Calibration Data Set Research Articles

Using surrogate models and mechanics to predict pressure from strain data

Many marine structures rely upon hull monitoring systems that utilize strain information to make inferences about structural integrity. One of the most significant factors affecting structural integrity are wave loads, specifically the resultant pressures. While direct measurement is not often available, this study explores how to translate strain information into localized pressure estimates through two different low-order modeling paradigms, physics-based and surrogate modeling. Strain data was collected via strain gauges mounted on a scaled ship model that was tested in towing tank under a variety of regular and irregular wave conditions. Using a set of strain calibration data, two different models are developed, one using principal stress and strain relations and the other using a machine learning approach (Gaussian process model), for predicting pressure time histories for different wave conditions. The two modeling approaches are compared in terms of both a variety of error metrics as well as their ability to handle measurement noise and disturbances commonly encountered during marine structures testing and operations.

ОЦЕНКА СИСТЕМ КООРДИНАТНЫХ МЕТОК С ИСПОЛЬЗОВАНИЕМ БОРТОВОЙ КАМЕРЫ МОБИЛЬНОГО РОБОТА «СЕРВОСИЛА ИНЖЕНЕР» В УСЛОВИЯХ ЗАКРЫТЫХ ПОМЕЩЕНИЙ

Modern tasks of special areas of robotics, including search and rescue operations in urbanareas, face a number of obstacles to mobile robotics, where the automatic performance of various functionsby mobile robots remains a key task. One of the important requirements for the algorithms andsoftware of the robot is the possibility of autonomous decision-making and automatic performance bythe robot of various functions, both low and high levels based on the embedded algorithms and informationreceived from the on-board sensors of the robot. To date, the most common on-board robot sensorsare cameras of various types, due to their technical capabilities and lower cost relative to lidarsand other sensors that provide visual information in the form of digital images. Camera calibration is anecessary process for extracting accurate information from digital images. This process is necessary toobtain an exact correspondence between the three-dimensional object space and the pixel space of theimage, for the possibility of subsequent use of computer vision algorithms, aggregation, and informationprocessing. Calibration of digital cameras is an integral part of a number of practical tasks of machinevision such as navigation of mobile robotic systems, medicine, reconstruction of dense and sparse threedimensionalmaps of the environment, video surveillance and visual inspection, visual simultaneouslocalization and mapping, etc. The urgency of the problem of camera calibration is defined by the presenceof many different methods of calibration and calibration templates. Each individual solution issuitable only for special conditions, e.g., lack of lighting, bad weather conditions, the presence of thirdpartyobjects blocking visibility, etc. In most cases, each calibration method uses a specific calibrationpattern. Camera calibration is usually associated with the use of special calibration templates. Theyallow to achieve the most accurate results due to a previously known geometric structure. Currently, theprocedure for camera calibrating of robotic systems is carried out in laboratory conditions using theclassic “chessboard” method. In addition, there are only a few alternative approaches that are in their infancy state both in Russia and abroad. On the other hand, research into camera calibration methodscontinues and new alternatives for camera calibration are emerging. One of the new directions is theuse of fiducial marker systems as a reference object. A variety of parameters such as the size of the calibrationtemplate, the dimension of the calibration data set, the distribution of distances from the camerato objects on the stage, etc. creates a vast area for experimental testing of optimal camera calibrationparameters. This paper presents a research of automatic camera calibration using fiducial marker systems(FMS), which are located on the surface of the robot. Based on the results of virtual experimentswith FMS in the Gazebo simulation environment of the robotic operating system ROS, two differenttypes of FMS were selected that are optimal relative to other types of FMS covered by our previousstudies in terms of the resistance of FMS to systematic occlusion of the marker area and the effect ofmarker size on quality of its recognition. The selected FMS were tested using the onboard camera of theRussian mobile robot Servosila Engineer in indoor settings to assess the correlation of results in virtualand real environments.

3D in situ stress state modelling and fault reactivation risk exemplified in the Ruhr region (Germany)

Throughout the 700-yearlong coal exploration period in the Ruhr region, an abundance of geological, geophysical, seismic, and in situ stress data was obtained from the Carboniferous strata. In this study, we take advantage of this unique dataset to develop a static 3D geomechanical model to predict the spatially continuous distribution of undisturbed in situ stress state and evaluate the reactivation risk of major fault zones. Compared to the point-wise stress information, the spatially continuous in situ stress state provides an effective tool for planning subsurface operations and assessing seismic hazards in areas where no stress information is available. The developed model was validated against a comprehensive calibration dataset including results from geophysical logging, borehole deformation, and fault-slip analysis, in situ hydrofracturing measurements, distribution of subsidence, microseismicity, and observations from coal mining activities. Consequently, interpretation and assessment of the model results including their uncertainties, reliability, limitations, and perspectives are discussed. The possible applications of the model approach for seismic hazard prediction and utilization of deep geothermal energy in the Ruhr region are outlined.

