Probability Estimates Research Articles

Broad-scale predictions of herpetofauna occupancy and colonization in an agriculturally dominated landscape.

Predictions of species occurrence allow land managers to focus conservation efforts on locations where species are most likely to occur. Such analyses are rare for herpetofauna compared to other taxa, despite increasing evidence that herptile populations are declining because of landcover change and habitat fragmentation. Our objective was to create predictions of occupancy and colonization probabilities for 15 herptiles of greatest conservation need in Iowa. From 2006-2014, we surveyed 295 properties throughout Iowa for herptile presence using timed visual-encounter surveys, coverboards, and aquatic traps. Data were analyzed using robust design occupancy modeling with landscape-level covariates. Occupancy ranged from 0.01 (95% CI = -0.01, 0.03) for prairie ringneck snake (Diadophis punctatus arnyi) to 0.90 (95% CI = 0.898, 0.904) for northern leopard frog (Lithobates pipiens). Occupancy for most species correlated to landscape features at the 1-km scale. General patterns of species' occupancy included negative effects of agricultural features and positive effects of water features on turtles and frogs. Colonization probabilities ranged from 0.007 (95% CI = 0.006, 0.008) for spiny softshell turtle (Apalone spinifera) to 0.82 (95% CI = 0.62, 1.0) for western fox snake (Pantherophis ramspotti). Colonization probabilities for most species were best explained by effects of water and grassland landscape features. Predictive models had strong support (AUC > 0.70) for six out of 15 species (40%), including all three turtles studied. Our results provide estimates of occupancy and colonization probabilities and spatial predictions of occurrence for herptiles of greatest conservation need across the state of Iowa.

Cyber risk management technology to strengthen the information security of the national economy

Purpose. Developing a technology for managing cyber risks based on their improved classification by the level of impact on the occurrence of an extreme situation. Methodology. To achieve the goal, general scientific and special methods of cognition were used in the study: dialectical and systemic approaches, analysis and synthesis, logical generalization and grouping, structural-logical method, iterative approach, modeling, method of formal representations of uncertainty. Findings. A cyber risk management technology has been developed, consisting of four main stages: analysis of cyber threats (context establishment; security audit; formation of scenario concepts); scenario modeling (threat decomposition; scenario formation; setting criteria; setting probability estimates of concepts (variables); building a network architecture; formation of a private threat model; scenario analysis); risk assessment; object classification. The proposed approach to cybersecurity risk management provides vulnerability detection and risk assessment (risk potential) and simplifies the development of management solutions to prevent events affecting cybersecurity. Originality. The proposed technology differs from the existing ones by focusing on identifying those vulnerabilities and cyber threats that, according to their improved classification by the level of impact on the occurrence of an extreme situation, can lead to serious disruptions in the functioning of critical information infrastructure of the national economy. Practical value. The practical significance of the study lies in the fact that the proposed cyber risk management technology is one of the tools for preventing the realization of risks in cyberspace and the basis for strengthening the information security of economic entities in particular and the national economy as a whole.

Rapid remote volcanic ashfall impact assessment for the 2022 eruption of Hunga volcano, Tonga: a bespoke approach and lessons identified

When disasters occur, rapid impact assessments are required to prioritise response actions, support in-country efforts and inform the mobilisation of aid. The 15 January 2022 eruption of Hunga volcano, Tonga, and the resulting atmospheric shockwave, ashfall, underwater mass disturbance and tsunami, caused substantial impacts across the Kingdom of Tonga. Volcanic impacts on the scale observed after the eruption are rare, necessitating a reliance on international advice and assistance. The situation was complicated by the loss of Tonga’s international submarine fibreoptic cable (causing a complete loss of communications for approximately 20 days) along with border closures due to the COVID-19 pandemic. A need emerged for a rapid remote volcanic impact assessment and provision of specialist advice to help inform the response of international partners. Here we present a novel methodology for conducting rapid remote volcanic ashfall impact assessments, conducted over a 10-day period following the eruption. We used three different hazard models for ashfall thickness across the main island of Tongatapu and available asset information and vulnerability functions for buildings, agriculture, electricity networks, water supply and roads, to provide initial estimates of losses due to ashfall from the 15 January eruption. For buildings, we estimated losses both as total losses and as percentages of the total replacement cost of buildings on Tongatapu. For agriculture, we made probabilistic estimates of production losses for three different crop classes. For ashfall clean-up, we estimated ranges of ashfall volumes requiring clean-up from road surfaces and roofs. For water supply, electricity networks and roads, our analysis was limited to assessing the exposure of important assets to ashfall, as we had insufficient information on system configurations to take the analysis further. Key constraints on our analysis were the limited nature of critical infrastructure asset inventories and the lack of volcanic vulnerability models for tropical regions including Pacific Island nations. Key steps towards iteratively improving rapid remote impact assessments will include developing vulnerability functions for tropical environments as well as ground-truthing estimated losses from remote approaches against in-person impact assessment campaigns.

