Uncertainty In Model Predictions Research Articles

Data-Driven Process Optimization Considering Surrogate Model Prediction Uncertainty: A Mixture Density Network-Based Approach

The artificial neural network (ANN) can be effectively used as a data-driven surrogate model in process optimization. However, there is a problem that the change of training set leads to prediction...

Bayesian parameter and joint probability distribution estimation for a hysteretic constitutive model of reinforcing steel

With the structural design paradigm shift since the early 2000′s from the traditional approach to performance-based design (PBD), there has been a growing need for reliable nonlinear finite element (FE) models that can accurately predict the response of structures when subjected to extreme loads, such as earthquakes. In the case of reinforced concrete (RC) structures, a proper representation of the hysteretic nonlinear behavior of reinforcing steel becomes crucial in order to carry out nonlinear time history analyses. The Giuffrè-Menegotto-Pinto (GMP) uniaxial steel constitutive law has been widely used by researchers and practitioners to model reinforcing steel bars. Despite the widespread in its implementation, a limited number of studies have proposed well-calibrated parameter values for this model. In addition, low identifiability of its governing parameters and the high cost of generating reliable experimental data have prevented a thorough probabilistic characterization of the GMP model parameters. Usually, only default parameter values from the early development of the model tend to be used. This paper uses experimental data from cyclic tests conducted on 36 reinforcing steel coupons manufactured in accordance to ASTM A615 and A706 Grade 60 reinforcing steel and proposes a joint probability density function (PDF) for the most influential parameters of the GMP material model. First, a local sensitivity analysis (LSA) is conducted to provide insight into the influence of each parameter in the model response. Also, global sensitivity analysis (GSA) is used to have a deep understanding of the composition of the variability in the model response due to parameter uncertainty and the level of interactions among parameters. The Bayesian approach is combined with the information obtained from GSA and LSA as input, to estimate model parameters and quantify the estimation uncertainties and propagate them to the material stress response. Uncertainty in model predictions obtained with the proposed PDF is assessed, and the impact of considering parameter correlations on the material response is investigated.

Using different size fractions to source fingerprint fine-grained channel bed sediment in a large drainage basin in Iran

Sediment fingerprinting in data sparse regions, such as the mountainous areas of Iran, is more suited to a confluence-based sample design wherein tributary sub-basins are characterised by sediment samples using different size fractions. Our objective was therefore to fingerprint spatial source contributions to the < 37 and 37–63 µm fractions of fine-grained channel bed sediment samples collected in a large erodible mountainous river basin in Iran based on a number of statistical and machine learning approaches. Geochemical elements were measured in channel bed surface drape sediments from seven sub-basins and in delivered sediments from the basin outlet. A Bayesian mass balance model (modified MixSIR) was applied to apportion sub-basin spatial sources with four composite signatures selected using the different statistical approaches. For the < 37 and 37–63 µm fractions, the signatures all indicated that sub-basins 3 (Andajroud; 46.7%) and 7 (Ninehroud; 36.8%) were the dominant spatial sources of the fine-grained bed drape sediment samples, respectively, identifying the most active erosional zones spatially. The statistical error between known and predicted spatial source contributions using virtual mixture tests for both size fractions demonstrated the importance of using multiple different composite fingerprints to decrease the model prediction uncertainties. Despite the difficult terrain in such data sparse areas, the source fingerprinting approach provides a basis for assembling new evidence which, in turn, is of interest to scientists and managers alike.

Roads, forest cover, and topography as factors affecting the occurrence of large carnivores: The case of the Andean bear (Tremarctos ornatus)