National soil organic carbon map of agricultural lands in Nepal

Reliable and accurate soil organic carbon (SOC) maps are needed to monitor and improve SOC status in croplands and for agro-environmental applications. Topsoil (0–20 cm) SOC content from agricultural lands was predicted and mapped with quantified uncertainty across Nepal using state-of-the-art soil mapping techniques. Altogether 25,312 SOC observations were used to build and evaluate prediction models derived from four machine learning algorithms, namely Random Forest (RF), Cubist, Extreme Gradient Boosting (XGB) and Support Vector Machines. Twenty two environmental variables were selected as SOC predictors based on their correlation with measured SOC contents and non-collinearity with other predictors. The predictive performance of these models was compared using calibration (80% observations) and validation (20% observations) datasets. The performance of the models was also compared against a global SOC dataset compiled by International Soil Reference and Information Centre (ISRIC). The best model among the four algorithms was used to map and quantify the spatial distribution of SOC contents, and the model uncertainty was assessed with the Quantile Regression Forests technique with standard deviation representing prediction uncertainty. The RF model performed the best among all tested models, closely followed by the Cubist, and then the XGB model. The predictive performance of all of these models was better than the global SOC prediction from ISRIC. This study provides a baseline map for the topsoil SOC contents from the croplands in Nepal, and also provides a reference for similar SOC mapping studies.

A Technique for Rapid Bacterial-Density Enumeration through Membrane Filtration and Differential Pressure Measurements.

In this article, we present a microfluidic technique for the rapid enumeration of bacterial density with a syringe filter to trap bacteria and the quantification of the bacterial density through pressure difference measurement across the membrane. First, we established the baseline differential pressure and hydraulic resistance for a filtration membrane by fully wetting the filter with DI water. Subsequently, when bacteria were infused and trapped at the pores of the membrane, the differential pressure and hydraulic resistance also increased. We characterized the infusion time required for the bacterial sample to achieve a normalized hydraulic resistance of 1.5. An equivalent electric-circuit model and calibration data sets from parametric studies were used to determine the general form of a calibration curve for the prediction of the bacterial density of a bacterial sample. As a proof of concept, we demonstrated through blind tests with Escherichia coli that the device is capable of determining the bacterial density of a sample ranging from 7.3 × 106 to 2.2 × 108 CFU/mL with mean and median accuracies of 87.21% and 91.33%, respectively. The sample-to-result time is 19 min for a sample with lower detection threshold, while for higher-bacterial-density samples the measurement time is further shortened to merely 8 min.

Soil organic carbon (SOC) prediction in Australian sugarcane fields using Vis–NIR spectroscopy with different model setting approaches

To maintain profitability in sugarcane areas of Australia, soil nutrients need to be applied to replace losses to biomass production. For example, nitrogen fertiliser requires consideration of soil organic carbon (SOC, %). However, determining SOC is time-consuming. An alternative is to use a visible–near infrared (Vis–NIR) spectroscopy library. Herein, a Vis–NIR library is developed to predict topsoil (0–0.3 m) SOC using partial least squares regression (PLSR) and machine learning (i.e., Cubist, random forest [RF] and support vector machine [SVM]) in four sugarcane districts (i.e., Mossman, Lannercost, Herbert, and Proserpine). Different approaches were compared (i.e., site-specific, site-independent, hold-out and spiking) with spike size also considered. In all comparisons, a consistent set of calibration and validation data were used. The calibration coefficient of determination (R2) was always strong (> 0.7), and generally better than the validation R2, regardless of the modelling approach, district, or spike size. For the validation, the Lin's concordance correlation coefficient (LCCC) showed PLSR (0.92, and 0.9) and Cubist (0.91 and 0.9) were close to perfect (> 0.9) for site-specific and site-independent, respectively. This was not the case for hold-out, with only strong R2 (0.71) and substantial agreement (0.80) in Herbert using Cubist and moderate overall using PLSR. Similar results were achieved in terms of the accuracy considering the ratio of performance to interquartile (RPIQ), whereby overall site-specific and site-independent approaches had excellent accuracy (> 2.5) with Cubist slightly more accurate than PLSR. Hold-out accuracy was generally very poor (< 1.4). Spiking the hold-out data sets produced mixed results with prediction R2, agreement and accuracy respectively best in Lannercost with 70 or more samples using PLSR (strong, substantial and excellent) and Hebert with 10 or more using Cubist (strong, near perfect and excellent), while in Mossman with 50 or more samples using SVM (very weak, poor and fair) and Proserpine with 30 or more samples using Cubist (weak, moderate and fair) the results were not as good. It can be concluded that either site-specific or site-independent approach to calibration and prediction using either PLSR or Cubist was best, with the use of the latter approach being more efficient and allowing for the potential to add to this spectral library when new samples from each area or new areas can be added.

