Model For Prediction Research Articles

Simulation of Friction Stir Welding of AZ31 Mg Alloys.

Friction stir welding has been extensively applied for the high-quality bonding of Mg alloys. The welding temperature caused by friction and plastic deformation is essential for determining the joint characteristics, especially the residual stress and weld microstructure. In this work, a modified moving heat source model was proposed by considering the variations in heat generation caused by friction shear stress at both the side and bottom surfaces of the tool. The application of this model was further extended to the entire welding process, especially in the plunging stage. The relative errors between the experimental and simulated peak temperatures at characteristic points were small, with a maximum of 10%, thereby validating the model for accurate temperature prediction. Furthermore, the influence of welding and rotational speed on temperature fields was systematically investigated. At relatively low welding and rotational speeds, the welding temperature increased significantly with either an increase in rotational speed or a decrease in welding speed. However, this effect gradually diminished at higher welding and rotational speeds. These results provide some valuable guidelines for controlling heat generation to improve the quality of Mg alloy welds.

Improving Predictive Efficacy for Drug Resistance in Novel HIV-1 Protease Inhibitors through Transfer Learning Mechanisms.

The human immunodeficiency virus presents a significant global health challenge due to its rapid mutation and the development of resistance mechanisms against antiretroviral drugs. Recent studies demonstrate the impressive performance of machine learning (ML) and deep learning (DL) models in predicting the drug resistance profile of specific FDA-approved inhibitors. However, generalizing ML and DL models to learn not only from isolates but also from inhibitor representations remains challenging for HIV-1 infection. We propose a novel drug-isolate-fold change (DIF) model framework that aims to predict drug resistance score directly from the protein sequence and inhibitor representation. Various ML and DL models, inhibitor representations, and protein representations were analyzed through realistic validation mechanisms. To enhance the molecular learning capacity of DIF models, we employ a transfer learning approach by pretraining a graph neural network (GNN) model for activity prediction on a data set of 4855 HIV-1 protease inhibitors (PIs). By performing various realistic validation strategies on internal and external genotype-phenotype data sets, we statistically show that the learned representations of inhibitors improve the predictive ability of DIF-based ML and DL models. We achieved an accuracy of 0.802, AUROC of 0.874, and r of 0.727 for the unseen external PIs. By comparing the DIF-based models with a null model consisting of isolate-fold change (IF) architecture, it is observed that the DIF models significantly benefit from molecular representations. Combined results from various testing strategies and statistical tests confirm the effectiveness of DIF models in testing novel PIs for drug resistance in the presence of an isolate.

Evaluation of Advances in Battery Health Prediction for Electric Vehicles from Traditional Linear Filters to Latest Machine Learning Approaches

In recent years, there has been growing interest in Li-ion battery State-of-Health (SOH) estimation due to its critical role in ensuring the safe and reliable operation of Electric Vehicles (EVs). Effective energy management and accurate SOH prediction are essential for the reliability and sustainability of EVs. This paper presents an in-depth review of SOH estimation techniques, starting with an overview of seminal methods that lay the theoretical groundwork for battery modeling and SOH prediction. The review then evaluates recent advancements in Machine Learning (ML) and Artificial Intelligence (AI) techniques, emphasizing their contributions to improving SOH estimation. Through a rigorous screening process, the paper systematically assesses the evolution of these advanced methods, addressing specific research questions to evaluate their effectiveness and practical implications. Key findings highlight the potential of hybrid models that integrate Equivalent Circuit Models (ECMs) with Deep Learning approaches, offering enhanced accuracy and real-time performance. Additionally, the paper discusses limitations of current methods, such as challenges in translating laboratory-based models to real-world conditions and the computational complexity of some prospective methods. In conclusion, this paper identifies promising future research directions aimed at optimizing hybrid models and overcoming existing constraints to advance SOH estimation and battery management in Electric Vehicles.