Global declines in large-bodied terrestrial vertebrates are widely linked to human disturbance and habitat loss. Identifying remaining opportunities to conserve areas wherein such species may persist or might be reintroduced has thus become a key challenge in biodiversity conservation globally. We used multi-year camera trap data and spatial capture-recapture models to test how individually-identified Andean bears (Tremarctos ornatus) responded to roads and a suite of forest cover and topographic variables that may be indicative of food resources in critically-threatened, equatorial dry forest habitats of northern Peru. Bear density averaged 4 individuals per 100 km2 in our 668 km2 study area, but proximity to roads significantly reduced density, possibly because they are avoided and/or facilitate demographic sinks or human hunting. In contrast, the effects of forest cover and topographic variables on bear density varied seasonally, reflecting documented variation in mating behavior that may result from seasonal food resources. Despite large uncertainties in model predictions on the occurrence of bear habitat outside our study area, extrapolating our results suggests that a substantial area of high-quality habitat for Andean bears exists outside the recognized IUCN range, including equatorial dry forest habitats known to harbor many threatened species. Because Andean bears are vulnerable to extinction and are an established flagship species in much of South America, the expanded range and potential habitats that we identify here require urgent validation for possible inclusion in landscape-level plans to conserve equatorial dry forests of the Andes.

Exploring the Cloud Top Phase Partitioning in Different Cloud Types Using Active and Passive Satellite Sensors

AbstractOne of the largest uncertainties in numerical weather prediction and climate models is the representation of mixed‐phase clouds. With the aim of understanding how the supercooled liquid fraction (SLF) in clouds with temperature from −40°C to 0°C is related to temperature, geographical location, and cloud type, our analysis contains a comparison of four satellite‐based datasets (one derived from active and three from passive satellite sensors), and focuses on SLF distribution near‐globally, but also stratified by latitude and continental/maritime regions. Despite the warm bias in cloud top temperature of the passive sensor compared to the active sensor and the phase mismatch in collocated data, all datasets indicate, at the same height‐level, an increase of SLF with cloud optical thickness, and generally larger SLF in the Southern Hemisphere than in the Northern Hemisphere (up to about 20% difference), with the exception of continental low‐level clouds, for which the opposite is true.

Temporal, Spatial, and Temperature Controls on Organic Carbon Mineralization and Methanogenesis in Arctic High-Centered Polygon Soils.

Warming temperatures in continuous permafrost zones of the Arctic will alter both hydrological and geochemical soil conditions, which are strongly linked with heterotrophic microbial carbon (C) cycling. Heterogeneous permafrost landscapes are often dominated by polygonal features formed by expanding ice wedges: water accumulates in low centered polygons (LCPs), and water drains outward to surrounding troughs in high centered polygons (HCPs). These geospatial differences in hydrology cause gradients in biogeochemistry, soil C storage potential, and thermal properties. Presently, data quantifying carbon dioxide (CO2) and methane (CH4) release from HCP soils are needed to support modeling and evaluation of warming-induced CO2 and CH4 fluxes from tundra soils. This study quantifies the distribution of microbial CO2 and CH4 release in HCPs over a range of temperatures and draws comparisons to previous LCP studies. Arctic tundra soils were initially characterized for geochemical and hydraulic properties. Laboratory incubations at −2, +4, and +8°C were used to quantify temporal trends in CO2 and CH4 production from homogenized active layer organic and mineral soils in HCP centers and troughs, and methanogen abundance was estimated from mcrA gene measurements. Results showed that soil water availability, organic C, and redox conditions influence temporal dynamics and magnitude of gas production from HCP active layer soils during warming. At early incubation times (2–9 days), higher CO2 emissions were observed from HCP trough soils than from HCP center soils, but increased CO2 production occurred in center soils at later times (>20 days). HCP center soils did not support methanogenesis, but CH4-producing trough soils did indicate methanogen presence. Consistent with previous LCP studies, HCP organic soils showed increased CO2 and CH4 production with elevated water content, but HCP trough mineral soils produced more CH4 than LCP mineral soils. HCP mineral soils also released substantial CO2 but did not show a strong trend in CO2 and CH4 release with water content. Knowledge of temporal and spatial variability in microbial C mineralization rates of Arctic soils in response to warming are key to constraining uncertainties in predictive climate models.

