Abstract. The spatial assessment of soil functions requires maps of basic soil properties. Unfortunately, these are either missing for many regions or are not available at the desired spatial resolution or down to the required soil depth. The field-based generation of large soil datasets and conventional soil maps remains costly. Meanwhile, legacy soil data and comprehensive sets of spatial environmental data are available for many regions.Digital soil mapping (DSM) approaches relating soil data (responses) to environmental data (covariates) face the challenge of building statistical models from large sets of covariates originating, for example, from airborne imaging spectroscopy or multi-scale terrain analysis. We evaluated six approaches for DSM in three study regions in Switzerland (Berne, Greifensee, ZH forest) by mapping the effective soil depth available to plants (SD), pH, soil organic matter (SOM), effective cation exchange capacity (ECEC), clay, silt, gravel content and fine fraction bulk density for four soil depths (totalling 48 responses). Models were built from 300-500 environmental covariates by selecting linear models through (1) grouped lasso and (2) an ad hoc stepwise procedure for robust external-drift kriging (georob). For (3) geoadditive models we selected penalized smoothing spline terms by component-wise gradient boosting (geoGAM). We further used two tree-based methods: (4) boosted regression trees (BRTs) and (5) random forest (RF). Lastly, we computed (6) weighted model averages (MAs) from the predictions obtained from methods 1-5.Lasso, georob and geoGAM successfully selected strongly reduced sets of covariates (subsets of 3-6 % of all covariates). Differences in predictive performance, tested on independent validation data, were mostly small and did not reveal a single best method for 48 responses. Nevertheless, RF was often the best among methods 1-5 (28 of 48 responses), but was outcompeted by MA for 14 of these 28 responses. RF tended to over-fit the data. The performance of BRT was slightly worse than RF. GeoGAM performed poorly on some responses and was the best only for 7 of 48 responses. The prediction accuracy of lasso was intermediate. All models generally had small bias. Only the computationally very efficient lasso had slightly larger bias because it tended to under-fit the data. Summarizing, although differences were small, the frequencies of the best and worst performance clearly favoured RF if a single method is applied and MA if multiple prediction models can be developed.Published by Copernicus Publications on behalf of the European Geosciences Union.

Magnetic resonance force microscopy (MRFM) is a scanning probe technique capable of detecting MRI signals from nanoscale sample volumes, providing a paradigmchanging potential for structural biology and medical research. Thus far, however, experiments have not reached sufficient spatial resolution for retrieving meaningful structural information from samples. In this work, we report MRFM imaging scans demonstrating a resolution of 0.9 nm and a localization precision of 0.6 nm in one dimension. Our progress is enabled by an improved spin excitation protocol furnishing us with sharp spatial control on the MRFM imaging slice, combined with overall advances in instrument stability. From a modeling of the slice function, we expect that our arrangement supports spatial resolutions down to 0.3 nm given sufficient signal-to-noise 1 arXiv:1908.04180v1 [physics.app-ph] 12 Aug 2019 ratio. Our experiment demonstrates the feasibility of sub-nanometer MRI and realizes an important milestone towards the three-dimensional imaging of macromolecular structures.

Soil legacy data rescue via GlobalSoilMap and other international and national initiatives The International Center for Tropical Agriculture (CIAT) believes that open access contributes to its mission of reducing hunger and poverty, and improving human nutrition in the tropics through research aimed at increasing the eco-efficiency of agriculture. CIAT is committed to creating and sharing knowledge and information openly and globally. We do this through collaborative research as well as through the open sharing of our data, tools, and publications.

We report the development of a scanning force microscope based on an ultrasensitive silicon nitride membrane optomechanical transducer. Our development is made possible by inverting the standard microscope geometry-in our instrument, the substrate is vibrating and the scanning tip is at rest. We present topography images of samples placed on the membrane surface. Our measurements demonstrate that the membrane retains an excellent force sensitivity when loaded with samples and in the presence of a scanning tip. We discuss the prospects and limitations of our instrument as a quantum-limited force sensor and imaging tool.

We present a "nanoladder" geometry that minimizes the mechanical dissipation of ultrasensitive cantilevers. A nanoladder cantilever consists of a lithographically patterned scaffold of rails and rungs with feature size ∼100 nm. Compared to a rectangular beam of the same dimensions, the mass and spring constant of a nanoladder are each reduced by roughly 2 orders of magnitude. We demonstrate a low force noise of 158 zN and 190 zN in a 1 Hz bandwidth for devices made from silicon and diamond, respectively, measured at temperatures between 100-150 mK. As opposed to bottom-up mechanical resonators like nanowires or nanotubes, nanoladder cantilevers can be batch-fabricated using standard lithography, which is a critical factor for applications in scanning force microscopy.