Flash flood prediction modeling in the hilly regions of Southeastern Bangladesh: A machine learning attempt on present and future climate scenarios

Flash floods are highly destructive, and their frequency and intensity are expected to escalate due to climatic changes. This study thus investigated flash flood susceptibility (FFS) by applying machine learning algorithms and climate projection to predict both present and future hazard scenarios in the southeastern hilly regions of Bangladesh. To predict FFS, we evaluated twelve flood-influencing variables: elevation (EL), slope (SL), aspect (AS), drainage density (DD), distance to stream (DS), topography roughness index (TRI), stream power index (SPI), topographic wetness index (TWI), soil permeability (SP), precipitation (PR), land use and land cover (LULC) and normalized difference vegetation index (NDVI). Earth observation data, field surveys, and past flood records were used to create a detailed flood inventory. Among the machine learning models tested, the random forest (RF) algorithm outperformed others, including support vector machine (SVC), logistic regression (LR), and extreme gradient boosting (XGBoost), and was subsequently used for flood susceptibility mapping based on future precipitation projections under two Sixth Coupled model intercomparison project (CMIP6) climate change scenarios: SSP1-2.6 and SSP5-8.5. Our findings indicated that the areas at high to very high risk of flooding are projected to increase significantly under both the SSP1-2.6 and SSP5-8.5 scenarios. Initially, around 38 % of the studied region had high to very high flood susceptibility, but this is expected to rise to 40–42 % over the projected time periods. These spatial delineations of flood-prone areas can provide guidance for developing effective mitigation and adaptation strategies to address the adverse impacts of flash flooding in the hilly river basins of Bangladesh.

A Contact Mechanics Model for Surface Wear Prediction of Parallel-Axis Polymer Gears.

As surface wear is one of the major failure mechanisms in many applications that include polymer gears, lifetime prediction of polymer gears often requires time-consuming and expensive experimental testing. This study introduces a contact mechanics model for the surface wear prediction of polymer gears. The developed model, which is based on an iterative numerical procedure, employs a boundary element method (BEM) in conjunction with Archard's wear equation to predict wear depth on contacting tooth surfaces. The wear coefficients, necessary for the model development, have been determined experimentally for Polyoxymethylene (POM) and Polyvinylidene fluoride (PVDF) polymer gear samples by employing an abrasive wear model by the VDI 2736 guidelines for polymer gear design. To fully describe the complex changes in contact topography as the gears wear, the prediction model employs Winkler's surface formulation used for the computation of the contact pressure distribution and Weber's model for the computation of wear-induced changes in stiffness components as well as the alterations in the load-sharing factors with corresponding effects on the normal load distribution. The developed contact mechanics model has been validated through experimental testing of steel/polymer engagements after an arbitrary number of load cycles. Based on the comparison of the simulated and experimental results, it can be concluded that the developed model can be used to predict the surface wear of polymer gears, therefore reducing the need to perform experimental testing. One of the major benefits of the developed model is the possibility of assessing and visualizing the numerous contact parameters that simultaneously affect the wear behavior, which can be used to determine the wear patterns of contacting tooth surfaces after a certain number of load cycles, i.e., different lifetime stages of polymer gears.

Comparative Efficiency of Conventional and Connectionist Models in Prediction of First Lactation 305-day Milk Yield in Murrah Buffaloes

Comparative Efficiency of Conventional and Connectionist Models in Prediction of First Lactation 305-day Milk Yield in Murrah Buffaloes

An interpretable machine learning-based model for shear resistance prediction of CFRP-strengthened RC beams using experimental and synthetic dataset

