NCTS-TCA Summer Student Program 2026
理論計算天文暑期學校

Recent combined analyses of Planck, DESI, and Type Ia supernovae suggest that a dynamical dark energy model may fit current data better than ΛCDM. If dark energy evolves with time, it modifies the cosmic expansion history and therefore the inferred distance scale.
In this project, we will perform a stringent internal consistency test using the Baryon Acoustic Oscillation (BAO) standard ruler. Assuming the best-fit dynamical dark energy model from DESI, the student will examine whether the BAO peak location remains self-consistent across redshift — i.e., whether the inferred BAO scaling parameter is consistent with unity when distances are computed within the model.
The student will:
* Recompute cosmological distances under ΛCDM and dynamical dark energy models
* Evaluate the BAO scaling parameter as a function of redshift
* Quantify any residual shifts and assess their implications
This project directly probes whether current hints of evolving dark energy represent a physically consistent cosmological model.

Pre-requisites: Strong programming skills in python
Keywords: Cosmology, Dark Energy, Large Scale Structure
Location: ASIAA/NTU or remote

The discovery of supermassive black holes in the early Universe challenges existing black hole formation models. Ultradense solitonic cores predicted in fuzzy dark matter offer a potential solution to this timing problem. Building on the proof-of-concept study of Chiu et al. (PRL 2025), this project will investigate whether the proposed soliton-driven accretion-boosting mechanism remains effective when key baryonic processes, such as turbulence, radiative cooling, and active galactic nucleus feedback, are included self-consistently. We will conduct three-dimensional hydrodynamic simulations using the multi-GPU-accelerated adaptive mesh refinement code GAMER.
This project has strong potential to lead to publication in a top-tier journal. However, achieving this goal will require sustained effort beyond the two-month summer program. Students who are highly motivated and interested in continuing beyond the summer are especially encouraged to apply; additional details will be discussed during the interview.

Pre-requisites: Calculus; Basic knowledge of C and Python
Keywords: Dark Matter, Supermassive Black Holes, Numerical Simulations
Location: NTU or remote

The accretion process onto black holes is one of the most powerful phenomena in the universe. Black hole systems often exhibit jets and outflows. To understand how these jets are ejected from black holes, we need to study the accretion process in detail. Over the years, various accretion disk models have been proposed. Recently, a significant accretion state known as the "magnetically arrested disk (MAD)" has gained attention. Observations of M87 and Sgr A* by the Event Horizon Telescope (EHT) collaboration have underscored the significance of the MAD state. It has been suggested that the accretion flow around these two black holes, which are active galactic nuclei (AGNs), is in the MAD state, based on comparisons between radio images and post-processed General Relativistic Magnetohydrodynamic (GRMHD) simulations. In this project, the student will carry out simulations using publicly available PLUTO code.
The objectives of this project are:
1. To run the simulation code, analyze the data, and interpret the results.
2. To understand the accretion process around black holes.
3. To explore the possible applications of the simulation results to black hole sources.
4. Successful completion of the project will lead to publication.

Pre-requisites: Understanding basic black hole physics, accretion flows, fluid dynamics, magnetohydrodynamics, and basic programming (e.g., Python, C, C++).
Keywords: Accretion; Black hole physics; Magnetohydrodynamics; Radiative magnetohydrodynamics; Relativistic jets; Supermassive black holes
Location: ASIAA/NTU

Recent JWST observations have captured highly structured, multiphase galactic winds in unprecedented detail. Yet, the presence of cold, dusty gas within these high-velocity outflows remains a puzzle. The observed near-constant dust-to-gas ratio contradicts current theoretical models, which suggest that dust should be rapidly destroyed by thermal sputtering. To address this, this project zooms in on the cloud-wind interface to uncover how dust survives in turbulent mixing layers. Utilizing the unique power of the hybrid CPU-GPU code GAMER, we will run high-resolution hydrodynamical simulations coupled with dust sputtering to resolve the entire turbulent inertial range all the way down to the conduction scale. This project offers students the chance to tackle a frontier research problem and gain crucial insights into the resilience of dust in galactic winds.

Pre-requisites: General astronomy; basic knowledge of Python, C, and Linux
Keywords: Hydrodynamical simulations; turbulent mixing; dust evolution
Location: NTU

