Assessment of the Effect of Soil Sample Preparation, Water Content and Excitation Time on Proximal X-ray Fluorescence Sensing.

X-ray fluorescence (XRF) spectroscopy offers a fast and efficient method for analysing soil elemental composition, both in the laboratory and the field. However, the technique is sensitive to spectral interference as well as physical and chemical matrix effects, which can reduce the precision of the measurements. We systematically assessed the XRF technique under different sample preparations, water contents, and excitation times. Four different soil samples were used as blocks in a three-way factorial experiment, with three sample preparations (natural aggregates, ground to ≤2 mm and ≤1 mm), three gravimetric water contents (air-dry, 10% and 20%), and three excitation times (15, 30 and 60 s). The XRF spectra were recorded and gave 540 spectra in all. Elemental peaks for Si, K, Ca, Ti, Fe and Cu were identified for analysis. We used analysis of variance (anova) with post hoc tests to identify significant differences between our factors and used the intensity and area of the elemental peaks as the response. Our results indicate that all of these factors significantly affect the XRF spectrum, but longer excitation times appear to be more defined. In most cases, no significant difference was found between air-dry and 10% water content. Moisture has no apparent effect on coarse samples unless ground to 1 mm. We suggested that the XRF measurements that take 60 s from dry samples or only slightly moist ones might be an optimum option under field conditions.

Distinct controls over the temporal dynamics of soil carbon fractions after land use change.

Soil organic carbon (SOC), the largest terrestrial carbon pool, plays a significant role in soil-related ecosystem services such as climate regulation, soil fertility and agricultural production. However, its fate under land use change is difficult to predict. A major issue is that SOC comprised of numerous organic compounds with potentially distinct and poorly understood turnover properties. Here we use spatiotemporal measurements of the particulate (POC), mineral-associated (MOC) and charred SOC (COC) fractions from 176 trials involving changes in land use to assess their underlying controls. We find that the initial pool sizes of each of the three fractions consistently and dominantly control their temporal dynamics after changes in land use (i.e. the baseline effects). The effects of climate, soil physicochemical properties and plant residues, however, are fraction- and time-dependent. Climate and soil properties show similar importance for controlling the dynamics of MOC and COC, while plant residue inputs (in term of their quantity and quality) are much less important. For POC, plant residues and management practices (e.g. the frequency of pasture in crop-pasture rotation systems) are substantially more important, overriding the influence of climate. These results demonstrate the pivotal role of measuring SOC composition and considering fraction-specific stabilization and destabilization processes for effective SOC management and reliable SOC predictions.

Microbial dynamics and soil physicochemical properties explain large-scale variations in soil organic carbon.

First-order organic matter decomposition models are used within most Earth System Models (ESMs) to project future global carbon cycling; these models have been criticized for not accurately representing mechanisms of soil organic carbon (SOC) stabilization and SOC response to climate change. New soil biogeochemical models have been developed, but their evaluation is limited to observations from laboratory incubations or few field experiments. Given the global scope of ESMs, a comprehensive evaluation of such models is essential using in situ observations of a wide range of SOC stocks over large spatial scales before their introduction to ESMs. In this study, we collected a set of in situ observations of SOC, litterfall and soil properties from 206 sites covering different forest and soil types in Europe and China. These data were used to calibrate the model MIMICS (The MIcrobial-MIneral Carbon Stabilization model), which we compared to the widely used first-order model CENTURY. We show that, compared to CENTURY, MIMICS more accurately estimates forest SOC concentrations and the sensitivities of SOC to variation in soil temperature, clay content and litter input. The ratios of microbial biomass to total SOC predicted by MIMICS agree well with independent observations from globally distributed forest sites. By testing different hypotheses regarding (using alternative process representations) the physicochemical constraints on SOC deprotection and microbial turnover in MIMICS, the errors of simulated SOC concentrations across sites were further decreased. We show that MIMICS can resolve the dominant mechanisms of SOC decomposition and stabilization and that it can be a reliable tool for predictions of terrestrial SOC dynamics under future climate change. It also allows us to evaluate at large scale the rapidly evolving understanding of SOC formation and stabilization based on laboratory and limited filed observation.

Novel Proximal Sensing for Monitoring Soil Organic C Stocks and Condition.