Multi-modality deep learning model for prediction of chronic obstructive coronary artery disease

Abstract Background/Introduction The incidence of coronary artery disease (CAD) is on the rise with an increasing burden of stable ischemic heart disease (SIHD) and high prevalence of obesity, diabetes, hyperlipidemia, and other cardiovascular risk factors. Pre-test probability estimation for CAD based on demographic features and characterization of chest pain is insufficient to accurately assess whether further evaluation for CAD is warranted. Purpose To develop and validate a deep learning (DL) algorithm utilizing electrocardiogram (ECG) waveforms and clinical features to predict obstructive CAD. Methods Clinical data from subjects undergoing invasive angiography over a 4-year period at a quaternary care center exclusive of subjects who had concern for acute coronary syndromes were used for model training. ECGs obtained up to 5 years prior to angiography were included; ECGs were partitioned 8:1:1 for training, validation, and test datasets. Obstructive CAD was defined as receiving percutaneous coronary intervention during elective angiography. Convolutional neural network models were developed for ECG waveforms alone (DL-ECG), clinical features from standard risk scores (DL-Clinical), and the combination of ECG waveforms and clinical features (DL-ECG+Clinical); a commonly used pre-test probability estimation tool derived from the CAD Consortium study was used as a comparison (PTP). The models were evaluated based on area under the curve receiver operating characteristics (AUC). Shapley additive explanations (SHAP) analysis was performed to understand feature importance in the DL models. Results A total of 40,602 ECGs and 8,361 coronary angiograms from 7,201 patients were included in the analysis. The median age was 66 (IQR 56-75), 71.9% were men, 34.9% had hypertension, 30.6% had diabetes mellitus, and 39.0% had dyslipidemia. The PTP model (AUC 0.733 [0.717-0.750]) had similar performance as the DL-Clinical model (AUC 0.762 [0.746-0.778]). The DL-ECG model (AUC 0.741 [0.726-0.758]) had similar performance as both the clinical feature models. The DL-ECG+Clinical model (AUC 0.807 [0.793-0.822]) trained on ECG waveforms and clinical features had a superior performance (Figure 1). SHAP analysis for the DL-ECG+Clinical model revealed that ECG waveforms had the largest influence on the model’s output followed by age and chest pain type (Figure 2). Conclusions Multi-modality DL model utilizing ECG and commonly used features can improve prediction of obstructive CAD in SIHD compared to traditional pre-test probability estimates. Prospective research is warranted to determine whether incorporation of DL methods to ECG analysis effectively improves diagnosis and outcomes in obstructive CAD.Figure 1.Figure 2.

A Brief Introduction on Latent Variable Based Ordinal Regression Models With an Application to Survey Data.

The analysis of survey data is a frequently arising issue in clinical trials, particularly when capturing quantities which are difficult to measure. Typical examples are questionnaires about patient's well-being, pain, or consent to an intervention. In these, data is captured on a discrete scale containing only a limited number of possible answers, from which the respondent has to pick the answer which fits best his/her personal opinion. This data is generally located on an ordinal scale as answers can usually be arranged in an ascending order, for example, "bad", "neutral", "good" for well-being. Since responses are usually stored numerically for data processing purposes, analysis of survey data using ordinary linear regression models are commonly applied. However, assumptions of these models are often not met as linear regression requires a constant variability of the response variable and can yield predictions out of the range of response categories. By using linear models, one only gains insights about the mean response which may affect representativeness. In contrast, ordinal regression models can provide probability estimates for all response categories and yield information about the full response scale beyond the mean. In this work, we provide a concise overview of the fundamentals of latent variable based ordinal models, applications to a real data set, and outline the use of state-of-the-art-software for this purpose. Moreover, we discuss strengths, limitations and typical pitfalls. This is a companion work to a current vignette-based structured interview study in pediatric anesthesia.