Existing analytical models for predicting the shear resistance of RC beams strengthened with externally bonded CFRP reinforcements exhibit deficient performance due to their inability to accurately capture the complex resisting mechanisms. Combined with significant statistical uncertainties in shear failure, driven by its brittle nature, this further undermines the reliability of these models. To address these limitations, this study leverages Machine Learning (ML) to develop more robust and reliable predictive tool. A rigorous feature-selection process identified eight predictors as the most influential. Subsequently, nine ML-algorithms were trained on a refined experimental dataset comprising 239 beams, with XGBoost emerging as the top performer. This model also outperformed established models likefib Bulletin-90 and ACI 2023 models. However, the limited scope of the experimental dataset constrained the model’s predictive performance especially when separately evaluated on beams strengthened with U-wraps, full wraps or side-bonded FRP configurations. Therefore, to achieve a more reliable model a synthetic dataset was generated using Tabular Variational Auto-Encoder. The XGBoost model trained with the synthetic dataset significantly improved the performance of the former model and exhibited better predictions for all strengthening configurations. Finally, to ensure the physical consistency of predictions, values obtained from the SHapley Additive exPlanations method were analysed.

A Hybrid CNN-Transformer Model for Predicting N Staging and Survival in Non-Small Cell Lung Cancer Patients Based on CT-Scan.

Accurate assessment of N staging in patients with non-small cell lung cancer (NSCLC) is critical for the development of effective treatment plans, the optimization of therapeutic strategies, and the enhancement of patient survival rates. This study proposes a hybrid model based on 3D convolutional neural networks (CNNs) and transformers for predicting the N-staging and survival rates of NSCLC patients within the NSCLC radiogenomics and Nsclc-radiomics datasets. The model achieved accuracies of 0.805, 0.828, and 0.819 for the training, validation, and testing sets, respectively. By leveraging the strengths of CNNs in local feature extraction and the superior performance of transformers in global information modeling, the model significantly enhances predictive accuracy and efficacy. A comparative analysis with traditional CNN and transformer architectures demonstrates that the CNN-transformer hybrid model outperforms N-staging predictions. Furthermore, this study extracts the one-year survival rate as a feature and employs the Lasso-Cox model for survival predictions at various time intervals (1, 3, 5, and 7 years), with all survival prediction p-values being less than 0.05, illustrating the time-dependent nature of survival analysis. The application of time-dependent ROC curves further validates the model's accuracy and reliability for survival predictions. Overall, this research provides innovative methodologies and new insights for the early diagnosis and prognostic evaluation of NSCLC.

Galformer: a transformer with generative decoding and a hybrid loss function for multi-step stock market index prediction

The prediction of stock market fluctuations is crucial for decision-making in various financial fields. Deep learning algorithms have demonstrated outstanding performance in stock market index prediction. Recent research has also highlighted the potential of the Transformer model in enhancing prediction accuracy. However, the Transformer faces challenges in multi-step stock market forecasting, including limitations in inference speed for long sequence prediction and the inadequacy of traditional loss functions to capture the characteristics of noisy, nonlinear stock history data. To address these issues, we introduce an innovative transformer-based model with generative decoding and a hybrid loss function, named “Galformer,” tailored for the multi-step prediction of stock market indices. Galformer possesses two distinctive characteristics: (1) a novel generative style decoder that predicts long time-series sequences in a single forward operation, significantly boosting the speed of predicting long sequences; (2) a novel loss function that combines quantitative error and trend accuracy of the predicted results, providing feedback and optimizing the transformer-based model. Experimental results on four typical stock market indices, namely the CSI 300 Index, S&P 500 Index, Dow Jones Industrial Average Index (DJI), and Nasdaq Composite Index (IXIC), affirm that Galformer outperforms other classical methods, effectively optimizing the Transformer model for stock market prediction.

Degradation Modeling and RUL Prediction of Hot Rolling Work Rolls Based on Improved Wiener Process.

