CPLUOS Astrophysics team meeting

Friday 22 Jul 2022, 14:00 → 15:10 Asia/Seoul

물리학과

https://uos-ac-kr.zoom.us/j/8264902652

14:00 → 14:50 COSMOLOGY

14:00 Hyeonmo's Report

Speaker: Hyeonmo Koo (University of Seoul)

Brief_Introduction_of_the_calculation_inside_GADGET2_simulation.pptx

Brief Introduction of the Calculation inside GADGET2 Simulation & Plotting

• Pynbody calculates the density ρ_i and smoothing length h_i using the position data made by GADGET2 and N_{sph (default = 32)}.
• Smoothing Length h_i would be decided from: N (r ≤ 2h_i) = N_sph
• Number of Particles inside r ≤ 2h
  • Default N_sph = 32.0
• |ρ_Calculated - ρ_Data|
  • 1e10 ~ 2e10
    • Effect of the outside of Kernel Function W(r, h)
• Number of Particles per each halo = 20000 → 50 Collision Samples (Each has identical initial condition) → Superposition → Number of Particles per each halo = 1000000
• Qty: What quantity we want to project to the grid? 'rho'
• n_x, n_y, n_z: The number of Grid '400(Same as PyUltraLight)'
• x₂: Half of the Boxsize '34'
• Kernel: 3D spline kernel
• Smooth: Calculated from pynbody 'smooth'
• For the position of particles (x_i, y_i)_{i = 1 ~ 10⁶...}
  • 1. Δx > 2h_i: Count one grid per each particle.
  • 2. Δx < 2h_i: Count all grids inside the smoothing domain.
  • Fortunately, the number of h_i larger than Δx/2 is about one fourth of total points.
• The result we are going to plot is as follows.
  • Result [n_{i, xpos}, n_{i, ypos}, n_{i, zpos}]+ = m_i X [Kernel Function]
• [kpc] graph comparison
  • N_sph = 32
  • N_sph = 1500

14:10 Young's report. FoF and MST algorithm

Speaker: Young Ju

FoF and MST algorithm

• Change 'sea level': Change minimum member of cluster (50 → 300).
  • Linking-length - Number of Clusters graph at Noise 50000
    • MGS has high kink.
• MGS Voronoi & MGS original → Calculate density → Galaxy Id - Density graph
• Used same galaxy.

14:20 Hannah's Report

See the slides from 8p

Speaker: Hannah Jhee (University of Seoul)

Axionyx: CDM Mixed Simulation

• No progress

14:30 Sumi's report

Speaker: Sumi Kim (University of Seoul)

Plan for IAU Poster: Cosmological Information from Higher Order Clustering Statistics

• Abstract
  • In this work, we analyze mock large scale structure data to estimate the cosmological information using higher order statistics. The mock data is generated via the Lagrangian perturbation theory-based code, PINOCCHIO
  • From a suite of dark matter halo light-cone catalogues with various cosmological parameters we measured various N-point correlation functions and trained a decision tree to predict cosmological parameters. In this work we probe up to the 4-point correlation function using the GRAMSCI (GRAph Made Statistics for Cosmological Information (Sabiu. C. G. 2019, ApJS, 242, 2)) code. We compare the cosmological information content of the 2-, 3-, and 4-point correlation functions and finally combine all the results to obtain the tightest constraints on the cosmological parameters.
• Result
  • To conclude how much information did higher order clustering statistics get from galaxy distribution, we compared the contours from output with 2 - point and 3 - point correlation function estimation results, in below comparison plot, all three plots the 2pcf contour has the biggest estimation range. This means 2 - point correlation function, which is relatively low order statistics, obtatins fewer information than the others. For contrast with lower statistics, we added contour shows much constrained region than only 2pcf estimated, suggests these higher order correlation function provides more information about cosmological parameters. Finally, we draw combined results of 2pcf, 3pcf, and 4pcf in parameter space to get most constrained contour and compare it with 2pcf to see how much higher statistics affected. In first and second plot, the combined result shows much reduced region compared with 2pcf, similar extent with higher statistics combined. But in third plot showing n_s, and h parameter, it's hard to tell that higher combined statistics are significantly constrained than 2pcf result. It's because these two parameters are not sensitive to the galaxy distribution, as written above. Through this method and result, we found out how higher order clustering statistics affect well in obtaining cosmological information. Further, we're expecting use DESI data for this work to get better outcome.
• Ω_m: -34.88%
• σ₈: -8.7%
• w₀: -54.7%
• w_a: -38.45%
• n_s: 8.15%
• h: -5.76%

14:40 Dr.Sabiu's Report

Speaker: Dr Cristiano Sabiu (University of Seoul)

Deep Learning the Deep Sky: Recovering Low Surface Brightness Objects with Machine Learning

• No progress

14:50 → 15:00 Discussion