ADVANCED PREDICTIVE MODELING FOR THE BIOFIRE FILMARRAY SYSTEM IN INFECTIOUS DISEASE DIAGNOSTICS

The Bio Fire Film Array System has emerged as a powerful tool in clinical diagnostics, offering rapid and multiplexed detection of infectious pathogens. This study presents the development of a comprehensive mathematical model aimed at predicting the accuracy and sensitivity of the BioFire FilmArray System under diverse conditions. Key factors such as sample type, pathogen load, specimen characteristics, and environmental variables were systematically analyzed to elucidate their impact on the system's performance. The model incorporates a range of parameters including sample volume, specimen complexity, microbial diversity, and system-specific variables to generate probabilistic estimates of diagnostic outcomes. By leveraging large datasets encompassing diverse clinical scenarios and pathogen profiles, the model achieves robustness and generalizability, thereby facilitating its applicability across different healthcare settings and populations. Validation of the model against independent datasets demonstrates its efficacy in accurately predicting the performance of the BioFire FilmArray System across a spectrum of conditions. Overall, this mathematical model represents a significant advancement in optimizing the utility of the BioFire FilmArray System for infectious disease diagnosis. Its integration into clinical practice holds promise for improving patient care, guiding sample processing protocols, and informing decision-making processes in infectious disease management.

Radiomics-Based Machine Learning with Natural Gradient Boosting for Continuous Survival Prediction in Glioblastoma

(1) Background: Glioblastoma (GBM) is the most common primary malignant brain tumor in adults, with an aggressive disease course that requires accurate prognosis for individualized treatment planning. This study aims to develop and evaluate a radiomics-based machine learning (ML) model to estimate overall survival (OS) for patients with GBM using pre-treatment multi-parametric magnetic resonance imaging (MRI). (2) Methods: The MRI data of 865 patients with GBM were assessed, comprising 499 patients from the UPENN-GBM dataset and 366 patients from the UCSF-PDGM dataset. A total of 14,598 radiomic features were extracted from T1, T1 with contrast, T2, and FLAIR MRI sequences using PyRadiomics. The UPENN-GBM dataset was used for model development (70%) and internal validation (30%), while the UCSF-PDGM dataset served as an external test set. The NGBoost Survival model was developed to generate continuous probability estimates as well as predictions for 6-, 12-, 18-, and 24-month OS. (3) Results: The NGBoost Survival model successfully predicted survival, achieving a C-index of 0.801 on internal validation and 0.725 on external validation. For 6-month OS, the model attained an AUROC of 0.791 (95% CI: 0.742–0.832) and 0.708 (95% CI: 0.654–0.748) for internal and external validation, respectively. (4) Conclusions: The radiomics-based ML model demonstrates potential to improve the prediction of OS for patients with GBM.

Cost-Effectiveness of a Shock Team Approach in Refractory Cardiogenic Shock.