Hot rolling work rolls are essential components in the hot rolling process. However, they are subjected to high temperatures, alternating stress, and wear under prolonged and complex working conditions. Due to these factors, the surface of the work rolls gradually degrades, which significantly impacts the quality of the final product. This paper presents an improved degradation model based on the Wiener process for predicting the remaining useful life (RUL) of hot rolling work rolls, addressing the critical need for accurate and reliable RUL estimation to optimize maintenance strategies and ensure operational efficiency in industrial settings. The proposed model integrates pulsed eddy current testing with VMD-Hilbert feature extraction and incorporates a Gaussian kernel into the standard Wiener process to effectively capture complex degradation paths. A Bayesian framework is employed for parameter estimation, enhancing the model's adaptability in real-time prediction scenarios. The experimental results validate the superiority of the proposed method, demonstrating reductions in RMSE by approximately 85.47% and 41.20% compared to the exponential Wiener process and the RVM model based on a Gaussian kernel, respectively, along with improvements in the coefficient of determination (CD) by 121% and 19.76%. Additionally, the model achieves reductions in MAE by 85.66% and 42.61%, confirming its enhanced predictive accuracy and robustness. Compared to other algorithms from the related literature, the proposed model consistently delivers higher prediction accuracy, with most RUL predictions falling within the 20% confidence interval. These findings highlight the model's potential as a reliable tool for real-time RUL prediction in industrial applications.

Contribution of the large-scale population cohort in disease risk prediction model study: taking United Kingdom Biobank as an example

The disease risk prediction model is the basis of precision prevention and an essential reference for clinical treatment decisions. The development of risk prediction models requires the support of a large amount of high-quality data. A large population cohort study is an important basis for this study. The United Kingdom Biobank (UKB), as a mega-population cohort and biobank, has played an essential role in the exploration of disease etiology and research related to disease prevention and control, with its rich baseline and follow-up data and concepts and mechanisms shared globally. This study followed PRISMA guidelines and included 210 articles with corresponding authors from 18 countries, of which 58 (27.62%) were from the UKB. A total of 491 disease risk prediction models were extracted for cancer, cardiovascular and cerebrovascular diseases, endocrine and metabolic diseases, respiratory diseases, and other diseases and their subgroups, of which 132 were developed by UKB without validation, 183 were developed by UKB with internal validation, 17 were developed by UKB with external validation, and 159 were developed by external development with UKB validation. A total of 188 models used only macro variables (38.29%), and 303 models combined macro and micro variables (61.71%). Model construction methods included survival outcome models, logistic regression, and machine learning. Survival outcome models were dominated by Cox proportional risk regression models and a few models considering competitive risk, accelerated failure models, or different baseline risk functions. Machine learning models included random forest, XGBoost, CatBoost, support vector machine, convolutional neural network, and other methods. The UKB is an essential resource for multiple disease risk prediction modeling studies.

Development and Validation of a Computed Tomography-Based Model for Noninvasive Prediction of the T Stage in Gastric Cancer: Multicenter Retrospective Study.

As part of the TNM (tumor-node-metastasis) staging system, T staging based on tumor depth is crucial for developing treatment plans. Previous studies have constructed a deep learning model based on computed tomographic (CT) radiomic signatures to predict the numberoflymphnodemetastases and survival in patients with resected gastric cancer (GC). However, few studies have reported the combination of deep learning and radiomics in predicting T staging in GC. This study aimed to develop a CT-based model for automatic prediction of the T stage of GC via radiomics and deep learning. A total of 771 GC patients from 3 centers were retrospectively enrolled and divided into training, validation, and testing cohorts. Patients with GC were classified into mild (stage T1 and T2), moderate (stage T3), and severe (stage T4) groups. Three predictive models based on the labeled CT images were constructed using the radiomics features (radiomics model), deep features (deep learning model), and a combination of both (hybrid model). The overall classification accuracy of the radiomics model was 64.3% in the internal testing data set. The deep learning model and hybrid model showed better performance than the radiomics model, with overall classification accuracies of 75.7% (P=.04) and 81.4% (P=.001), respectively. On the subtasks of binary classification of tumor severity, the areas under the curve of the radiomics, deep learning, and hybrid models were 0.875, 0.866, and 0.886 in the internal testing data set and 0.820, 0.818, and 0.972 in the external testing data set, respectively, for differentiating mild (stage T1~T2) from nonmild (stage T3~T4) patients, and were 0.815, 0.892, and 0.894 in the internal testing data set and 0.685, 0.808, and 0.897 in the external testing data set, respectively, for differentiating nonsevere (stage T1~T3) from severe (stage T4) patients. The hybrid model integrating radiomics features and deep features showed favorable performance in diagnosing the pathological stage of GC.

