TTTerrence S. Tricco

Terrence S. Tricco

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Terrence S. Tricco has a total of 46 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    A Comparison of Two- and One-fluid Dust Solvers for Dusty, Supersonic Turbulence

    by Narges Vadood, Terrence S. Tricco

    We model 3 and 10 μ m dust grains in supersonic, turbulent molecular cloud conditions to compare two different numerical methods solving dust coupled to gas through a drag term. One method models dust and gas as two separate species (two fluid) and the other the combination of dust and gas as a single mixture (one fluid) using the diffusion approximation. Simulations are performed in a 3D periodic box using the Phantom code. The gas probability distributions are consistent with a lognormal distribution for both methods and grain sizes. The dust distributions are different in the two methods, showing discrepancies particularly in regions of low density. The most significant difference between the two methods is in the dust-to-gas ratio distributions. Both methods peak at the mean dust-to-gas ratio of 0.01, but the two-fluid method has wider distributions than the one-fluid method suggesting excess dust concentration in dense filaments. Filaments are where the one-fluid method results are most accurate. The narrowing of the one-fluid distribution is also caused by the stopping time limiter used to maintain the terminal velocity approximation. This artificially makes the mixture more coupled in low-density regions. Our overall conclusion is that both methods are viable for the study of dust in molecular clouds, but that the correct method should be chosen based on the numerical Stokes number regime of the calculation to avoid numerical artifacts.

  • Open Access

    Mapping Galaxy Images across Ultraviolet, Visible, and Infrared Bands Using Generative Deep Learning

    by Youssef Zaazou, Alex Bihlo, Terrence S. Tricco

    We demonstrate that generative deep learning can translate galaxy observations across ultraviolet, visible, and infrared photometric bands. Leveraging mock observations from the Illustris simulations, we develop and validate a supervised image-to-image model capable of performing both band interpolation and extrapolation. The resulting trained models exhibit high fidelity in generating outputs, as verified by both general image comparison metrics (mean absolute error, SSIM, PSNR) and specialized astronomical metrics (Gini coefficient, M20). Moreover, we show that our models can be used to predict real-world observations, using data from the DECaLS survey as a case study. These findings highlight the potential of generative learning to augment astronomical data sets, enabling efficient exploration of multiband information in regions where observations are incomplete. This work opens new pathways for optimizing mission planning, guiding high-resolution follow-ups, and enhancing our understanding of galaxy morphology and evolution.


  • Open Access

    Mapping Galaxy Images Across Ultraviolet, Visible and Infrared Bands Using Generative Deep Learning

    by Youssef Zaazou, Alex Bihlo, Terrence S. Tricco

    We demonstrate that generative deep learning can translate galaxy observations across ultraviolet, visible, and infrared photometric bands. Leveraging mock observations from the Illustris simulations, we develop and validate a supervised image-to-image model capable of performing both band interpolation and extrapolation. The resulting trained models exhibit high fidelity in generating outputs, as verified by both general image comparison metrics (MAE, SSIM, PSNR) and specialized astronomical metrics (GINI coefficient, M20). Moreover, we show that our model can be used to predict real-world observations, using data from the DECaLS survey as a case study. These findings highlight the potential of generative learning to augment astronomical datasets, enabling efficient exploration of multi-band information in regions where observations are incomplete. This work opens new pathways for optimizing mission planning, guiding high-resolution follow-ups, and enhancing our understanding of galaxy morphology and evolution.

  • Open Access

    Smoothed particle magnetohydrodynamics

    by Terrence S. Tricco

    Smoothed particle magnetohydrodynamics has reached a level of maturity that enables the study of a wide range of astrophysical problems. In this review, the numerical details of the modern SPMHD method are described. The three fundamental components of SPMHD are methods to evolve the magnetic field in time, calculate accelerations from the magnetic field, and maintain the divergence-free constraint on the magnetic field (no monopoles). The connection between these three requirements in SPMHD will be highlighted throughout. The focus of this review is on the methods that work well in practice, with discussion on why they work well and other approaches do not. Numerical instabilities will be discussed, as well as strategies to overcome them. The inclusion of non-ideal MHD effects will be presented. A prospective outlook on possible avenues for further improvements will be discussed.


  • Ego-Aware Graph Neural Network

    by Zhihao Dong, Yuanzhu Chen, Terrence S. Tricco, Cheng Li, Ting Hu

    In scientific exploration and daily life, inter-connectivity between entities is ubiquitous, such as species via the food webs, people among social networks, and products and customers in e-commerce. Graphs and networks are natural constructs to model such linkages, which can capture the relational information in addition to features provided by individual nodes. Graph neural networks (GNNs) are a powerful computation tool for reasoning about interconnected data and can help us make better use of the information imbued in data collectively. GNNs have found many successful applications, such as predicting drug side-effects, classifying diseases, and predicting the function of proteins. Most existing GNN methods are designed for datasets with moderate node features, where individual node features provide limited information, and assortative graph datasets, where nodes and their neighbours are more likely to be similar. However, for graphs with rich node features, or for the disassortative graphs, where nodes and their neighbours tend to be different, the message passing process in GNNs might be susceptible to inferences from the neighbourhood, leading to performance deterioration. To address this issue, we find that ego networks contain an extra layer of information to further distinguish different ego nodes. Based on that, we propose a generic GNN model that can better utilize structural information of nodes' proximity to extract informative messages and resist contradictory ones from their neighbourhood. We analyze how node features and graph structure can influence the performance of GNN models. Experimental results provide insight on how our method outperforms the baselines. The presented model paves the way for incorporating ego networks' structural information into the learned graph representations, which brings GNNs with better performance and higher robustness over different datasets.