Soil information is needed for environmental monitoring to address current concerns over food, water and energy securities, land degradation, and climate change. We developed the Soil Condition ANalysis System (SCANS) to help address these needs. It integrates an automated soil core sensing system (CSS) with statistical analytics and modeling to characterize soil at fine depth resolutions and across landscapes. The CSS's sensors include a γ-ray attenuation densitometer to measure bulk density, digital cameras to image the measured soil, and a visible-near-infrared (vis-NIR) spectrometer to measure iron oxides and clay mineralogy. The spectra are also modeled to estimate total soil organic carbon (C), particulate, humus, and resistant organic C (POC, HOC, and ROC, respectively), clay content, cation exchange capacity (CEC), pH, volumetric water content, available water capacity (AWC), and their uncertainties. Measurements of bulk density and organic C are combined to estimate C stocks. Kalman smoothing is used to derive complete soil property profiles with propagated uncertainties. The SCANS provides rapid, precise, quantitative, and spatially explicit information about the properties of soil profiles with a level of detail that is difficult to obtain with other approaches. The information gained effectively deepens our understanding of soil and calls attention to the central role soil plays in our environment.

Baseline map of organic carbon in Australian soil to support national carbon accounting and monitoring under climate change.

We can effectively monitor soil condition-and develop sound policies to offset the emissions of greenhouse gases-only with accurate data from which to define baselines. Currently, estimates of soil organic C for countries or continents are either unavailable or largely uncertain because they are derived from sparse data, with large gaps over many areas of the Earth. Here, we derive spatially explicit estimates, and their uncertainty, of the distribution and stock of organic C in the soil of Australia. We assembled and harmonized data from several sources to produce the most comprehensive set of data on the current stock of organic C in soil of the continent. Using them, we have produced a fine spatial resolution baseline map of organic C at the continental scale. We describe how we made it by combining the bootstrap, a decision tree with piecewise regression on environmental variables and geostatistical modelling of residuals. Values of stock were predicted at the nodes of a 3-arc-sec (approximately 90 m) grid and mapped together with their uncertainties. We then calculated baselines of soil organic C storage over the whole of Australia, its states and territories, and regions that define bioclimatic zones, vegetation classes and land use. The average amount of organic C in Australian topsoil is estimated to be 29.7 t ha(-1) with 95% confidence limits of 22.6 and 37.9 t ha(-1) . The total stock of organic C in the 0-30 cm layer of soil for the continent is 24.97 Gt with 95% confidence limits of 19.04 and 31.83 Gt. This represents approximately 3.5% of the total stock in the upper 30 cm of soil worldwide. Australia occupies 5.2% of the global land area, so the total organic C stock of Australian soil makes an important contribution to the global carbon cycle, and it provides a significant potential for sequestration. As the most reliable approximation of the stock of organic C in Australian soil in 2010, our estimates have important applications. They could support Australia's National Carbon Accounting System, help guide the formulation of policy around carbon offset schemes, improve Australia's carbon balances, serve to direct future sampling for inventory, guide the design of monitoring networks and provide a benchmark against which to assess the impact of changes in land cover, land management and climate on the stock of C in Australia. In this way, these estimates would help us to develop strategies to adapt and mitigate the effects of climate change.

Soil organic carbon dust emission: an omitted global source of atmospheric CO2.

Soil erosion redistributes soil organic carbon (SOC) within terrestrial ecosystems, to the atmosphere and oceans. Dust export is an essential component of the carbon (C) and carbon dioxide (CO(2)) budget because wind erosion contributes to the C cycle by removing selectively SOC from vast areas and transporting C dust quickly offshore; augmenting the net loss of C from terrestrial systems. However, the contribution of wind erosion to rates of C release and sequestration is poorly understood. Here, we describe how SOC dust emission is omitted from national C accounting, is an underestimated source of CO(2) and may accelerate SOC decomposition. Similarly, long dust residence times in the unshielded atmospheric environment may considerably increase CO(2) emission. We developed a first approximation to SOC enrichment for a well-established dust emission model and quantified SOC dust emission for Australia (5.83 Tg CO(2)-e yr(-1)) and Australian agricultural soils (0.4 Tg CO(2)-e yr(-1)). These amount to underestimates for CO(2) emissions of ≈10% from combined C pools in Australia (year = 2000), ≈5% from Australian Rangelands and ≈3% of Australian Agricultural Soils by Kyoto Accounting. Northern hemisphere countries with greater dust emission than Australia are also likely to have much larger SOC dust emission. Therefore, omission of SOC dust emission likely represents a considerable underestimate from those nations' C accounts. We suggest that the omission of SOC dust emission from C cycling and C accounting is a significant global source of uncertainty. Tracing the fate of wind-eroded SOC in the dust cycle is therefore essential to quantify the release of CO(2) from SOC dust to the atmosphere and the contribution of SOC deposition to downwind C sinks.

