, M., Tekeste, M. Z., Nordstorm, R. A., Barnard, D. J., and
Birrel, S. J.: A mechanical-dielectric-high frequency acoustic sensor fusion
for soil physical characterization, Comput. Electron.
Agr., 156, 10–23, 2019. a
Oh, Y.-Y., Yun, S.-T., Yu, S., Kim, H.-J., and Jun, S.-C.: A novel
wavelet-based approach to characterize dynamic environmental factors
controlling short-term soil surface CO2 flux: Application to a controlled CO2
release test site (EIT) in South Korea, Geoderma, 337, 76–90, 2019. a
Padarian, J., Minasny, B., and McBratney, A.: Using deep learning to predict
soil properties from regional spectral data, Geoderma Regional, 16,
e00198, https://doi.org/10.1016/j.geodrs.2018.e00198, 2019b. a, b
Padarian, J., Minasny, B., and McBratney, A. B.: Using deep learning for
digital soil mapping, Soil, 5, 79–89, 2019c. a
Pasini, A.: Artificial neural networks for small dataset analysis, J.
Thoracic Dis., 7, 953–960, 2015. a
Perlich, C., Provost, F., and Simonoff, J. S.: Tree induction vs. logistic
regression: A learning-curve analysis, J. Mach. Learn. Res.,
4, 211–255, 2003. a
Poggio, L., Gimona, A., Spezia, L., and Brewer, M. J.: Bayesian spatial
modelling of soil properties and their uncertainty: The example of soil
organic matter in Scotland using R-INLA, Geoderma, 277, 69–82, 2016. a
Prasad, R., Deo, R. C., Li, Y., and Maraseni, T.: Ensemble committee-based
data intelligent approach for generating soil moisture forecasts with
multivariate hydro-meteorological predictors, Soil Till. Res.,
181, 63–81, 2018. a
Probst, P., Wright, M. N., and Boulesteix, A.-L.: Hyperparameters and tuning
strategies for random forest, WIRES Data Min.
Knowl., 9, e1301, https://doi.org/10.1002/widm.1301, 2019. a
Pueyo, M., Rauret, G., Bacon, J., Gomez, A., Muntau, H., Quevauviller, P., and
López-Sánchez, J.: A new organic-rich soil reference material
certified for its EDTA-and acetic acid-extractable contents of Cd, Cr, Cu,
Ni, Pb and Zn, following collaboratively tested and harmonised procedures,
J. Environ. Monit., 3, 238–242, 2001. a
Rauber, P. E., Fadel, S. G., Falcao, A. X., and Telea, A. C.: Visualizing the
hidden activity of artificial neural networks, IEEE T.
Vis. Comput. Gr., 23, 101–110, 2017. a
Reale, C., Gavin, K., Librić, L., and Jurić-Kaćunić, D.:
Automatic classification of fine-grained soils using CPT measurements and
Artificial Neural Networks, Adv. Eng. Inform., 36, 207–215,
2018. a
Reeves, M. K., Perdue, M., Munk, L. A., and Hagedorn, B.: Predicting risk of
trace element pollution from municipal roads using site-specific soil samples
and remotely sensed data, Sci. Total Environ., 630, 578–586,
2018. a
Rial, M., Cortizas, A. M., Taboada, T., and Rodríguez-Lado, L.: Soil
organic carbon stocks in Santa Cruz Island, Galapagos, under different
climate change scenarios, Catena, 156, 74–81, 2017. a
Röder, M., Both, A., and Hinneburg, A.: Exploring the space of topic
coherence measures, in: Proceedings of the eighth ACM international
conference on Web search and data mining, ACM, 399–408, 2015. a
Rudin, C. and Wagstaff, K. L.: Machine learning for science and society, Mach. Learn., 95, 1–9,
2014. a
Sagasti, F. R.: Underdevelopment, science and technology: the point of view of
the underdeveloped countries, Sci. Stud., 3, 47–59, 1973. a
Schaap, M. G. and Bouten, W.: Modeling water retention curves of sandy soils
using neural networks, Water Resour. Res., 32, 3033–3040, 1996. a
Schillaci, C., Acutis, M., Lombardo, L., Lipani, A., Fantappie, M., Märker,
M., and Saia, S.: Spatio-temporal topsoil organic carbon mapping of a
semi-arid Mediterranean region: The role of land use, soil texture,
topographic indices and the influence of remote sensing data to modelling,
Sci. Total Environ., 601, 821–832, 2017a. a
Schillaci, C., Lombardo, L., Saia, S., Fantappiè, M., Märker, M., and
Acutis, M.: Modelling the topsoil carbon stock of agricultural lands with
the Stochastic Gradient Treeboost in a semi-arid Mediterranean region,
Geoderma, 286, 35–45, 2017b. a
Shavlik, J. W., Mooney, R. J., and Towell, G. G.: Symbolic and neural learning
algorithms: An experimental comparison, Mach. Learn., 6, 111–143, 1991. a
Shaw, J., West, L., Radcliffe, D., and Bosch, D.: Preferential flow and
pedotransfer functions for transport properties in sandy Kandiudults, Soil
Sci. Soc. Am. J., 64, 670–678, 2000. a
Snoek, J., Larochelle, H., and Adams, R. P.: Practical bayesian optimization
of machine learning algorithms, Adv. Neur. In., 25, 2951–2959, 2012. a
Snoek, J., Rippel, O., Swersky, K., Kiros, R., Satish, N., Sundaram, N.,
Patwary, M., Prabhat, M., and Adams, R.: Scalable bayesian optimization
using deep neural networks, Proceedings of the 32nd International Conference on Machine Learning, Int. Conf. Mach.