Revision of the Stochastically Perturbed Parametrisations model uncertainty scheme in the Integrated Forecasting System

AbstractThe Stochastically Perturbed Parametrisations scheme (SPP) represents model uncertainty in numerical weather prediction by introducing stochastic perturbations into the physical parametrisation schemes. The perturbations are constructed in such a way that the internal consistency of the physical parametrisation schemes is preserved. We developed a revised version of SPP for the Integrated Forecasting System of the European Centre for Medium‐Range Weather Forecasts (ECMWF). The revised version introduces perturbations to additional quantities and modifies the probability distributions sampled by the scheme. Medium‐range ensemble forecasts with the revised SPP are considerably more skilful than ensemble forecasts with the original implementation of SPP. The revised version of SPP is similar, in terms of forecast skill, to the Stochastically Perturbed Parametrisation Tendency scheme (SPPT), which is currently used to represent model uncertainty in the operational ECMWF ensemble forecasts.

Information-Theoretic Approach for Model Reduction Over Finite Time Horizon

This paper presents an information-theoretic approach for model reduction for finite time simulation. Although system models are typically used for simulation over a finite time, most of the metrics (and pseudo-metrics) used for model accuracy assessment consider asymptotic behavior e.g., Hankel singular values and Kullback-Leibler(KL) rate metric. These metrics could further be used for model order reduction. Hence, in this paper, we propose a generalization of KL divergence-based metric called n-step KL rate metric, which could be used to compare models over a finite time horizon. We then demonstrate that the asymptotic metrics for comparing dynamical systems may not accurately assess the model prediction uncertainties over a finite time horizon. Motivated by this finite time analysis, we propose a new pragmatic approach to compute the influence of a subset of states on a combination of states called information transfer (IT). Model reduction typically involves the removal or truncation of states. IT combines the concepts from the n-step KL rate metric and model reduction. Finally, we demonstrate the application of information transfer for model reduction. Although the analysis and definitions presented in this paper assume linear systems, they can be extended for nonlinear systems.

ASSESSING UNCERTAINTY IN THE MODELING OF RUNUP AND SWASH MORPHODYNAMICS USING XBEACH

Phase-resolving numerical models are frequently used tools to investigate short and long wave transformation, nonlinear wave interactions, and wave runup. Moreover, nearshore morphodynamics can be explored with the recent advancement of the models and computational resources. Sea surface elevation time series that force phase-resolving models at the offshore boundary are often unavailable. Therefore, time series are usually recreated from wave energy-frequency spectra through the superposition of harmonics. The wave phases of the harmonics are unknown and therefore assumed to be randomly distributed. This implies that an infinite number of time series with different sequencing of waves can be recreated from a single wave-energy spectrum and, for that reason, recreated time series are a source of uncertainty in model predictions. This intrinsic uncertainty has been found to cause variability in wave overtopping of structures (e.g., Pearson et al, 2002; Williams et al., 2014; Romano et al., 2015) and in setup and runup at beaches (McCabe et al., 2011; Torres-Freyermuth et al., 2019). Torres-Freyermuth et al. (2019) investigated the effect of intrinsic uncertainty on runup at planar beaches for different wave conditions and beach slopes and suggested that uncertainty is especially important under dissipative conditions. Yet unknown is the effect of intrinsic uncertainty on bed evolution. Here we assess the effect of intrinsic uncertainty on inner surf and swash zone evolution at three beaches with different beach morphology.Recorded Presentation from the vICCE (YouTube Link): https://youtu.be/w3zi3Yoo170

Benchmarking empirical models for THMs formation in drinking water systems: An application for decision support in Barcelona, Spain

Drinking water treatment plants (DWTPs) face changes in raw water quality, which affect the formation of disinfection by-products. Several empirical modelling approaches have been reported in the literature, but most of them have been developed with lab-scale data, which may not be representative of real water systems. Therefore, the application of these models for real-time operation of DWTPs might be limited. At the present study, multiple linear regression (MLR) and multi-layer perceptrons (MLP) were benchmarked using field-scale data for predicting the THMs formation in a case-study DWTP in Barcelona, Spain. After fitting the studied models, MLR exhibited good fit with the validation data set (R2 = 0.88 and MAE = 4.0 μg·L−1) and described the most plausible input-output relationships with field-scale data. The MLR predictive model was incorporated into an environmental decision support system (EDSS) for assessing the THMs formation at two critical points of the distribution network. A Monte Carlo scheme was applied for quantifying uncertainty of model predictions at these points, considering low and high water quality scenarios and different degrees of treatment by an electrodialysis reversal process. The results show that the use of the proposed EDSS can help in real operation of complex drinking water systems, which face important changes in water quality throughout the year.