Multidisciplinary Shock Teams have improved clinical outcomes for cardiogenic shock, but their implementation costs have not been studied. This study's objective was to compare costs between patients treated with and without a Shock Team and determine if the team's implementation is cost-effective compared with standard of care. We examined patients with refractory cardiogenic shock treated with or without a Shock Team at a tertiary academic hospital from 2009 to 2018. Real-world hospital data were used to compare costs and outcomes, including survival at discharge, 1-year survival, and quality-adjusted life years gained at 1 year. Incremental cost-effectiveness ratios were calculated over a 1-year time horizon, with parameter uncertainty evaluated through probabilistic sensitivity analysis using 1000 second-order Monte Carlo simulations. The study involved 244 patients, with 123 treated by the Shock Team and 121 receiving standard of care. Patients were predominantly male (77.5%), with a mean age of 58 (18-92) years. The Shock Team approach improved survival rates at hospital discharge and 1-year follow-up (61.0% versus 47.9%; P=0.04 and 55.0% versus 40.5%; P=0.03, respectively). The incremental cost-effectiveness ratio for increases in survival probability at discharge for the multidisciplinary Shock Team compared with standard of care was $102 088. The incremental cost-effectiveness ratio for increases in survival probability at 1-year was estimated at $96 152 and at $127 862 per 1 quality-adjusted life year gained. Probabilistic sensitivity analysis estimates showed that the Shock Team was cost-effective in the majority of simulations using a willingness-to-pay threshold of $150 000, while it was also dominant in almost one-third of the simulations. The Shock Team approach for treating refractory cardiogenic shock may be a cost-effective alternative to traditional standard of care. These findings can help prioritize the implementation of Shock Team initiatives to further improve cardiogenic shock outcomes.

A petrophysically driven seismic inversion method for the total organic carbon content of source rocks

Organic carbon content is the main index for evaluating organic matter abundance of source rocks, and it is still difficult to quantitatively predict total organic carbon (TOC) in source rocks based on seismic data. The study of seismic fluid identification driven by petrophysics can help to understand the fluid characteristics and distribution patterns of subsurface oil and gas reservoirs. First, this paper clarifies the sweet spot parameters (parameters characterizing hydrocarbon enrichment) and sensitive elastic parameters (a parameter characterizing the nature of an ideal elastic body) of source rocks through theoretical petrophysical modeling. Next, the paper establishes the relationship between sensitive elastic parameters and sweet spot parameters TOC and constructs a statistical petrophysical model that can characterize the relationship between the two on this basis. And then we construct the joint distribution of TOC and elastic impedance through the Bayesian theoretical framework to obtain the maximum posterior probability estimate as the final TOC inversion results of source rocks. Our method successfully predicts the spreading of high-quality source rocks in the Wen4 Section of Lufeng 13 Subsag, and the inversion results are within an uncertainty range of ±14 m for well data, which proves the reliability of the method. The prediction results show that the organic matter abundance of source rocks in the Wen4 Section is high, and the organic carbon content is generally higher than 2%, which provides a reliable basis for the further implementation of the resource scale of the depression and the clarification of the hydrocarbon-rich area, which provides technical support for the evaluation of the source rocks of the new depression in the new area.

Quantitative Assessments of the Liquefaction Hazard of Soils considering Possible Strong Earthquakes in Seismically Active Regions of Russia

Introduction The seismogenic liquefaction of the soil poses a great hazard to society and the environment. Therefore, it is actively studied in many countries. In Russian engineering and seismological practice, this area is not sufficiently developed. The deterministic approach still prevails in Russian research on this topic. More modern probabilistic estimates are very rare. Methods This paper describes examples of both deterministic and probabilistic assessments of the seismogenic liquefaction hazard performed in certain areas of the Russian territory with different seismogeological conditions. Deterministic estimates were made using the Iwasaki-Seed-Finn methods and their modifications. Probability distribution functions of a random variable, the “seismic potential of liquefaction” (SPL), were developed for probabilistic estimates. These functions are regional in nature and take into account two types of uncertainties. The first is the uncertainty in achieving “critical” values by the SPL value in the event of potentially dangerous earthquake sources for a given location. The second is the uncertainty in the very occurrence of these sources in a given place for a given period of time. The “critical” SPL values are determined by the strength properties of the site soils. All estimates are based on multivariate calculations using various models of strong ground motions and seismicity. In all cases, the probability of liquefaction of water-saturated sandy and sandy-loam deposits was estimated which were found near the seabed and at depths of up to 80 m in the waters of Pogibi cape (the coast of Sakhalin island), in the districts of Sochi and Novorossiysk, as well as in land conditions (Stavropol, Krasnodar). Results The results of the research made it possible to correctly (at the quantitative level) take into account this component of the seismic hazard of the studied territories. Conclusion The variants of practical use of the obtained data are offered. An assessment of the possibilities and limitations of the developed methodology is made, and ways to improve it are outlined.

Can Expert Prediction Markets Forecast Climate-Related Risks?