Abstract. Spatial assessment of soil functions requires maps of basic soil properties. Unfortunately, these are either missing for many regions or are not available at the desired spatial resolution or down to required soil depth. Conventional soil map generation remains costly. Field based generation of large soil data sets and of conventional soil maps remains costly. Meanwhile, soil legacy data and comprehensive sets of spatial environmental data are available for many regions. Digital soil mapping (DSM) approaches – relating soil data (responses) to environmental data (covariates) – are facing the challenge to build statistical models from large sets of covariates originating for example from airborne imaging spectroscopy or multi-scale terrain analysis. We evaluated six approaches for DSM in three study regions in Switzerland (Berne, Greifensee, ZH forest) by mapping effective soil depth available to plants (SD), pH, soil organic matter (SOM), effective cation exchange capacity (ECEC), clay, silt, gravel content and bulk density for four soil layers (totalling 48 responses). Models were built from 300–500 environmental covariates by selecting linear models by (1) grouped lasso and by an ad-hoc stepwise procedure for (2) robust external-drift kriging (EDK). For (3) geoadditive models we selected penalized smoothing spline terms by componentwise gradient boosting (geoGAM). We further used two tree-based methods: (4) boosted regression trees (BRT) and (5) Random Forest (RF). Lastly, we computed (6) weighted model averages (MA) from predictions obtained from methods 1–5. Lasso, georob and geoGAM successfully selected strongly reduced sets of covariates (subsets of 3–6 % of all covariates). To automatically select a sparse trend model for EDK was however difficult, and the applied ad hoc procedure was computationally inefficient and over-fitted the data. Differences in predictive performance, tested on independent validation data, were mostly small and did not reveal a single best method for 48 responses. Nevertheless, RF was on average often best among methods 1–5 (28 of 48 responses), but was outcompeted by MA for 14 of these 28 responses. RF tended to over-fit the data. Performance of BRT was slightly worse than RF. GeoGAM performed poorly on some responses and was only best for 7 of 48 responses. Predictive precision of lasso was intermediate. All models generally had small bias. Only the computationally very efficient lasso had slightly larger bias likely because it tended to under-fit the data. Summarizing, although differences were small, the frequencies of best and worst performance clearly favoured RF if a single method is applied MA if multiple prediction models can be developed.

Biomedical devices employed in therapy, diagnostics and for self-monitoring often require a high degree of flexibility and compactness. Many near infrared (NIR) optical fiber-coupled systems meet these requirements and are employed on a daily basis. However, mid-infrared (MIR) fibers-based systems have not yet found their way to routine application in medicine. In this work we present the implementation of the first MIR fiber-coupled photoacoustic sensor for the investigation of condensed samples in the MIR fingerprint region. The light of an external-cavity quantum-cascade laser (1010–1095 cm−1) is delivered by a silver halide fiber, which is attached to the PA cell. The PA chamber is conically shaped to perfectly match the beam escaping the fiber and to minimize the cell volume. This results in a compact and handy sensor for investigations of biological samples and the monitoring of constituents both in vitro and in vivo. The performance of the fiber-coupled PA sensor is demonstrated by sensing glucose in aqueous solutions. These measurements yield a detection limit of 57 mg/dL (SNR = 1). Furthermore, the fiber-coupled sensor has been applied to record human skin spectra at different body sites to illustrate its flexibility.

We report on mechanical dissipation measurements carried out on thin (∼100 nm), single-crystal silicon cantilevers with varying chemical surface termination. We find that the 1-2 nm-thick native oxide layer of silicon contributes about 85% to the friction of the mechanical resonance. We show that the mechanical friction is proportional to the thickness of the oxide layer and that it crucially depends on oxide formation conditions. We further demonstrate that chemical surface protection by nitridation, liquid-phase hydrosilylation, or gas-phase hydrosilylation can inhibit rapid oxide formation in air and results in a permanent improvement of the mechanical quality factor between three-and five-fold. This improvement extends to cryogenic temperatures. Presented recipes can be directly integrated with standard cleanroom processes and may be especially beneficial for ultrasensitive nanomechanical force-and mass sensors, including silicon cantilevers, membranes, or nanowires.