A Review of Tank Model and Its Applicability to Various Korean Catchment Conditions

This paper reviews a conceptual rainfall-runoff model called Tank which has been widely used over the last 20 years in Korea as a part of a water resource modelling framework for assessing and developing long-term water resource polices. In order to examine the uncertainty of model predictions and the sensitivity of model’s parameters, Monte Carlos and Markov chain-based approaches are applied to five catchments of various Korean geographical and climatic conditions where the catchment sizes are ranged from 83 to 4786 km2. In addition, three optimization algorithms—dynamically dimensioned search (DDS), robust parameter estimation (ROPE), and shuffled complex evolution (SCE)—are selected to test whether the model parameters can be optimized consistently within a narrower range than the uncertainty bounds. From the uncertainty analysis, it is found that there is limited success in refining the priori distributions of the model parameters, indicating there is a high degree of equifinality for some parameters or at least there are large numbers of parameter combinations leading to good solutions within model’s uncertainty bounds. Out of the three optimization algorithms, SCE meets the criteria of the consistency best. It is also found that there are still some parameters that even the SCE method struggles to refine the priori distributions. It means that their contribution to model results is minimal and can take a value within a reasonable range. It suggests that the model may be reconceptualized to be parsimonious and to rationalize some parameters without affecting model’s capacity to replicate historical flow characteristics. Cross-validation indicates that sensitive parameters to catchment characteristics can be transferred when geophysical similarity exists between two catchments. Regionalization can be further improved by using a regression or geophysical similarity-based approach to transfer model parameters to ungauged catchments. It may be beneficial to categorize the model parameters depending on the level of their sensitivities, and a different approach to each category may be applied to regionalize the calibrated parameters.

Variance-based sensitivity analysis for uncertainties in proton therapy: A framework to assess the effect of simultaneous uncertainties in range, positioning, and RBE model predictions on RBE-weighted dose distributions.

Treatment plans in proton therapy are more sensitive to uncertainties than in conventional photon therapy. In addition to setup uncertainties, proton therapy is affected by uncertainties in proton range and relative biological effectiveness (RBE). While to date a constant RBE of 1.1 is commonly assumed, the actual RBE is known to increase toward the distal end of the spread-out Bragg peak. Several models for variable RBE predictions exist. We present a framework to evaluate the combined impact and interactions of setup, range, and RBE uncertainties in a comprehensive, variance-based sensitivity analysis (SA). The variance-based SA requires a large number (104 -105 ) of RBE-weighted dose (RWD) calculations. Based on a particle therapy extension of the research treatment planning system CERR we implemented a fast, graphics processing unit (GPU) accelerated pencil beam modeling of patient and range shifts. For RBE predictions, two biological models were included: The mechanistic repair-misrepair-fixation (RMF) model and the phenomenological Wedenberg model. All input parameters (patient position, proton range, RBE model parameters) are sampled simultaneously within their assumed probability distributions. Statistical formalisms rank the input parameters according to their influence on the overall uncertainty of RBE-weighted dose-volume histogram (RW-DVH) quantiles and the RWD in every voxel, resulting in relative, normalized sensitivity indices (S=0: noninfluential input, S=1: only influential input). Results are visualized as RW-DVHs with error bars and sensitivity maps. The approach is demonstrated for two representative brain tumor cases and a prostate case. The full SA including RWD calculations took 39, 11, and 55min, respectively. Range uncertainty was an important contribution to overall uncertainty at the distal end of the target, while the relatively smaller uncertainty inside the target was governed by biological uncertainties. Consequently, the uncertainty of the RW-DVH quantile D98 for the target was governed by range uncertainty while the uncertainty of the mean target dose was dominated by the biological parameters. The SA framework is a powerful and flexible tool to evaluate uncertainty in RWD distributions and DVH quantiles, taking into account physical and RBE uncertainties and their interactions. The additional information might help to prioritize research efforts to reduce physical and RBE uncertainties and could also have implications for future approaches to biologically robust planning and optimization.