Task Description and Goals
We develop hydrodynamic and magnetohydrodynamic codes and solvers for astrophysical problems in the ASIAA CompAS project. We invite students to explore and validate state-of-the-art numerical codes and solvers based on fundamental numerical methods and physics that are currently being developed, commissioned, and deployed. These codes and solvers range from Newtonian to General Relativistic regimes, covering young stars to black holes. The students will gain hands-on experience testing the accuracy and performance of numerical methods for HD, MHD, and particle problems, using them to validate and benchmark their results.
We seek highly motivated students interested in numerical methods, code development techniques, scientific verification, potential applications, and the exploration and use of machine learning/AI approaches. Actual projects will be assigned commensurate with students' academic background, preparation, and readiness. The experiences gained will be highly valuable for future career development in science, engineering, computing, astronomy, and astrophysics.
Student Opportunity: Explorations of Numerical Codes and Solvers in Astrophysical Systems
The Computational Astrophysical Sciences (CompAS) group invites applications for an exciting student opportunity to explore and validate cutting-edge numerical codes and solvers designed for astrophysical systems. This program offers hands-on experience in computational astrophysics, focusing on hydrodynamics (HD), magnetohydrodynamics (MHD), and particle-based simulations, spanning Newtonian to General Relativistic regimes.
You will
Work with state-of-the-art numerical methods and physics-based solvers under active development.
Engage in projects covering astrophysical phenomena such as young stars, black holes, and plasma physics.
Test and benchmark the performance of advanced numerical methods for HD, MHD, and particle problems.
Explore emerging fields such as machine learning (ML) and artificial intelligence (AI) applications in computational astrophysics.
Tailored project assignments based on academic background and readiness.
You will
Gain valuable numerical methods, code development, and scientific verification skills.
Build a strong foundation for science, engineering, computing, or astrophysics research careers. Collaborate with leading scientists at ASIAA.

Pre-requisites: Fourth year in college or above; students with advanced mathematics and numerical skills, Master's or beginning PhD level.
Keywords: Computational fluid dynamics, computational astrophysics, Numerical Astrophysics, numerical algorithms, code development, micro-physics, subgrid models, AI, Machine Learning
Location: NTU campus/remote

Task Description and Goals
The origins of planetary systems like our own have intrigued generations to investigate the underlying astrophysical mechanisms of formation. With the recent advances in observational facilities and high-performance computing capabilities, we enter an unprecedented era leaping toward a better understanding of how sun-like stars form in high spatial and temporal resolution.
We study young stellar objects and their formation processes using theoretical, computational, and observational methods--where theory meets observation. We develop state-of-the-art simulation codes and use high-performance computing techniques to model complex astrophysical processes. We develop diagnostic approaches in hydrodynamics, magnetohydrodynamics, radiative transfer, and chemical evolution modeling of the intricate dynamical evolution. In particular, we focus on the active physical processes operating simultaneously with the ongoing formation and evolution of the envelopes, protoplanetary disks, and powerful jets and outflows occurring in the earliest phases of the lifetimes of sun-like stars.
For more details, please visit our group site: CHARMS Group Page. We look for highly motivated students interested in foundational physics and astrophysics, numerical methods, computational and analytical tools, code development techniques, radiative diagnostics, and their applications to star-formation problems that are under active investigation and development. Depending on the participating students' background knowledge and experience, various hands-on experiences in code development, numerical simulation, modeling, or observational diagnostic comparisons could be arranged. Starting in 2025, we will incorporate machine learning and AI-accelerated automation into our development pipelines.
Are you fascinated by the birth of stars and planetary systems? Join us in unraveling the mysteries of stellar formation using cutting-edge computational methods, theories, and observational techniques. You will be exploring the frontiers of astrophysical research from multidisciplinary perspectives.

Pre-requisites: Good working knowledge of programming languages, in Python or C/C++ Strong college-level physics, chemistry, applied mathematics, or numerical methods background is a prerequisite. Fluency in English--reading, writing, and oral communication--is required. Knowledge and usage of interactive visualization software and tools Knowledge of Machine Learning or AI Techniques Passion, diligence, and curiosity.
Keywords: Star and Planet Formation, Origins of Solar System, Dust, Meteorites, JWST, ALMA, Collapse, accretion, outflows, winds, jets, disks, pseudodisks, envelopes, magnetic field, gravity, turbulence, radiation, chemistry, astrochemistry, Infrared Spectroscopy, Molecular Lines, Machine Learning, Artificial Intelligence
Location: NTU campus/Remote

Stars are often born in clusters as they condense out from the molecular clouds. These stars are by no means stationary. Previous research showed that the complex interaction between the stars crucially determines their mass distributions and their multiplicity (e.g. binary pairs). One important aspect revealed by observations is that the most massive stars tend to lie at the centre of the clusters, whereas the lower mass stars tend to lie at the edge - a phenomenon known as "mass segregation". There is an ongoing debate about whether the massive stars have formed at the centre at birth (Bonnell & Davies 1998), or if they've formed somewhere else but migrated towards the centre via N-body interaction (Guszejnov et al. 2022). This problem is of significant importance to the theory of star formation.
This project will involve using (or developing) a simple N-body code to simulate stellar dynamics in clusters. The goal is to investigate the shortest possible time for stars in clusters to organise themselves, from which we could infer the likelihood of dynamical mass segregation during the early stages of stellar evolution.
References:
Bonnell I. A., Davies M. B., 1998, MNRAS, 295, 691
Guszejnov D., Markey C., Offner S. S. R., et al., 2022, MNRAS, 515, 167