Abstract Prediction markets use a betting-like mechanism to aggregate information from disparate sources to produce probability forecasts that represent the collective view of participants. They have similarities with financial markets but are designed specifically to discover information rather than transfer assets or risks. Contracts that pay out a fixed amount if a well-defined event occurs are traded in these markets and, under certain conditions, the price at which these event-contracts trade can be interpreted as a probability estimate that the event will occur. This article examines the performance of 24 prediction markets for climate-related variables that have been run over the past 5 years. The predicted variables included the monthly temperature and rainfall for the United Kingdom, an index of El Niño–Southern Oscillation, Atlantic hurricane activity, and U.K. wheat yield. The markets had horizons of 2–12 months. Invitations to participate in the markets were extended to individuals and teams with relevant expertise. Participants did not have to pay to take part but did receive cash rewards based on their performance. Trades were made through an automated market maker overcoming the problems of low activity that have affected previous prediction markets for specialized topics. The predictions of the markets were consistent with good reliability, given the resolving power afforded by the sample size. Whether this level of reliability would persist for longer, multiyear, horizons relevant to climate change cannot be known without running markets on such time scales, something that we strongly advocate.

A Knowledge Distillation Method Based on Evidence Theory to Prevent Catastrophic Forgetting

Abstract Incremental learning is an emerging machine-learning approach designed to prevent catastrophic forgetting while learning a new task with knowledge retained from previous tasks. However, existing methods often must fully utilize the distributional information of the old task model’s outputs. To alleviate this, we propose a class incremental learning method grounded in evidence theory to leverage the distributional information of outputs from the old task model, which integrates uncertainty estimation into the knowledge distillation process to ensure the new task model effectively learns from the old task model’s distribution information. Specifically, we incorporate uncertainty estimation based on evidence theory to calculate knowledge distillation loss during training. We compute the uncertainty estimates of outputs from both the new and old task models, and then they are used in the knowledge distillation loss calculation. We propose a novel classification strategy that considers outputs’ probability and uncertainty estimates, which could determine the sample category without requiring an additional training phase or supplementary models. Experiments on the CIFAR-100 and ImageNet datasets demonstrate the effectiveness of the proposed method, showing that it outperforms existing methods by improving accuracy by 1%-2%. Results suggest that leveraging the distribution information of model outputs can effectively mitigate catastrophic forgetting in deep learning models within incremental learning scenarios.

Stepwise updating of probabilities is neither universal nor fully explained by motor costs.

Often our expectations are set by sampling many times from the same distribution. When that distribution changes, so should our expectations. If we want to decide whether to take our umbrella today we need to have tracked, and updated, our estimate of the probability of rain by reference to recent temperatures and precipitation. Under debate is whether we update our mental estimates of probabilities given each new incoming piece of evidence or whether we stick with a current estimate until it becomes clearly in need of change. Previous research has suggested that participant estimates of running probabilities are not updated on every trial, but only intermittently. This has been used to support change-point models of probability updating. However, such a pattern could also be explained by the way common laboratory procedures impose a motor cost to update the probability report. This study was designed to remove the motor confound. Our procedure required similar motor actions for both changing and maintaining one's probability estimate. At a group level, motor cost did affect updating frequency and removing the default response option encouraged more frequent updating. However, intermittent updating response patterns did not disappear completely. Despite this equivalence in response effort, some participants, even when forced to make a new estimate on every trial, continued to update rarely while other participants meticulously updated every trial. We conclude deliberate updating frequency is heterogenous but intermittent updating is not simply an artifact of motor cost.

Extremal fixed points and Diophantine equations

The coupling constants of fixed points in the ϵ expansion at one loop are known to satisfy a quadratic bound due to Rychkov and Stergiou. We refer to fixed points that saturate this bound as extremal fixed points. The theories which contain such fixed points are those which undergo a saddle-node bifurcation, entailing the presence of a marginal operator. Among bifundamental theories, a few examples of infinite families of such theories are known. A necessary condition for extremality is that the sizes of the factors of the symmetry group of a given theory satisfy a specific Diophantine equation, given in terms of what we call the extremality polynomial. In this work we study such Diophantine equations and employ a combination of rigorous and probabilistic estimates to argue that these infinite families constitute rare exceptions. The Pell equation, Falting’s theorem, Siegel’s theorem, and elliptic curves figure prominently in our analysis. In the cases we study here, more generic classes of multi-fundamental theories saturate the Rychkov-Stergiou bound only in sporadic cases or in limits where they degenerate into simpler known examples.