Use of sedimentary algal pigment analyses to infer past lake-water total phosphorus concentrations

We tested the feasibility of using sedimentary algal pigment analyses by spectral deconvolution to infer past lake-water total phosphorus concentrations. We established equations that link lake-water nutrient concentrations and sediment pigment concentrations, using a combination of calibration in both space and time, with a training set of 31 Swedish lakes. The calibration dataset yielded a significant positive relationship between total carotenoid concentrations and lake-water total phosphorus concentrations. We also compared sediment-pigment-based nutrient inferences with time series of water column monitoring data to evaluate whether temporal changes in total phosphorus concentrations are well captured by analysis of sedimentary pigments. We found that changes in pigment preservation through time can alter the relationship between concentrations of lake-water nutrients and sedimentary pigments, thus limiting the reliability of historical ecological conditions inferred from pigments in the sediment. Our data suggested that ratios of Chlorophyll derivatives to total carotenoids (CD/TC ratio) and Chlorophyll a to Chlorophyll derivatives (CPI) can be used as proxies for pigment preservation. Using our approach, inferred temporal changes in water-column total phosphorus concentrations in lakes are promising, but require further development, specifically with respect to the influence of pigment degradation in both the water column and sediments, as well as the factors that control such degradation.

Less is more: dimensionality reduction as a general strategy for more precise luminescence thermometry

Thermal resolution (also referred to as temperature uncertainty) establishes the minimum discernible temperature change sensed by luminescent thermometers and is a key figure of merit to rank them. Much has been done to minimize its value via probe optimization and correction of readout artifacts, but little effort was put into a better exploitation of calibration datasets. In this context, this work aims at providing a new perspective on the definition of luminescence-based thermometric parameters using dimensionality reduction techniques that emerged in the last years. The application of linear (Principal Component Analysis) and non-linear (t-distributed Stochastic Neighbor Embedding) transformations to the calibration datasets obtained from rare-earth nanoparticles and semiconductor nanocrystals resulted in an improvement in thermal resolution compared to the more classical intensity-based and ratiometric approaches. This, in turn, enabled precise monitoring of temperature changes smaller than 0.1 °C. The methods here presented allow choosing superior thermometric parameters compared to the more classical ones, pushing the performance of luminescent thermometers close to the experimentally achievable limits.

Mantle geothermometry: experimental evaluation and recalibration of Fe–Mg geothermometers for garnet-clinopyroxene and garnet-orthopyroxene in peridotite, pyroxenite and eclogite systems