On the interpretability of predictors in spatial data science: the information horizon.

Two important theories in spatial modelling relate to structural and spatial dependence. Structural dependence refers to environmental state-factor models, where an environmental property is modelled as a function of the states and interactions of environmental predictors, such as climate, parent material or relief. Commonly, the functions are regression or supervised classification algorithms. Spatial dependence is present in most environmental properties and forms the basis for spatial interpolation and geostatistics. In machine learning, modelling with geographic coordinates or Euclidean distance fields, which resemble linear variograms with infinite ranges, can produce similar interpolations. Interpolations do not lend themselves to causal interpretations. Conversely, with structural modeling, one can, potentially, extract knowledge from the modelling. Two important characteristics of such interpretable environmental modelling are scale and information content. Scale is relevant because very coarse scale predictors can show nearly infinite ranges, falling out of what we call the information horizon, i.e. interpretation using domain knowledge isn't possible. Regarding information content, recent studies have shown that meaningless predictors, such as paintings or photographs of faces, can be used for spatial environmental modelling of ecological and soil properties, with accurate evaluation statistics. Here, we examine under which conditions modelling with such predictors can lead to accurate statistics and whether an information horizon can be derived for scale and information content.

Modelling potentially toxic elements in forest soils with vis-NIR spectra and learning algorithms.

The surface organic horizons in forest soils have been affected by air and soil pollutants, including potentially toxic elements (PTEs). Monitoring of PTEs requires a large number of samples and adequate analysis. Visible-near infrared (vis-NIR: 350-2500 nm) spectroscopy provides an alternative method to conventional laboratory measurements, which are time-consuming and expensive. However, vis-NIR spectroscopy relies on an empirical calibration of the target attribute to the spectra. This study examined the capability of vis-NIR spectra coupled with machine learning (ML) techniques (partial least squares regression (PLSR), support vector machine regression (SVMR), and random forest (RF)) and a deep learning (DL) approach called fully connected neural network (FNN) to assess selected PTEs (Cr, Cu, Pb, Zn, and Al) in forest organic horizons. The dataset consists of 2160 samples from 1080 sites in the forests over all the Czech Republic. At each site, we collected two samples from the fragmented (F) and humus (H) organic layers. The content of all PTEs was higher in horizon H compared to F horizon. Our results indicate that the reflectance of samples tended to decrease with increased PTEs concentration. Cr was the most accurately predicted element, regardless of the algorithm used. SVMR provided the best results for assessing the H horizon (R2 = 0.88 and RMSE = 3.01 mg/kg for Cr). FNN produced the best predictions of Cr in the combined F + H layers (R2 = 0.89 and RMSE = 2.95 mg/kg) possibly due to the larger number of samples. In the F horizon, the PTEs were not predicted adequately. The study shows that PTEs in forest soils of the Czech Republic can be accurately estimated with vis-NIR spectra and ML approaches. Results hint in availability of a large sample size, FNN provides better results.

The relevant range of scales for multi-scale contextual spatial modelling.

Spatial autocorrelation in the residuals of spatial environmental models can be due to missing covariate information. In many cases, this spatial autocorrelation can be accounted for by using covariates from multiple scales. Here, we propose a data-driven, objective and systematic method for deriving the relevant range of scales, with distinct upper and lower scale limits, for spatial modelling with machine learning and evaluated its effect on modelling accuracy. We also tested an approach that uses the variogram to see whether such an effective scale space can be approximated a priori and at smaller computational cost. Results showed that modelling with an effective scale space can improve spatial modelling with machine learning and that there is a strong correlation between properties of the variogram and the relevant range of scales. Hence, the variogram of a soil property can be used for a priori approximations of the effective scale space for contextual spatial modelling and is therefore an important analytical tool not only in geostatistics, but also for analyzing structural dependencies in contextual spatial modelling.