Acute cholecystitis diagnosis in the emergency department: an artificial intelligence-based approach.

This study aimed to assess the diagnostic performance of a support vector machine (SVM) algorithm for acute cholecystitis and evaluate its effectiveness in accurately diagnosing this condition. Using a retrospective analysis of patient data from a single center, individuals with abdominal pain lasting one week or less were included. The SVM model was trained and optimized using standard procedures. Model performance was assessed through sensitivity, specificity, accuracy, and AUC-ROC, with probability calibration evaluated using the Brier score. Among 534 patients, 198 (37.07%) were diagnosed with acute cholecystitis. The SVM model showed balanced performance, with a sensitivity of 83.08% (95% CI: 71.73-91.24%), a specificity of 80.21% (95% CI: 70.83-87.64%), and an accuracy of 81.37% (95% CI: 74.48-87.06%). The positive predictive value (PPV) was 73.97% (95% CI: 65.18-81.18%), the negative predictive value (NPV) was 87.50% (95% CI: 80.19-92.37%), and the AUC-ROC was 0.89 (95% CI: 0.85 to 0.93). The Brier score indicated well-calibrated probability estimates. The SVM algorithm demonstrated promising potential for accurately diagnosing acute cholecystitis. Further refinement and validation are needed to enhance its reliability in clinical practice.

A Modeling Framework for Quantifying Spatial Recruitment Dynamics Using Abundance Estimation and Sibship Analysis

Quantifying recruitment at the sibling group offers a powerful methodology for understanding density-dependent and environmental drivers of recruitment. We propose a modeling framework that combines sibship and abundance estimation datasets to estimate mean sibling group size, sibling group size process error, environmental and density-dependent effects on sibling group size, dispersal, and mortality rate. Geographic states in the model consist of discrete habitat patches connected via dispersal. Simulations were used to investigate the influence of sampling processes and sibling group size on parameter estimation within our modeling framework. Mean sibling-group size, environmental effects on recruitment, and dispersal rate among habitat patches were estimated with high accuracy under a wide range of sampling conditions, including imprecise out-of-model estimates of capture probability and subsampling both within and among habitat patches. Density-dependent effects on recruitment and process error tended to be estimated with lower accuracy, though accuracy improved as sibling group size or sampling intensity increased. The main contribution of this research is a flexible quantitative modeling framework for parameterizing mechanistic models of recruitment dynamics with empirical sibship data.

Neural dynamics underlying the illusion of control during reward processing.

The illusion of control refers to a behavioral bias in which people believe they have greater control over completely stochastic events than they actually do, leading to an inflated estimate of reward probability than objective probability warrants. In this study, we examined how reward system is modulated by the illusion of control through the lens of neural dynamics. Participants in a behavioral task exhibited a classical illusion of control, assigning a higher value to the gambling wheels they picked themselves than to those given randomly. An event-related potential study of the same task revealed that this behavioral bias is associated with reduced reward anticipation, as indexed by the stimulus-preceding negativity, diminished positive prediction error signals, as reflected by the reward positivity, and enhanced motivational salience, as revealed by the P300. Our findings offer a mechanistic understanding of the illusion of control in terms of reward dynamics.

Neural dynamics of reversal learning in the prefrontal cortex and recurrent neural networks.