The reliability of eight Fe–Mg exchange geothermobarometers for garnet-bearing peridotites, pyroxenites and eclogites has been examined using a database comprised of more than 300 published peridotite, pyroxenite and eclogite experiments conducted from 10 to 70 kbar and 850 to > 1650^circ{rm C}. We have tested Fe–Mg exchange geothermometers suitable for a range of mantle lithologies, including websterite, harzburgite, wehrlite and eclogite. All geothermometers maintained an average difference in experimental and calculated temperature (T) left( {Delta {text{T}} = {text{T}}_{{exp - { }}} {text{T}}_{{{text{calc}}}} } right) of less than pm 50 °C with a standard deviation of Delta {text{T}} between pm 50 to 150 °C. Most geothermometers performed well across a narrow range in ln {text{Kd}}_{{{text{Fe}} - {text{Mg}}}}^{{{text{A}} - {text{B}}}} (where {text{Kd}}_{{{text{Fe}} - {text{Mg}}}}^{{{text{A}} - {text{B}}}} =frac{{left( {{text{Fe}}_{{text{A}}} times {text{Mg}}_{{text{B}}} } right)}}{{({text{Fe}}_{{text{B}}} times {text{Mg}}_{{text{A}}} )}}), however, systematic overestimation and underestimation of T were observed outside of the optimal range of {text{lnKd}}_{{{text{Fe}} - {text{Mg}}}}^{{{text{A}} - {text{B}}}}. Increases in experimental pressure (P) adversely affected several geothermometers, particularly those calibrated empirically using natural samples. All previously published calibrations of the garnet-clinopyroxene geothermometer were unable to reliably reproduce the experimental T for both peridotite and eclogite experimental compositions, which hinders their confident application to natural datasets. To improve the state of mantle geothermobarometry we have used our experimental database to recalibrate the (1) garnet-clinopyroxene Fe–Mg exchange geothermometer, and (2) garnet-orthopyroxene Fe–Mg exchange geothermometer. Each geothermometer has been recalibrated across an extended P, T, and compositional range. The inclusion of eclogitic experiments in the calibration for the garnet-clinopyroxene geothermometer permits application to both eclogitic and peridotitic/pyroxenitic assemblages equilibrated under a wide range of PT conditions in the upper mantle. Using multiple linear regression to solve for lnKd, we found the following expressions best reproduced the experimental T (℃) of our dataset: TFe-Mggrt-cpx(∘C)=3356.34-0.008×Pkbar+0.259×XCagrt+0.914×XMggrt+-0.159×Jdcpx+lnKdFe-Mggrt-cpx+1.265-273\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$${\ ext{T}}_{{{\ ext{Fe}} - {\ ext{Mg}}}}^{{{\ ext{grt}} - {\ ext{cpx}}}} { (}^{ \\circ } {\ ext{C)}} = { }\\frac{3356.34}{{\\left( {\\left( { - 0.008 \ imes {\ ext{P }}\\left( {{\ ext{kbar}}} \\right)} \\right) + \\left( {0.259 \ imes {\ ext{X}}_{{{\ ext{Ca}}}}^{{{\ ext{grt}}}} } \\right) + \\left( {0.914 \ imes {\ ext{X}}_{{{\ ext{Mg}}}}^{{{\ ext{grt}}}} } \\right) + \\left( { - 0.159 \ imes {\ ext{Jd}}^{{{\ ext{cpx}}}} } \\right) + \\left( {{\ ext{ln}}\\left( {{\ ext{ Kd}}_{{{\ ext{Fe}} - {\ ext{Mg}}}}^{{{\ ext{grt}} - {\ ext{cpx}}}} } \\right) + 1.265} \\right)} \\right)}} - 273{ }$$\\end{document}TFe-Mggrt-opx(∘C)=1851.85-0.007×Pkbar+-1.83×XCagrt+lnKdFe-Mggrt-cpx+1.08-273\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$${\ ext{T}}_{{{\ ext{Fe}} - {\ ext{Mg}}}}^{{{\ ext{grt}} - {\ ext{opx}}}} { (}^{ \\circ } {\ ext{C)}} = { }\\frac{1851.85}{{\\left( {\\left( { - 0.007 \ imes {\ ext{P }}\\left( {{\ ext{kbar}}} \\right)} \\right) + \\left( { - 1.83 \ imes {\ ext{X}}_{{{\ ext{Ca}}}}^{{{\ ext{grt}}}} } \\right) + \\left( {{\ ext{ln}}\\left( {{\ ext{ Kd}}_{{{\ ext{Fe}} - {\ ext{Mg}}}}^{{{\ ext{grt}} - {\ ext{cpx}}}} } \\right) + 1.08} \\right)} \\right)}} - 273$$\\end{document}. where, {text{X}}_{{{text{Ca}}}}^{{{text{grt}}}} =frac{{{text{Ca}}}}{{left( {{text{Ca}} + {text{Fe}} + {text{Mg}}} right)}}, {text{Kd}}_{{{text{Fe}} - {text{Mg}}}}^{{{text{grt}} - {text{opx}}}} =frac{{left( {{text{Fe}}_{{{text{grt}}}} times {text{Mg}}_{{{text{opx}}}} } right)}}{{({text{Fe}}_{{{text{opx}}}} times {text{Mg}}_{{{text{grt}}}} )}} , {text{X}}_{{{text{Mg}}}}^{{{text{grt}}}} =frac{{{text{Mg}}}}{{left( {{text{Ca}} + {text{Fe}} + {text{Mg}}} right)}}, {text{Jd}}^{{{text{cpx}}}} = {text{Na}} - {text{Cr}} - 2 times {text{Ti}}, {text{Kd}}_{{{text{Fe}} - {text{Mg}}}}^{{{text{grt}} - {text{cpx}}}} =frac{{left( {{text{Fe}}_{{{text{grt}}}} times {text{Mg}}_{{{text{cpx}}}} } right)}}{{({text{Fe}}_{{{text{cpx}}}} times {text{Mg}}_{{{text{grt}}}} )}}, with all elements calculated on the basis of 12 oxygen anions in garnet and 6 oxygen anions in clino- and orthopyroxene. Fe2+ = total Fe. Our updated calibrations resolve several issues with earlier calibrations, including a poor performance at elevated P and compositional limitations. An improvement in precision and accuracy has been demonstrated through application to the experimental calibration dataset, a second independent set of published experimental data, and to natural peridotites, pyroxenites and eclogites from on and off craton settings. Iterative PT estimates on natural datasets calculated using our updated calibrations compare well with estimates from widely used calibrations such as the Taylor (1998) two-pyroxene solvus geothermometer. We anticipate that this contribution will provide an important reference for the reliability of mantle geothermometers and that our updated calibrations will be used in future studies on peridotite, pyroxenite and eclogite inclusions in diamond and mantle-derived xenoliths.