  • Open Access

    Synthetic Demographic Data Generation for Card Fraud Detection Using GANs

    by Shuo Wang, Terrence Tricco, Xianta Jiang, Charles Robertson, John Hawkin

    Using machine learning models to generate synthetic data has become common in many fields. Technology to generate synthetic transactions that can be used to detect fraud is also growing fast. Generally, this synthetic data contains only information about the transaction, such as the time, place, and amount of money. It does not usually contain the individual user's characteristics (age and gender are occasionally included). Using relatively complex synthetic demographic data may improve the complexity of transaction data features, thus improving the fraud detection performance. Benefiting from developments of machine learning, some deep learning models have potential to perform better than other well-established synthetic data generation methods, such as microsimulation. In this study, we built a deep-learning Generative Adversarial Network (GAN), called DGGAN, which will be used for demographic data generation. Our model generates samples during model training, which we found important to overcame class imbalance issues. This study can help improve the cognition of synthetic data and further explore the application of synthetic data generation in card fraud detection.

  • Open Access

    An Integral-based Approach for the Vector Potential in Smoothed Particle Magnetohydrodynamics

    by Terrence S. Tricco, Daniel J. Price

    A new implementation for the time evolution of the magnetic vector potential is obtained for smoothed particle magnetohydrodynamics by considering the induction equation in integral form. Galilean invariance is achieved through proper gauge choice. This new discretisation is tested using the Orszag-Tang MHD vortex in a 3D configuration. The corresponding conservative equations of motion are derived, but are not found to solve the MHD equations in the continuum limit. Tests are performed using a hybrid approach instead, whereby the equations of motion based on the magnetic field instead of vector potential are used. Test results experience the same numerical instability as with the Price (2010) formulation. We conclude that this new formulation is non-viable.


  • Open Access

    Sarracen: a Python package for analysis and visualization of smoothed particle hydrodynamics data

    by Andrew Harris, Terrence S. Tricco

    Sarracen is a Python package for analyzing and visualizing smoothed particle hydrodynamics (SPH) data. SPH is a method of fluid simulation that discretizes fluid elements into a collection of particles (Gingold & Monaghan, 1977; Lucy, 1977; Monaghan, 2005; Price, 2012). This approach works well for many astrophysical problems of interest, and as such there are a number of SPH codes widely used for astrophysical simulations, e.g., Wadsley et al. (2017), Price et al. (2018), Schaller et al. (2018). Sarracen offers a variety of SPH interpolation methods to aid in analysis and visualization of SPH data. It is built in Python so that users can leverage the robust scientific libraries that are available. Much of the core of Sarracen is built upon pandas and Matplotlib. Users familiar with these packages should be able to use Sarracen to do complex analyses without difficulty. Our intended use is for astrophysical SPH data, but anticipate that our package may be useful in other scientific domains.


  • The Kelvin–Helmholtz instability and smoothed particle hydrodynamics

    by Terrence S Tricco

    We perform simulations of the Kelvin–Helmholtz instability using smoothed particle hydrodynamics (SPH). The instability is studied both in the linear and strongly non-linear regimes. The smooth, well-posed initial conditions of Lecoanet et al. (2016) are used, along with an explicit Navier–Stokes viscosity and thermal conductivity to enforce the evolution in the non-linear regime. We demonstrate convergence to the reference solution using SPH. The evolution of the vortex structures and the degree of mixing, as measured by a passive scalar ‘colour’ field, match the reference solution. Tests with an initial density contrast produce the correct qualitative behaviour. The $\mathcal {L}_2$ error of the SPH calculations decreases as the resolution is increased. The primary source of error is numerical dissipation arising from artificial viscosity, and tests with reduced artificial viscosity have reduced $\mathcal {L}_2$ error. A high-order smoothing kernel is needed in order to resolve the initial velocity amplitude of the seeded mode and inhibit excitation of spurious modes. We find that standard SPH with an artificial viscosity has no difficulty in correctly modelling the Kelvin–Helmholtz instability and yields convergent solutions.


  • Open Access

    The Kelvin-Helmholtz instability and smoothed particle hydrodynamics

    by Terrence S. Tricco

    We perform simulations of the Kelvin-Helmholtz instability using smoothed particle hydrodynamics (SPH). The instability is studied both in the linear and strongly non-linear regimes. The smooth, well-posed initial conditions of Lecoanet et al. (2016) are used, along with an explicit Navier-Stokes viscosity and thermal conductivity to enforce the evolution in the non-linear regime. We demonstrate convergence to the reference solution using SPH. The evolution of the vortex structures and the degree of mixing, as measured by a passive scalar `colour' field, match the reference solution. Tests with an initial density contrast produce the correct qualitative behaviour. The L2 error of the SPH calculations decreases as the resolution is increased. The primary source of error is numerical dissipation arising from artificial viscosity, and tests with reduced artificial viscosity have reduced L2 error. A high-order smoothing kernel is needed in order to resolve the initial velocity amplitude of the seeded mode and inhibit excitation of spurious modes. We find that standard SPH with an artificial viscosity has no difficulty in correctly modelling the Kelvin-Helmholtz instability and yields convergent solutions.