Tree-based classification model for Long-COVID infection prediction with age stratification using data from the national COVID cohort collaborative (N3C)

Abstract Objectives We propose and validate a domain knowledge-driven classification model for diagnosing Post-acute sequelae of SARS-CoV-2 infection (PASC), also known as Long COVID, using Electronic Health Records (EHR) data. Materials and Methods We developed a robust model that incorporates features strongly indicative of PASC or associated with the severity of COVID-19 symptoms as identified in our literature review. The XGBoost tree-based architecture was chosen for its ability to handle class-imbalanced data and its potential for high interpretability. Using the training data provided by the Long COVID Computation Challenge (L3C), which was a sample of the National COVID Collaborative (N3C), our models were fine-tuned and calibrated to optimize Area Under the Receiver Operating Characteristic curve (AUROC) and the F1 score, following best practices for the class-imbalanced N3C data. Results Our age-stratified classification model demonstrated strong performance with an average 5-fold cross-validated AUROC of 0.844 and F1 score of 0.539 across the young adult, mid-aged, and older-aged populations in the training data. In an independent testing dataset, which was made available after the challenge was over, we achieved an overall AUROC score of 0.814 and F1 score of 0.545. Discussion The results demonstrated the utility of knowledge-driven feature engineering in a sparse EHR data and demographic stratification in model development to diagnose a complex and heterogeneously presenting condition like PASC. The model's architecture, mirroring natural clinician decision-making processes, contributed to its robustness and interpretability, which are crucial for clinical translatability. Further, the model’s generalizability was evaluated over a new cross-sectional data as provided in the later stages of the L3C challenge. Conclusion The study proposed and validated the effectiveness of age-stratified, tree-based classification models to diagnose PASC. Our approach highlights the potential of machine learning in addressing the diagnostic challenges posed by the heterogeneity of Long COVID symptoms.

Efficient self-attention based joint optimization for lithology and petrophysical parameter estimation in the Athabasca Oil Sands

Accurately identifying lithology and petrophysical parameters, such as porosity and water saturation, are essential in reservoir characterization. Manual interpretation of well-log data, the conventional approach, is not only labor-intensive but also susceptible to human errors. To address these challenges of lithology identification and petrophysical parameter estimation in the Athabasca Oil Sands area, this study introduces an AutoRegressive Vision Transformer (ARViT) model for lithology and petrophysical parameter prediction. The effectiveness of ARViT lies in its self-attention mechanism and its ability to handle data sequentially, allowing the model to capture important spatial dependencies within the well-log data. This mechanism enables the model to identify subtle spatial and temporal relationships among various geophysical measurements. The model is also interpretable and can serve as an assistive tool for geoscientists, enabling faster interpretation while reducing human bias. The interpretable nature of the model should assist geoscientists in conducting faster quality checks of the predictions, ensuring that errors are not propagated to subsequent stages. This study adopts a multitask learning approach, jointly optimizing the model's performance across multiple tasks simultaneously. To evaluate the effectiveness of the ARViT model, we conducted series of experiments and comparisions, testing it against traditional artificial neural networks (ANN), Long Short-Term Memory (LSTM), and Vision Transformer (ViT) models. To showcase the versatility of ARViT, we apply Low-Rank Adaptation (LoRA) to a different smaller dataset, showing its potential to adapt to different geological contexts. LoRA not only helps in model adaptability but also helps to reduce the number of trainable parameters. Our findings demonstrate that ARViT outperforms ANN, LSTM, and ViT in estimating lithological and petrophysical parameters. While lithology prediction has been a well-explored field, ARViT's unique blend of features, including its self-attention mechanism, autoregression, and multitask approach along with efficient fine tuning using LoRA, sets it apart as a valuable tool for the complex task of lithology prediction and petrophysical parameter estimation.