Ecology of testate amoebae along an environmental gradient from bogs to calcareous fens in East-Central Europe: development of transfer functions for palaeoenvironmental reconstructions

Testate amoebae play an important role in biomonitoring and the understanding of peatland (palaeo)ecology. However, their application has been mainly limited to Sphagnum-dominated peatlands, especially ombrotrophic bogs. To facilitate wider use of these microorganisms, we explored their ecology along a gradient from mineral-poor acidic bogs to mineral-rich calcareous fens with tufa formation using a new calibration dataset of over 250 samples from East-Central Europe. Specifically, we examined environmental controls on testate amoebae along the complete gradient and separately in the minerotrophic (pH >5.5) and ombrotrophic (pH <5) parts. Based on these findings, we developed and statistically evaluated transfer function models for the dominant factors controlling the community composition of these protists. Multivariate statistical analysis showed that groundwater pH was the main driver of species composition in the full and minerotrophic datasets. In contrast, testate amoeba communities in the ombrotrophic dataset were primarily structured by depth to water table (DWT). Under leave-one-out cross-validation (LOO), the best-performing model for DWT gave a prediction error (RMSEP) of 5.43 cm. The most robust models for groundwater mineral richness yielded RMSEPLOO of 0.45 and 0.27 pH units, for the full and the minerotrophic dataset respectively. This is the first study to present a complete set of testate amoeba-based transfer functions that allow reconstructions of Holocene hydrochemical and hydrological (hydroclimatic) variability in different mire types within one region. The new models open up the possibility for testate amoebae to become a valuable tool in palaeoecological research, although further work is needed to fully assess their performance and usefulness in practice.

Evaluation of the Analysis of Record for Calibration (AORC) Rainfall across Louisiana

The use of a long-term and high-quality precipitation dataset is crucial for hydrologic modeling and flood risk management. This study evaluates the Analysis of Period of Record for Calibration (AORC) dataset, a newly released product with high temporal and spatial resolutions. Our study region is centered on Louisiana because of the major flooding it has been experiencing. We compare the AORC hourly precipitation to other widely used gridded rainfall products and rain-gauge observations. To evaluate the performance of rainfall products according to different weather conditions causing severe flooding, we stratify the analyses depending on whether precipitation is associated with a tropical cyclone (TC) or not. Compared to observations, our results show that the AORC has the highest correlation coefficients (i.e., values above 0.75) with respect to observations among all rainfall products for both TC and non-TC periods. When the skill metric is decomposed into the potential skill and biases, the AORC clearly shows the highest potential skill with relatively small biases for the whole period. In addition, the AORC performs better for the TC period compared to the non-TC ones. Our results suggest that AORC precipitation shows good potential to be viable for hydrologic modeling and simulations of TC and non-TC events.

A bimodal extension of the ARYA&amp;PARIS approach for predicting hydraulic properties of structured soils