Climate Cloud Model Forecast Verification-an Engineering Perspective

The processes relevant to the verification of cloud forecasts generated by climate models are discussed from an engineering perspective. These processes include an assessment of cloud product requirements to be evaluated, the creation of a verification test plan including procedures and data to be analyzed, the development of independent sources of validation or truth datasets, and the quantitative comparisons between the cloud forecast products and the truth data needed to establish model performance. The engineering perspective means minimal effort is focused on assessing the veracity of the physics contained in the cloud forecast model, rather emphasis is upon evaluating the results produced by it. It is postulated that these procedures are critical to improve the reliability of climate model predictions. The World Meteorological Organization has stated accuracy requirements for cloud products created from satellite observations, through the Global Climate Observing System (GSOC) program; however, no similar requirements have been defined for cloud forecast products. A statement of accuracy requirements is urgently needed. Meanwhile, it is assumed herein that cloud observation and cloud forecast requirements are identical. The assessment of model performance exploits high quality, manually-generated cloud truth products created from remotely-sensed satellite data which serve as truth data. Results show clouds under-specified in reanalysis cloud datasets created for use to initialize climate models but an over-specification of clouds by the cloud forecast model, in short-range predictions. This system level analysis demonstrates the need to improve the accuracy of cloud forecasts, especially lower-level water clouds which are responsible for most of the uncertainty in climate model predictions.

Data-Driven Uncertainty Quantification for Cardiac Electrophysiological Models: Impact of Physiological Variability on Action Potential and Spiral Wave Dynamics.

Computational modeling of cardiac electrophysiology (EP) has recently transitioned from a scientific research tool to clinical applications. To ensure reliability of clinical or regulatory decisions made using cardiac EP models, it is vital to evaluate the uncertainty in model predictions. Model predictions are uncertain because there is typically substantial uncertainty in model input parameters, due to measurement error or natural variability. While there has been much recent uncertainty quantification (UQ) research for cardiac EP models, all previous work has been limited by either: (i) considering uncertainty in only a subset of the full set of parameters; and/or (ii) assigning arbitrary variation to parameters (e.g., ±10 or 50% around mean value) rather than basing the parameter uncertainty on experimental data. In our recent work we overcame the first limitation by performing UQ and sensitivity analysis using a novel canine action potential model, allowing all parameters to be uncertain, but with arbitrary variation. Here, we address the second limitation by extending our previous work to use data-driven estimates of parameter uncertainty. Overall, we estimated uncertainty due to population variability in all parameters in five currents active during repolarization: inward potassium rectifier, transient outward potassium, L-type calcium, rapidly and slowly activating delayed potassium rectifier; 25 parameters in total (all model parameters except fast sodium current parameters). A variety of methods was used to estimate the variability in these parameters. We then propagated the uncertainties through the model to determine their impact on predictions of action potential shape, action potential duration (APD) prolongation due to drug block, and spiral wave dynamics. Parameter uncertainty had a significant effect on model predictions, especially L-type calcium current parameters. Correlation between physiological parameters was determined to play a role in physiological realism of action potentials. Surprisingly, even model outputs that were relative differences, specifically drug-induced APD prolongation, were heavily impacted by the underlying uncertainty. This is the first data-driven end-to-end UQ analysis in cardiac EP accounting for uncertainty in the vast majority of parameters, including first in tissue, and demonstrates how future UQ could be used to ensure model-based decisions are robust to all underlying parameter uncertainties.