Using multi-modal spectroscopy technology and microscopic analysis to explore the regulation of ultra-high pressure heat-assisted treatment on the texture of ready-to-eat shrimp during storage

This study employed multi-modal spectroscopy technology and microscopic analysis to investigate the effects of various treatments on the texture of ready-to-eat (RTE) shrimps during storage. The results indicated that ultra-high pressure heat-assisted treatment (UHP-HAT) significantly improved the texture properties of RTE shrimps while simultaneously reducing the carbonyl and trichloroacetic acid-soluble properties associated with protein oxidative degradation. Furthermore, UHP-HAT resulted in a denser and more ordered protein structure, which contributed to an increased content of bound and immobile water. A partial least squares regression model for texture prediction was developed using data obtained from hyperspectral imaging, demonstrating a strong correlation between spectral information and texture parameters. Additionally, infrared imaging was utilized to elucidate the distribution of functional groups (OH, CO, CH) and carbonyl compounds within the proteins. Overall, the findings suggested that UHP-HAT could effectively delay the deterioration of texture by mitigating protein degradation.

Prediction of formation fracture pressure based on reinforcement learning and XGBoost

Abstract Clearly determining the magnitude of fracture pressure is a crucial indicator for fracturing design. Traditional methods for predicting fracture pressure suffer from challenges such as difficulties in obtaining required data, low prediction accuracy, and local limitations in application. In light of these issues, the article proposes a fracture pressure prediction model based on reinforcement learning and XGBoost utilizing geophysical well logging data. Based on the relevance analysis, optimal input parameters, including DEPTH, DEN, AC, GR, CRL, and RT, are selected from geophysical well logging data. We have developed a framework for a fracture pressure prediction model based on XGBoost, wherein hyperparameters are fine-tuned using an improved Q-learning algorithm. The optimized XGBoost model for fracture pressure prediction attains outstanding performance metrics, including an R 2 value of 0.992, a root mean square error of 0.006%, and a mean absolute error of 0.539%. In direct comparison with grid search, Bayesian optimization, and ant colony optimization, the improved Q-learning algorithm emerges as the most effective optimization approach. The predictions generated by the proposed method exhibit remarkable consistency with fracture pressure data measured on-site. This approach successfully addresses the shortcomings encountered with traditional fracture pressure prediction methods, such as inadequate accuracy, demanding data prerequisites, and constrained applicability.

Ecotoxicological risk assessment of active pharmaceutical ingredients (APIs) against different aquatic species leveraging intelligent consensus prediction and i-QSTTR modeling

The increasing presence of active pharmaceutical ingredients (APIs) in aquatic ecosystems, driven by widespread human use, poses significant risks, including acute and chronic toxicity to aquatic species. However, the scarcity of experimental toxicity data on APIs and related compounds due to the high costs, time requirements, and ethical concerns associated with animal testing hinders comprehensive risk assessment. In response, we developed quantitative structure-toxicity relationship (QSTR) and interspecies quantitative structure toxicity-toxicity relationship (i-QSTTR) models for three key aquatic species: zebrafish, water fleas, and green algae, using NOEC as an endpoint, following OECD guidelines. Algae, daphnia, and fish, recognized as standard organisms in toxicity testing, are crucial bio-indicators due to their size, transparency, adaptability, and regulatory acceptance. We used partial least squares (PLS) and multiple linear regression (MLR) methods for model development alongside machine learning techniques such as Random Forest (RF), Support Vector Machines (SVM), K-nearest Neighbor (kNN), and Neural Networks (NN) to enhance the predictivity. Lipophilicity, electronegativity, unsaturation, a molecular cyclized degree in molecular structure, large fragments, aliphatic secondary C(sp2), and R–CR–R groups were identified as critical biomarkers for API toxicity. Screening of the PPDB (pesticide properties databases) and DrugBank validated the practical application of these models, offering valuable tools for regulatory decisions, safer API design, and the preservation of aquatic biodiversity.