• Bimodal Arya&Paris hydraulic properties were developed based on soil structure. • The bimodal PTF provides better prediction of un/saturated hydraulic conductivity. • The bimodal PTF predicts well the spatial variability of the hydraulic properties. The main purpose of this paper is to develop a bimodal pedotransfer function to obtain soil water retention (WRC) and hydraulic conductivity (HCC) curves. The proposed pedo-transfer function (PTF) extends the Arya and Paris (AP) approach, which is based on particle size distribution (PSD), by incorporating aggregate-size distribution (ASD) into the PTF to obtain the bimodal WRC. A bimodal porosity approach was developed to quantify the fraction of each of the porous systems (matrix and macropores) in overall soil porosity. Saturated hydraulic conductivity, K 0 , was obtained from WRC using the Kozeny-Carman equation, whose parameters were inferred from the behaviour of the bimodal WRC close to saturation. Finally, the Mualem model was applied to obtain the HCC. In order to calibrate the PTF, measured soil physical and hydraulic properties data were used, coming from field infiltration experiments from an irrigation sector of 140 ha area in the “Sinistra Ofanto” irrigation system in Apulia, southern Italy. The infiltration data were fitted by using both bimodal and unimodal hydraulic properties by an inverse solution of the Richards equation. The bimodal “measured” hydraulic properties were then used to calibrate the scaling parameter (α AP ) of the proposed bimodal AP ( bimAP ) PTF. Similarly, for the sake of comparison with the bimodal results, the unimodal hydraulic properties were used to calibrate the α AP of the classical unimodal AP ( unimAP ) PTF. Compared to the unimAP PTF, the proposed bimAP significantly improves the predictions of the mean WRC parameters and K 0 , as well as the prediction of the shape of the whole HCC. Moreover, compared to the unimodal approach, it also allows to reproduce statistics of the hydraulic parameters (for example, variance) similar to those observed in the calibration dataset. Multiple linear regression (MLR) was also applied to analyse the sensitivity of the bimodal α AP parameter to textural and structural features, confirming significant predictive effects of soil structure.

Identifying low risk and profitable crop management practices for irrigated Teff production in northwestern Ethiopia

Teff (Eragrostis tef (Zucc.) Trotter) is one of the most important staple crops in Ethiopia. However, the optimal agronomic practices of the crop under irrigation remain unclear. The objectives of this study were to evaluate the Cropping System Model (CSM)-NWheat-Teff and to determine the optimum planting date, nitrogen (N) application rate, and irrigation threshold in northwestern Ethiopia. The model was calibrated and evaluated using published data from field experiments conducted from 1996 to 1998 in the Adet and Bichena districts and from 2001 to 2003 in the Dangila district. The model was then used to simulate different planting dates (Jan 1, Jan 15, Feb 1, Feb 15, Mar 1), N rates (0, 20, 40, 60, 80 kg ha−1), and irrigations (irrigation when required/automated, 50% and 25% available soil moisture thresholds) levels for Adet, Bichena and Dangila using historical weather data (1983–2020). The model was able to simulate teff phenology and yield adequately as indicated by low root mean square error (RMSE) and a high index of agreement (d) for both the calibration and evaluation datasets. Grain and straw yield increased with increasing N, but the rate of increment was higher under irrigation when required (automated) compared to 50% and 25% thresholds. Late planting (March 1) at Adet and early planting (January 1 and February 15) at Bichena and Dangila resulted in the highest profit with a minimum variability in output. The combined use of 80 kg N ha−1 and irrigation when required gave the most profitable option with low risk of irrigated teff production at the three sites studied. This study showed that the new CSM-NWheat-Teff model can be used to optimize agronomic practices for irrigated teff production systems in Ethiopia. Further research is needed to improve model performance in predicting teff yield for contrasting biophysical and socio-economical environments.

Reducing Energy Consumption of Pressure Sensor Calibration Using Polynomial HyperNetworks with Fourier Features

Our research aims to reduce the cost of pressure sensor calibration through machine learning. Pressure sensor calibration is a standard process whereby freshly manufactured pressure sensors are subjected to various controlled temperature and pressure setpoints to compute a mapping between the sensor's output and true pressure. Traditionally this mapping is calculated by fitting a polynomial with calibration data. Obtaining this data is costly since a large spectrum of temperature and pressure setpoints are required to model the sensor's behavior. We present a machine learning approach to predict a pre-defined calibration polynomial's parameters while requiring only one-third of the calibration data. Our method learns a pattern from past calibration sessions to predict the calibration polynomial's parameters from partial calibration setpoints for any newly manufactured sensor. We design a novel polynomial hypernetwork coupled with Fourier features and a weighted loss to solve this problem. We perform extensive evaluations and show that the current industry-standard method fails under similar conditions. In contrast, our approach saves two-thirds of the calibration time and cost. Furthermore, we conduct comprehensive ablations to study the effect of Fourier mapping and weighted loss. Code and a novel calibration dataset validated by calibration engineers are also made public.

Monitoring Leaf Nitrogen Accumulation With Optimized Spectral Index in Winter Wheat Under Different Irrigation Regimes.