Pre-requisites: Programming skills are required. Fortran and/or Python are preferred.
Keywords: Stellar clusters, N-body simulation, numerical simulation
Location: NTU or NTNU

Massive stars born in Giant Molecular Clouds (GMCs) are the largest sources of energy and momentum in the interstellar medium (ISM). Throughout their lifetime, they release energy in the form of stellar winds, ionizing radiation, jets and supernovae, which we collectively call "stellar feedback". In particular, the ionizing radiation can heat up the ISM to 10^4 K and significantly impact the star formation activity in the cloud (e.g. Walch et al. 2012, Dale et al. 2007). It was also found in previous research that the shape of these feedback bubbles largely determines their ability to perturb the ISM.
The goal of this project is to characterise the morphologies (shapes) and structures of the ionized regions around the massive stars. The student will be provided with the datasets produced from hydrodynamical simulations of GMCs. Your job is to analyse the feedback-impacted regions, and possibly come up with methods to quantify their morphologies.
References:
Dale J. A., Bonnell I. A., Whitworth A. P., 2007, MNRAS, 375, 1291
Walch S. K. Whitworth A. P., Bisbas T., et al., 2012, MNRAS, 427, 625

Pre-requisites: Python programming skills are required.
Keywords: Molecular clouds, stellar feedback, photoionization, numerical simulation
Location: NTU or NTNU

Protoplanetary disks are the birthplaces of planets and organic material. Their vertical structure regulates how radiation, dust, and chemistry interact to shape the environments in which planets form. In edge-on disks, this vertical structure can be directly observed, providing a unique opportunity to study how material is processed as a function of height above the disk midplane. The project sits at the intersection of astrochemistry, radiative transfer, and planet-forming disks and will directly contribute to ongoing research aimed at linking disk physical structure to the chemical processing of organic material. This project will investigate the emission of dust irradiation and disk surface chemistry using spatially resolved infrared observations from the James Webb Space Telescope (JWST). The student will combine these observations with two- dimensional radiative transfer models and Bayesian analysis to retrieve physical properties of these disks.

Pre-requisites: Python knowledge, Physics/Astronomy Undergraduate
Keywords: Astrochemistry, Radiative transfer, Planet-forming, Dust, Radiative transfer, and Bayesian analysis
Location: NTHU/remote

Young star clusters eject massive stars of spectral types O and B at considerable speeds (several tens to hundreds km/s). These stars, which are called runaway stars, are produced in close dynamical encounters containing three and more stars. Previous theoretical studies indicate that most runaway stars are produced in encounters with a single binary star at the cluster centre. Independently, it was shown that scattering experiments on binary stars tend to eject stars preferentially in the plane of the binary orbit instead of isotropically. This implies that runaway stars might be preferentially ejected along the orbital plane of the most massive binary star in the cluster. Concentration of runaway stars within a planar structure might explain some of the peculiarities reported by the Gaia mission.
The student will test the possible correlation between escape directions of runaway stars produced inside a particular cluster and the orbital plane of the most massive binary star inside the cluster. The study will be performed on numerical models of open star clusters, which are already available (they were calculated for the purpose of another project).

Pre-requisites: Undergraduate students in physics (astronomy is an advantage, but not necesity), basics in python are welcome
Keywords: Star clusters, runaway stars, binary stars, stellar dynamics
Location: NTNU or remote

Open star clusters are dense systems where stars gravitationally interact with each other. Interactions between massive stars can be particularly strong, often causing some of the massive stars to be ejected from the cluster at speeds of tens to hundreds km/s. Both observations and numerical studies demonstrate that up to 40% of the ejected stars have a companion; they are actually binary stars. Since the ejection from the cluster was caused by a hard interaction, it is possible that the orbital plane of the ejected stars preserves its orientation relative to the cluster; in other words, all the binaries ejected from a particular cluster might be similarly oriented.
Based on present numerical models, the student will measure the orientation of the orbital planes of escaping stars relative to the direction to the cluster which ejected the stars. Then, the student will discuss possible relationship between these two quantities.

Pre-requisites: Undergraduate students in physics (astronomy is an advantage, but not necesity), basics in python are welcome
Keywords: Star clusters, runaway stars, binary stars, stellar dynamics
Location: NTNU or remote

Mars was once a blue world with a thick atmosphere, but today it is a cold, red desert. Where did the air go? The answer is hidden in the Martian ionosphere—a high-altitude chemical laboratory where solar radiation transforms gas into reactive plasma.
In this project, you will act as a planetary detective solving the diffusion equations to simulate the complex "chemical engine" driving atmospheric escape. You’ll track how ions like O2+ and CO2+ react under the influence of mysterious crustal magnetic fields and solar wind forcing.
By comparing your models to real data from the spacecraft data, you will help decode how chemistry dictates the life and death of a planet. Join us to explore the edge of space and solve the mystery of the Red Planet’s evolution!

Pre-requisites: English/Reading Skill, Programming
Keywords: Mars, Atmospheric Evolution
Location: NTNU