Multi-scale digital soil mapping with deep learning.

We compared different methods of multi-scale terrain feature construction and their relative effectiveness for digital soil mapping with a Deep Learning algorithm. The most common approach for multi-scale feature construction in DSM is to filter terrain attributes based on different neighborhood sizes, however results can be difficult to interpret because the approach is affected by outliers. Alternatively, one can derive the terrain attributes on decomposed elevation data, but the resulting maps can have artefacts rendering the approach undesirable. Here, we introduce 'mixed scaling' a new method that overcomes these issues and preserves the landscape features that are identifiable at different scales. The new method also extends the Gaussian pyramid by introducing additional intermediate scales. This minimizes the risk that the scales that are important for soil formation are not available in the model. In our extended implementation of the Gaussian pyramid, we tested four intermediate scales between any two consecutive octaves of the Gaussian pyramid and modelled the data with Deep Learning and Random Forests. We performed the experiments using three different datasets and show that mixed scaling with the extended Gaussian pyramid produced the best performing set of covariates and that modelling with Deep Learning produced the most accurate predictions, which on average were 4-7% more accurate compared to modelling with Random Forests.

Current and future assessments of soil erosion by water on the Tibetan Plateau based on RUSLE and CMIP5 climate models.

Soil erosion by water is accelerated by a warming climate and negatively impacts water security and ecological conservation. The Tibetan Plateau (TP) has experienced warming at a rate approximately twice that observed globally, and heavy precipitation events lead to an increased risk of erosion. In this study, we assessed current erosion on the TP and predicted potential soil erosion by water in 2050. The study was conducted in three steps. During the first step, we used the Revised Universal Soil Equation (RUSLE), publicly available data, and the most recent earth observations to derive estimates of annual erosion from 2002 to 2016 on the TP at 1-km resolution. During the second step, we used a multiple linear regression (MLR) model and a set of climatic covariates to predict rainfall erosivity on the TP in 2050. The MLR was used to establish the relationship between current rainfall erosivity data and a set of current climatic and other covariates. The coefficients of the MLR were generalised with climate covariates for 2050 derived from the fifth phase of the Coupled Model Intercomparison Project (CMIP5) models to estimate rainfall erosivity in 2050. During the third step, soil erosion by water in 2050 was predicted using rainfall erosivity in 2050 and other erosion factors. The results show that the mean annual soil erosion rate on the TP under current conditions is 2.76tha-1y-1, which is equivalent to an annual soil loss of 559.59×106t. Our 2050 projections suggested that erosion on the TP will increase to 3.17tha-1y-1 and 3.91tha-1y-1 under conditions represented by RCP2.6 and RCP8.5, respectively. The current assessment and future prediction of soil erosion by water on the TP should be valuable for environment protection and soil conservation in this unique region and elsewhere.

A new detailed map of total phosphorus stocks in Australian soil.

Accurate data are needed to effectively monitor environmental condition, and develop sound policies to plan for the future. Globally, current estimates of soil total phosphorus (P) stocks are very uncertain because they are derived from sparse data, with large gaps over many areas of the Earth. Here, we derive spatially explicit estimates, and their uncertainty, of the distribution and stock of total P in Australian soil. Data from several sources were harmonized to produce the most comprehensive inventory of total P in soil of the continent. They were used to produce fine spatial resolution continental maps of total P in six depth layers by combining the bootstrap, a decision tree with piecewise regression on environmental variables and geostatistical modelling of residuals. Values of percent total P were predicted at the nodes of a 3-arcsecond (approximately 90 m) grid and mapped together with their uncertainties. We combined these predictions with those for bulk density and mapped the total soil P stock in the 0-30 cm layer over the whole of Australia. The average amount of P in Australian topsoil is estimated to be 0.98 t ha(-1) with 90% confidence limits of 0.2 and 4.2 t ha(-1). The total stock of P in the 0-30 cm layer of soil for the continent is 0.91 Gt with 90% confidence limits of 0.19 and 3.9 Gt. The estimates are the most reliable approximation of the stock of total P in Australian soil to date. They could help improve ecological models, guide the formulation of policy around food and water security, biodiversity and conservation, inform future sampling for inventory, guide the design of monitoring networks, and provide a benchmark against which to assess the impact of changes in land cover, land use and management and climate on soil P stocks and water quality in Australia.