In probabilistic reversal learning, the choice option yielding reward at higher probability switches at a random trial. To perform optimally in this task, one has to accumulate evidence across trials to infer the probability that a reversal has occurred. In this study, we investigated how this reversal probability is represented in cortical neurons by analyzing the neural activity in prefrontal cortex of monkeys and recurrent neural networks trained on the task. We found that neural trajectories encoding reversal probability had substantial dynamics associated with intervening behaviors necessary to perform the task. Furthermore, the neural trajectories were translated systematically in response to whether outcomes were rewarded, and their position in the neural subspace captured information about reward outcomes. These findings suggested that separable dynamic trajectories, instead of fixed points on a line attractor, provided a better description of neural representation of reversal probability. Near the behavioral reversal, in particular, the trajectories shifted monotonically across trials with stable ordering, representing varying estimates of reversal probability around the reversal point. Perturbing the neural trajectory of trained networks biased when the reversal trial occurred, showing the role of reversal probability activity in decision-making. In sum, our study shows that cortical neurons encode reversal probability in a family of dynamic neural trajectories that accommodate flexible behavior while maintaining separability to represent distinct probabilistic values.

Using the Observed Variations of the Start Date of the Rainy Season over Central America for Its Reliable Seasonal Outlook

Abstract We introduce a simple method to define the start and the end of the rainiest part of the year as the first and the last day of the year when the daily rain rate is more or less than the annual mean climatological rain rate for a region or at a given grid point of the rainfall analysis, respectively. A novelty of this work is the adoption of a perturbation technique to generate a total of 1001 ensemble members to account for observational and analysis uncertainties. This allows for a probabilistic estimate of the start and retreat dates of the rainy season at the granularity of the Integrated Multi-satellitE Retrievals for Global Precipitation Measurement (IMERG), version 6, rainfall analysis over Central America. The seasonal cycle of the IMERG rainfall analysis is also found to verify with in situ observations in the region. Many large-scale climate drivers affect regional rainfall, often with complex interactions that affect the onset date, retreat date, and magnitude of the seasonal rainfall cycle, making it difficult to predict the length or total quantity of seasonal rainfall using climate drivers alone. Once an onset date is established, however, this metric alone can be more indicative of both the length and total seasonal rainfall anomaly than predicting how the climate drivers will interact to affect the quantity and duration of upcoming seasonal rainfall. The local relationships of the start date with seasonal length and rainfall anomaly are leveraged to produce effective seasonal outlooks of the rainy season for the region by just monitoring the start date variations. Significance Statement The start date and retreat date of the rainy season in Central America are defined every year, precisely to a specific date from our proposed definition of it. This is possible because the region exhibits a strong seasonality of the rainfall. As a result, the year-to-year (interannual) variation of the seasonal rainfall during the rainy season is also determined by the variations in the length of the season that are usually overlooked in fixed calendar seasons. We define the start or retreat date of the rainy season as the first or the last day of the year when the daily rain rate exceeds or falls below the average daily rainfall from 2001 to 2022, respectively. We also find that both start and retreat date variations independently influence the length and total seasonal rainfall of the rainy season. Consequently, these relationships are leveraged to provide an outlook of the forthcoming rainy season from the diagnosis of the variations of its onset date, which is shown to be an effective predictor with significant useful seasonal prediction skills.

Simultaneous identification of propeller fluctuating forces and unknown parameters based on a Bayesian inversion approach

Accurate quantification of propeller excitation is essential for effective vibro-acoustic analysis and mitigation in marine vessels. This study introduces a Bayesian statistical framework that indirectly reconstruct uncertain forces and system parameters by fusing available prior information, forward modeling and measured vibration responses through Bayesian theorem. The Markov Chain Monte Carlo (MCMC) algorithm, enhanced with Polynomial Chaos Kriging (PC-Kriging) surrogates, is employed to facilitate efficient estimation of likelihood functions and robust uncertainty propagation, providing posterior probabilistic estimates and credible intervals for the identified forces. Additionally, the framework incorporates model class selection based on model evidence to determine the most plausible empirical spectrum model for the propeller broadband thrust spectrum. Specifically, this study proposes and evaluates two candidate empirical spectrum models, the Ornstein–Uhlenbeck (OU)-Gaussian and OU-Weibull models, which effectively capture the frequency–amplitude characteristics of unsteady thrust. Numerical results demonstrate substantial parameter uncertainty reduction, accurate thrust spectrum reconstruction, and effective quantification of statistical properties. Moreover, the OU-Gaussian model is revealed to be more plausible for characterizing the propeller thrust spectrum and response prediction. This framework may offer a practical solution for propeller excitation identification, enabling rapid vibro-acoustic performance assessments of marine vessels.