A Bayesian end-to-end model with estimated uncertainties for simple question answering over knowledge bases

Existing methods for question answering over knowledge bases (KBQA) ignore the consideration of the model prediction uncertainties. We argue that estimating such uncertainties is crucial for the reliability and interpretability of KBQA systems. Therefore, we propose a novel end-to-end KBQA model based on Bayesian Neural Network (BNN) to estimate uncertainties arose from both model and data. To our best knowledge, we are the first to consider the uncertainty estimation problem for the KBQA task using BNN. The proposed end-to-end model integrates entity detection and relation prediction into a unified framework, and employs BNN to model entity and relation under the given question semantics, transforming network weights into distributions. Therefore, predictive distributions can be estimated by sampling weights and forward inputs through the network multiple times. Uncertainties can be further quantified by calculating the variances of predictive distributions. The experimental results demonstrate the effectiveness of uncertainties in both the misclassification detection task and cause of error detection task. Furthermore, the proposed model also achieves comparable performance compared to the existing state-of-the-art approaches on SimpleQuestions dataset.

Structure-based QSAR models to predict repeat dose toxicity points of departure

Human health risk assessment for environmental chemical exposure is limited by a vast majority of chemicals with little or no experimental in vivo toxicity data. Data gap filling techniques, such as quantitative structure–activity relationship (QSAR) models based on chemical structure information, can predict hazard in the absence of experimental data. Risk assessment requires identification of a quantitative point-of-departure (POD) value, the point on the dose-response curve that marks the beginning of a low-dose extrapolation. This study presents two sets of QSAR models to predict POD values (PODQSAR) for repeat dose toxicity. For training and validation, a publicly available in vivo toxicity dataset for 3592 chemicals was compiled using the U.S. Environmental Protection Agency’s Toxicity Value database (ToxValDB). The first set of QSAR models predict point-estimates of POD values (PODQSAR) using structural and physicochemical descriptors for repeat dose study types and species combinations. A random forest QSAR model using study type and species as descriptors showed the best performance, with an external test set root mean square error (RMSE) of 0.71 log10-mg/kg/day and coefficient of determination (R2) of 0.53. The second set of QSAR models predict the 95% confidence intervals for PODQSAR using a constructed POD distribution with a mean equal to the median POD value and a standard deviation of 0.5 log10-mg/kg/day, based on previously published typical study-to-study variability that may lead to uncertainty in model predictions. Bootstrap resampling of the pre-generated POD distribution was used to derive point-estimates and 95% confidence intervals for each POD prediction. Enrichment analysis to evaluate the accuracy of PODQSAR showed that 80% of the 5% most potent chemicals were found in the top 20% of the most potent chemical predictions, suggesting that the repeat dose POD QSAR models presented here may help inform screening level human health risk assessments in the absence of other data.

Bathymetric Inversion and Uncertainty Estimation from Synthetic Surf-Zone Imagery with Machine Learning

Resolving surf-zone bathymetry from high-resolution imagery typically involves measuring wave speeds and performing a physics-based inversion process using linear wave theory, or data assimilation techniques which combine multiple remotely sensed parameters with numerical models. In this work, we explored what types of coastal imagery can be best utilized in a 2-dimensional fully convolutional neural network to directly estimate nearshore bathymetry from optical expressions of wave kinematics. Specifically, we explored utilizing time-averaged images (timex) of the surf-zone, which can be used as a proxy for wave dissipation, as well as including a single-frame image input, which has visible patterns of wave refraction and instantaneous expressions of wave breaking. Our results show both types of imagery can be used to estimate nearshore bathymetry. However, the single-frame imagery provides more complete information across the domain, decreasing the error over the test set by approximately 10% relative to using timex imagery alone. A network incorporating both inputs had the best performance, with an overall root-mean-squared-error of 0.39 m. Activation maps demonstrate the additional information provided by the single-frame imagery in non-breaking wave areas which aid in prediction. Uncertainty in model predictions is explored through three techniques (Monte Carlo (MC) dropout, infer-transformation, and infer-noise) to provide additional actionable information about the spatial reliability of each bathymetric prediction.