Reliability of DFT and multiscale modelling for prediction of emission properties of blue TADF luminophores

Donor-acceptor molecules exhibiting TADF are in the heart of third-generation OLEDs. In this study, we apply (TD)DFT and multiscale modelling to assess the reliability of these approaches for prediction of the emission properties of carbazole-triazine (Cz-TRZ) based TADF luminophores. We found that a series of hybrid and doubly hybrid functionals reproduce the experimentally observed trend in the S1→S0 transition energy for single molecules of Cz-TRZ series, verifying the use of these calculations for the emission wavelength prediction. Considering molecular environment of luminophores within QM/MM and QM/EFP methods allowed us to model inhomogeneous broadening of the emission spectra; impact of long-range corrected functionals was revealed. The results obtained provide useful tips for computational studying of the structure-property relationships for TADF luminophores and semiquantitative prediction of their properties. We anticipate that it will stimulate rational design of promising TADF luminophores for doped and non-doped OLEDs.

Probabilistic model for high-level intention estimation and trajectory prediction in urban environments

To enable successful automated driving, precise behavior prediction of surrounding vehicles is indispensable in urban traffic scenarios. Furthermore, given that a vehicle’s behavior is influenced by the movements of other road users, it becomes crucial to estimate their intentions to anticipate precise future motion. However, the elevated complexity resulting from interdependencies among traffic participants and the uncertainty arising from the object recognition errors present additional challenges. Despite extensive research on inferring intentions, many studies have concentrated on estimating intentions from interactions, resulting in a lack of practicality in urban traffic environments due to low computational efficiency and low robustness against recognition failure of strongly interacting road users. In this paper, we introduce a practical stochastic model for intention estimation and trajectory prediction of surrounding vehicles in automated driving under urban traffic environments. The trajectory is forecasted based on hierarchically computed and probabilistically estimated intentions, which represent an interpretation of vehicle behavior, utilizing only the kinematic state of the focal vehicle and HD maps to ensure real-time performance and enhance robustness. The evaluated results demonstrate that the proposed model surpasses straightforward methods in terms of accuracy while maintaining computational efficiency and exhibits robustness against the recognition failure of traffic participants which strongly influence the focal vehicle.

Evaluating MPAS-A Performance for Mesoscale Simulation in a Tropical Region: A Case Study of Extreme Heat in Jakarta, Indonesia

The Model for Prediction Across Scales–Atmosphere (MPAS-A) has been widely used for larger scale simulations, but its performance in mesoscale, particularly in tropical regions, is less evaluated. This study aimed to assess MPAS-A in simulating extreme surface air temperature in Jakarta during the hot spells of October 2023 with eight different simulation setups. Several validation metrics were applied to near-surface meteorological variables, land surface temperature (LST), and vertical atmospheric profile. From the eight simulations, MPAS-A captured diurnal patterns of the near-surface variables well, except for wind direction. The model also performed well in LST simulations. Moreover, the biases in the vertical profiles varied with height and were sensitive to the initial/boundary conditions used. Simulations with modified terrestrial datasets showed higher LST and air temperatures over the sprawling urban areas. MPAS-A successfully simulated the extreme event, showing higher air temperatures in southern Jakarta (over 36 °C) compared to the northern part. Negative temperature advection by sea breeze helped lower air temperature in the northern area. This study highlights the role of sea breezes as natural cooling mechanisms in coastal cities. Additionally, MPAS-A is feasible for several applications for urban climate studies and climate projection, although further development is needed.