Learn., 2171–2180, 2015. a
Somarathna, P., Minasny, B., and Malone, B. P.: More data or a better model?
Figuring out what matters most for the spatial prediction of soil carbon,
Soil Sci. Soc. Am. J., 81, 1413–1426, 2017. a
Song, X.-D., Yang, F., Ju, B., Li, D.-C., Zhao, Y.-G., Yang, J.-L., and Zhang,
G.-L.: The influence of the conversion of grassland to cropland on changes
in soil organic carbon and total nitrogen stocks in the Songnen Plain of
Northeast China, Catena, 171, 588–601, 2018. a
Stevens, A., van Wesemael, B., Bartholomeus, H., Rosillon, D., Tychon, B., and
Ben-Dor, E.: Laboratory, field and airborne spectroscopy for monitoring
organic carbon content in agricultural soils, Geoderma, 144, 395–404, 2008. a
Stevens, K., Kegelmeyer, P., Andrzejewski, D., and Buttler, D.: Exploring
topic coherence over many models and many topics, in: Proceedings of the
2012 Joint Conference on Empirical Methods in Natural Language Processing and
Computational Natural Language Learning, Association for
Computational Linguistics, 952–961, 2012. a, b
Stine, R. A.: Bootstrap prediction intervals for regression, J.
Am. Stat. Assoc., 80, 1026–1031, 1985. a
Stumpf, F., Keller, A., Schmidt, K., Mayr, A., Gubler, A., and Schaepman, M.:
Spatio-temporal land use dynamics and soil organic carbon in Swiss
agroecosystems, Agr. Ecosyst. Environ., 258, 129–142, 2018. a
Subburayalu, S., Jenhani, I., and Slater, B.: Disaggregation of component soil
series on an Ohio County soil survey map using possibilistic decision trees,
Geoderma, 213, 334–345, 2014. a
Sugimoto, C. R., Li, D., Russell, T. G., Finlay, S. C., and Ding, Y.: The
shifting sands of disciplinary development: Analyzing North American Library
and Information Science dissertations using latent Dirichlet allocation,
J. Am. Soc. Inf. Sci. Tech., 62,
185–204, 2011. a
Taghizadeh-Mehrjardi, R., Nabiollahi, K., Minasny, B., and Triantafilis, J.:
Comparing data mining classifiers to predict spatial distribution of
USDA-family soil groups in Baneh region, Iran, Geoderma, 253, 67–77, 2015. a
Taghizadeh-Mehrjardi, R., Nabiollahi, K., and Kerry, R.: Digital mapping of
soil organic carbon at multiple depths using different data mining techniques
in Baneh region, Iran, Geoderma, 266, 98–110, 2016. a
Tomasella, J., Hodnett, M. G., and Rossato, L.: Pedotransfer Functions for the
Estimation of Soil Water Retention in Brazilian Soils, Soil Sci. Soc. Am. J., 64, 327–338, 2000. a
Tranter, G., Minasny, B., and McBratney, A.: Estimating Pedotransfer Function
Prediction Limits Using Fuzzy k-Means with Extragrades, Soil Sci. Soc. Am.