Rapid and non-destructive estimation of leaf nitrogen accumulation (LNA) is essential to field nitrogen management. Currently, many vegetation indices have been used for indicating nitrogen status. Few studies systematically analyzed the performance of vegetation indices of winter wheat in estimating LNA under different irrigation regimes. This study aimed to develop a new spectral index for LNA estimation. In this study, 2 years of field experiments with different irrigation regimes were conducted from 2015 to 2017. The original reflectance (OR) and three transformed spectra [e.g., the first derivative reflectance (FDR), logarithm of the reciprocal of the spectra (Log(1/R)), and continuum removal (CR)] were used to calculate two- and three-band spectral indices. Correlation analyses and univariate linear and non-linear regression between transformed-based spectral indices and LNA were performed. The performance of the optimal spectral index was evaluated with classical vegetation index. The results showed that FDR was the most stable transformation method, which can effectively enhance the relationships to LNA and improve prediction performance. With a linear relationship with LNA, FDR-based three-band spectral index 1 (FDR-TBI1) (451, 706, 688) generated the best performance with coefficient of determination (R2) of 0.73 and 0.79, the root mean square error (RMSE) of 1.267 and 1.266 g/m2, and the ratio of performance to interquartile distance (RPIQ) of 2.84 and 2.71 in calibration and validation datasets, respectively. The optimized spectral index [FDR-TBI1 (451, 706, 688)] is more effective and might be recommended as an indicator for estimating winter wheat LNA under different irrigation regimes.

Coupled Coccolith-Based Temperature and Productivity High-Resolution Reconstructions in the Eastern Equatorial Pacific During the Last Deglaciation and the Holocene

We present a new high-resolution reconstruction of annual sea-surface temperatures (SSTa) and net primary productivity (NPP) using novel coccolithophore-based models developed for the Eastern Equatorial Pacific (EEP). We combined published coccolithophore census counts from core-tops in the Eastern Pacific with 32 new samples from the Equatorial region, to derive a new statistical model to reconstruct SSTa. Results show that the addition of the new EEP samples improves existing coccolithophore-based SST-calibrations, and allow reconstructing SSTa in the EEP with higher confidence. We also merged the relative abundance of deep-photic species Florisphaera profunda in the same surface sediment samples with existing calibration datasets for tropical regions, to reconstruct annual NPP. Both temperature and productivity calibrations were successfully applied to fossil coccolith data from Ocean Drilling Project Site 1240, in the EEP. The coccolith-based SSTa estimates show a cooling during the Last Glacial Maximum (LGM) and the Younger Dryas, and warming at the start of the Holocene. This pattern differs in the timing and magnitude of the temperature changes from other available SST-reconstructions based on biogeochemical and faunal proxies. We discuss these discrepancies to be the result of different proxy sensitivities to insolation forcing, seasonal bias, and/or preservation artifacts. Reconstructed annual NPP shows a general decreasing trend from the late last glacial period to recent times, which we relate to the weakening of wind-driven equatorial upwelling towards the Holocene. We also calculated carbon export using our SSTa and NPP reconstructions, and compared to other geochemical-based reconstructions for the same location. Our coupled SSTa-NPP reconstruction provides key data to more fully assess the evolution of primary and export productivity as well as organic carbon burial in the EEP, with implications for its role in global biogeochemical cycles across glacial terminations.

On the use of transfer modeling to design new steels with excellent rotating bending fatigue resistance even in the case of very small calibration datasets

In this research a machine learning model for predicting the rotating bending fatigue strength and the high-throughput design of fatigue resistant steels is proposed. In this transfer prediction framework, machine learning models are first trained to estimate tensile properties (yield strength, tensile strength and elongation) on the basis of composition and critical process conditions. Then, based on the predicted tensile properties, transfer models are trained to estimate fatigue strength. The results are compared with those of a similar model not having such a transfer layer. The transfer prediction framework shows high accuracy for fatigue strength prediction with a remarkably high tolerance to limitations in the amount of calibration data available for training. By combining the transfer prediction framework with evolutionary algorithms, a robust high-throughput alloy design model is achieved requiring only tens of fatigue data points to get a decent reliability. The newly designed steel showed the predicted high fatigue strength. The method as presented here might also be applicable to other alloy design challenges in which only a limited database for the property to be optimized is available.