A Model-Based Volume Estimator that Accounts for Both Land Cover Misclassification and Model Prediction Uncertainty

Forest/non-forest and forest species maps are often used by forest inventory programs in the forest estimation process. For example, some inventory programs establish field plots only on lands corresponding to the forest portion of a forest/non-forest map and use species-specific area estimates obtained from those maps to support the estimation of species-specific volume (V) totals. Despite the general use of these maps, the effects of their uncertainties are commonly ignored with the result that estimates might be unreliable. The goal of this study is to estimate the effects of the uncertainty of forest species maps used in the sampling and estimation processes. Random forest (RF) per-pixel predictions were used with model-based inference to estimate V per unit area for the six main forest species of La Rioja, Spain. RF models for predicting V were constructed using field plot information from the Spanish National Forest Inventory and airborne laser scanning data. To limit the prediction of V to pixels classified as one of the main forest species assessed, a forest species map was constructed using Landsat and auxiliary information. Bootstrapping techniques were implemented to estimate the total uncertainty of the V estimates and accommodated both the effects of uncertainty in the Landsat forest species map and the effects of plot-to-plot sampling variability on training data used to construct the RF V models. Standard errors of species-specific total V estimates increased from 2–9% to 3–22% when the effects of map uncertainty were incorporated into the uncertainty assessment. The workflow achieved satisfactory results and revealed that the effects of map uncertainty are not negligible, especially for open-grown and less frequently occurring forest species for which greater variability was evident in the mapping and estimation process. The effects of forest map uncertainty are greater for species-specific area estimation than for the selection of field plots used to calibrate the RF model. Additional research to generalize the conclusions beyond Mediterranean to other forest environments is recommended.

Full Parameterisation Matters for the Best Performance of Crop Models: Inter-comparison of a Simple and a Detailed Maize Model

Process-based crop growth models have become indispensable tools for investigating the effects of genetic, management, and environmental factors on crop productivity. One source of uncertainty in crop model predictions is model parameterization, i.e. estimating the values of model input parameters, which is carried out very differently by crop modellers. One simple (SSM-iCrop) and one detailed (APSIM) maize (Zea mays L.) model were partially or fully parameterized using observed data from a 2-year field experiment conducted in 2016 and 2017 at the UFT (Universitäts- und Forschungszentrum Tulln, BOKU) in Austria. Model initialisation was identical for both models based on field measurements. Partial parameterization (ParLevel_1) was first performed by estimating only those parameters related to crop phenology. Full parameterization (ParLevel_2) was then conducted by estimating parameters related to phenology plus those affecting dry mass production and partitioning, nitrogen uptake, and grain yield formation. With ParLevel_1, both models failed to provide accurate estimation of LAI, dry mass accumulation, nitrogen uptake and grain yield, but the performance of APSIM was generally better than SSM-iCrop. Full parameterization greatly improved the performance of both crop models, but it was more effective for the simple model, so that SSM-iCrop was equally well or even better compared to APSIM. It was concluded that full parameterization is indispensable for improving the accuracy of crop model predictions regardless whether they are simple or detailed. Simple models seem to be more vulnerable to incomplete parameterization, but they better respond to full parameterization. This needs confirmation by further research.

New physics at the MUonE experiment at CERN

A confirmation of the long-standing muon $g$-2 discrepancy requires both experimental and theoretical progress. On the theory side, the hadronic corrections are under close scrutiny, as they induce the leading uncertainty of the Standard Model prediction. Recently, the MUonE experiment has been proposed at CERN to provide a new determination of the leading hadronic contribution to the muon $g$-2 via the measurement of the differential cross section of muon-electron scattering. The precision expected at this experiment raises the question whether possible new-physics (NP) could affect its measurements. We address this issue studying possible NP signals in muon-electron collisions due to heavy or light mediators, depending on whether their mass is higher or lower than ${\cal O} (1{\rm GeV})$. We analyze the former in a model-independent way via an effective field theory approach, whereas for the latter we focus on scenarios with light scalar and vector bosons. Using existing experimental bounds, we show that possible NP effects in muon-electron collisions are expected to lie below MUonE's sensitivity. This result confirms and reinforces the physics case of the MUonE proposal.