J., 74, 1967–1975, 2010. a
Tziachris, P., Aschonitis, V., Chatzistathis, T., and Papadopoulou, M.:
Assessment of spatial hybrid methods for predicting soil organic matter
using DEM derivatives and soil parameters, Catena, 174, 206–216, 2019. a
Vaysse, K. and Lagacherie, P.: Using quantile regression forest to estimate
uncertainty of digital soil mapping products, Geoderma, 291, 55–64, 2017. a
Vincent, S., Lemercier, B., Berthier, L., and Walter, C.: Spatial
disaggregation of complex Soil Map Units at the regional scale based on
soil-landscape relationships, Geoderma, 311, 130–142, 2018. a
Viscarra-Rossel, R. and Behrens, T.: Using data mining to model and interpret
soil diffuse reflectance spectra, Geoderma, 158, 46–54, 2010. a
Řehůřek, R. and Sojka, P.: Software Framework for Topic Modelling
with Large Corpora, in: Proceedings of the LREC 2010 Workshop on New
Challenges for NLP Frameworks, ELRA, Valletta, Malta,
45–50, http://is.muni.cz/publication/884893/en (last access: 5 February 2020), 2010. a
Wang, B., Waters, C., Orgill, S., Gray, J., Cowie, A., Clark, A., and Li Liu,
D.: High resolution mapping of soil organic carbon stocks using remote
sensing variables in the semi-arid rangelands of eastern Australia, Sci. Total Environ., 630, 367–378, 2018b. a
Ware, M. and Mabe, M.: The STM report: An overview of scientific and scholarly
journal publishing, available at: https://www.stm-assoc.org/2015_02_20_STM_Report_2015.pdf (last access: 5 February 2020), 2015.
a
Warner, D. L., Guevara, M., Inamdar, S., and Vargas, R.: Upscaling
soil-atmosphere CO2 and CH4 fluxes across a topographically complex forested
landscape, Agr. Forest Meteorol., 264, 80–91, 2019. a
Watson, S. J., Luck, G. W., Spooner, P. G., and Watson, D. M.: Land-use
change: incorporating the frequency, sequence, time span, and magnitude of
changes into ecological research, Front. Ecol. Environ.,
12, 241–249, 2014. a
Were, K., Bui, D. T., Dick, Ø. B., and Singh, B. R.: A comparative
assessment of support vector regression, artificial neural networks, and
random forests for predicting and mapping soil organic carbon stocks across
an Afromontane landscape, Ecol. Ind., 52, 394–403, 2015. a
Wu, G., Kechavarzi, C., Li, X., Wu, S., Pollard, S. J., Sui, H., and Coulon,
F.: Machine learning models for predicting PAHs bioavailability in compost
amended soils, Chem. Engin. J., 223, 747–754, 2013. a
Wu, Q., Zhang, C., Hong, Q., and Chen, L.: Topic evolution based on LDA and
HMM and its application in stem cell research, J. Inf.
Sci., 40, 611–620, 2014. a
Xie, X.-L. and Li, A.-B.: Identification of soil profile classes using
depth-weighted visible-near-infrared spectral reflectance,
Geoderma, 325, 90–101, https://doi.org/10.1016/j.geoderma.2018.03.029, 2018. a
Xing, L., Li, L., Gong, J., Ren, C., Liu, J., and Chen, H.: Daily soil
temperatures predictions for various climates in United States using
data-driven model, Energy, 160, 430–440, 2018. a
Xu, Y., Smith, S. E., Grunwald, S., Abd-Elrahman, A., and Wani, S. P.:
Incorporation of satellite remote sensing pan-sharpened imagery into digital
soil prediction and mapping models to characterize soil property variability
in small agricultural fields, ISPRS J Photogramm., 123, 1–19, 2017. a
Zeynoddin, M., Bonakdari, H., Ebtehaj, I., Esmaeilbeiki, F., Gharabaghi, B.,
and Haghi, D. Z.: A reliable linear stochastic daily soil temperature
forecast model, Soil Till. Res., 189, 73–87, 2019. a
Zhang, C., Mishra, D. R., and Pennings, S. C.: Mapping salt marsh soil
properties using imaging spectroscopy, ISPRS J. Photogramm., 148, 221–234, 2019. a
Zhang, Q., Nian Wu, Y., and Zhu, S.-C.: Interpretable convolutional neural
networks, in: Proceedings of the IEEE Conference on Computer Vision and
Pattern Recognition, 8827–8836, 2018a. a
Zhou, D., Ji, X., Zha, H., and Giles, C. L.: Topic evolution and social
interactions: how authors effect research, in: Proceedings of the 15th ACM
international conference on Information and knowledge management,
ACM, 248–257, 2006. a
Short summary
The application of machine learning (ML) has shown an accelerated adoption in soil sciences. It is a difficult task to manually review all papers on the application of ML. This paper aims to provide a review of the application of ML aided by topic modelling in order to find patterns in a large collection of publications. The objective is to gain insight into the applications and to discuss research gaps. We found 12 main topics and that ML methods usually perform better than traditional ones.
The application of machine learning (ML) has shown an accelerated adoption in soil sciences. It...