Disease Localization and Severity Assessment in Chest X-Ray Images using Multi-Stage Superpixels Classification

Background and ObjectivesChest X-ray (CXR) is a non-invasive imaging modality used in the prognosis and management of chronic lung disorders like tuberculosis (TB), pneumonia, coronavirus disease (COVID-19), etc. The radiomic features associated with different disease manifestations assist in detection, localization, and grading the severity of infected lung regions. The majority of the existing computer-aided diagnosis (CAD) system used these features for the classification task, and only a few works have been dedicated to disease-localization and severity scoring. Moreover, the existing deep learning approaches use class activation map and Saliency map, which generate a rough localization. This study aims to generate a compact disease boundary, infection map, and grade the infection severity using proposed multistage superpixel classification-based disease localization and severity assessment framework. Methods: The proposed method uses a simple linear iterative clustering (SLIC) technique to subdivide the lung field into small superpixels. Initially, the different radiomic texture and proposed shape features are extracted and combined to train different benchmark classifiers in a multistage framework. Subsequently, the predicted class labels are used to generate an infection map, mark disease boundary, and grade the infection severity. The performance is evaluated using a publicly available Montgomery dataset and validated using Friedman average ranking and Holm and Nemenyi post-hoc procedures. Results: The proposed multistage classification approach achieved accuracy (ACC)= 95.52%, F-Measure (FM)= 95.48%, area under the curve (AUC)= 0.955 for Stage-I and ACC=85.35%, FM=85.20%, AUC=0.853 for Stage-II using calibration dataset and ACC = 93.41%, FM = 95.32%, AUC = 0.936 for Stage-I and ACC = 84.02%, FM = 71.01%, AUC = 0.795 for Stage-II using validation dataset. Also, the model has demonstrated the average Jaccard Index (JI) of 0.82 and Pearson's correlation coefficient (r) of 0.9589. Conclusions: The obtained classification results using calibration and validation dataset confirms the promising performance of the proposed framework. Also, the average JI shows promising potential to localize the disease, and better agreement between radiologist score and predicted severity score (r) confirms the robustness of the method. Finally, the statistical test justified the significance of the obtained results.

A Low-Cost Wireless Bite Force Measurement Device.

Assessing maximum voluntary bite force is important to characterize the functional state of the masticatory system. Due to several factors affecting the estimation of the maximum bite force, a unique solution combining desirable features such as reliability, accuracy, precision, usability, and comfort is not available. The aim of the present study was to develop a low-cost bite force measurement device allowing for subject-specific customization, comfortable bite force expression, and reliable force estimation over time. The device was realized using an inexpensive load cell, two 3D printed ergonomic forks hosting reusable subject-specific silicone molds, a read-out system based on a low-cost microcontroller, and a wireless link to a personal computer. A simple model was used to estimate bite force taking into account individual morphology and device placement in the mouth. Measurement reliability, accuracy, and precision were assessed on a calibration dataset. A validation procedure on healthy participants was performed to assess the repeatability of the measurements over multiple repetitions and sessions. A 2% precision and 2% accuracy were achieved on measurements of forces in the physiological range of adult bite forces. Multiple recordings on healthy participants demonstrated good repeatability (coefficient of variation 11%) with no significant effect of repetition and session. The novel device provides an affordable and reliable solution for assessing maximum bite force that can be easily used to perform clinical evaluations in single sessions or in longitudinal studies.

Estimation of canopy nitrogen content in winter wheat from Sentinel-2 images for operational agricultural monitoring

Canopy nitrogen content (CNC, kg/ha) provides crucial information for site-specific crop fertilization and the usability of Sentinel-2 (S2) satellite data for CNC monitoring at high fertilization levels in managed agricultural fields is still underexplored. Winter wheat samples were collected in France and Belgium in 2017 (n = 126) and 2018 (n = 18), analysed for CNC and S2-spectra were extracted at the sample locations. A comparison of three established remote sensing methods to retrieve CNC was carried out: (1) look-up-table (LUT) inversion of the canopy reflectance model PROSAIL, (2) Partial Least Square Regression (PLSR) and (3) nitrogen-sensitive vegetation indices (VI). The spatial and temporal model transferability to new data was rigorously assessed. The PROSAIL-LUT approach predicted CNC with a root mean squared error of 33.9 kg/ha on the 2017 dataset and a slightly larger value of 36.8 kg/ha on the 2018 dataset. Contrary, PLSR showed an error of 27.9 kg N/ha (R2 = 0.52) in the calibration dataset (2017) but a substantially larger error of 38.4 kg N/ha on the independent dataset (2018). VIs revealed calibration errors were slightly larger than the PLSR results but showed much higher validation errors for the independent dataset (> 50 kg/ha). The PROSAIL inversion was more stable and robust than the PLSR and VI methods when applied to new data. The obtained CNC maps may support farmers in adapting their fertilization management according to the actual crop nitrogen status.