Upload
society-of-women-engineers
View
162
Download
0
Tags:
Embed Size (px)
DESCRIPTION
Presented by: Tabitha Voytek
Citation preview
From Dark to Light Hydrogen and the First Stars
Tabitha Voytek, Carnegie Mellon University
1
SCI-HI Project Collaborators: Jeff Peterson (CMU), Aravind Natarajan (CMU/Pitt) *Omar Lopez-Cruz (INAOE), Jose-Miguel Jauegui-Garcia (INAOE), Edgar Castillo (INAOE) *Other future collaborators in Mexico
2
Background
Instrument
Deployment
Analysis
Results
Improvements
Observations with the SCI-HI experiment focus on first star formation during the Cosmic Dawn.
Image: http://www.nasa.gov/mission_pages/planck/multimedia/pia16876b.html
SCI-HI Observations
3 Background Deployments Results
Instrument Analysis Improvements
4
Background
All 21-cm observations are made using Hyperfine splitting of the ground state of hydrogen.
1s 1S1/2
ΔE → 1420 MHz → 21-cm
5 Background Deployments Results
Instrument Analysis Improvements
Hyperfine splitting is energy differentiation caused by the spin alignment of the proton and electron.
Image from: http://aliceingalaxyland.blogspot.com/2009/04/mini-spinni-antics-of-atoms-in-space.html
6 Background Deployments Results
Instrument Analysis Improvements
Neutral hydrogen (HI) gas can be found in the Interstellar and Intergalactic Medium (IGM).
7
Hydrogen 21-cm Emission Contours for M51
Image from: http://www.haydenplanetarium.org
Background Deployments Results
Instrument Analysis Improvements
All-sky 21-cm spectrum experiments measure the average redshift evolution of HI gas in the IGM.
Image: Pritchard and Loeb, Nature, 2010
8
2/1)1)(1( zTs
initial
T
T
spaceb
Background Deployments Results
Instrument Analysis Improvements
obs
obsemitz
In particular, the SCI-HI experiment focuses on the all-sky averaged signal during the Cosmic Dawn (z~20).
Image: Pritchard and Loeb, Nature, 2010
9 Background Deployments Results
Instrument Analysis Improvements
SCI-HI Measures Here
10
Instrument
Three primary components were considered in designing the SCI-HI instrument.
Antenna Working Frequencies: 40-130 MHz
RF Electronics Switch, amplifiers, filters and cables
Computer Analog to Digital Conversion and Fourier Transform
11 Background Deployments Results
Instrument Analysis Improvements
The SCI-HI antenna went through several different iterations before settling on the HIbiscus design.
Trombone Antenna HIbiscus Antenna v1 HIbiscus Antenna v2
12 Background Deployments Results
Instrument Analysis Improvements
Data was collected with the preliminary SCI-HI instrument in June 2013 on Isla Guadalupe.
13 Background Deployments Results
Instrument Analysis Improvements
At the time, the instrument had a single antenna and relied on 12V car batteries for DC power.
14 Background Deployments Results
Instrument Analysis Improvements
15
Deployment
Terrestrial radio frequency interference (RFI) is a significant problem in radio astronomy.
16 Background Deployments Results
Instrument Analysis Improvements
We evaluated current and potential experiment sites for their RFI environmental quality.
17 Background Deployments Results
Instrument Analysis Improvements
At each site, quality was investigated using a portable antenna and spectrum analyzer.
18
Algonquin Guadalupe
Background Deployments Results
Instrument Analysis Improvements
We chose to deploy SCI-HI at Isla Guadalupe based upon the excellent RFI environment.
19
FM band as measured with the SCI-HI instrument
Background Deployments Results
Instrument Analysis Improvements
20
Analysis
Data collection includes both the sky signal and calibration source signals.
21 Background Deployments Results
Instrument Analysis Improvements
Sky signals collected with the HIbiscus antenna are dominated by the Milky Way Galaxy.
22 Background Deployments Results
Instrument Analysis Improvements
This Milky Way Galaxy signal varies over each day due to the antenna beam sky coverage.
Daily Signal Variation at 70 MHz
23 Background Deployments Results
Instrument Analysis Improvements
Calibration utilizes the Milky Way Galaxy foreground model and daily mean/variance.
Beam Averaged Sky Temperature:
𝑇𝐺𝑆𝑀 𝑡, 𝜐 = 𝑑Ω 𝐺𝑆𝑀 𝜃, 𝜙, 𝜐 𝐵(𝜃 − 𝜃0 𝑡 , 𝜙 − 𝜙0 𝑡 , 𝜐)
𝑑Ω 𝐵(𝜃 − 𝜃0 𝑡 , 𝜙 − 𝜙0 𝑡 , 𝜐)
GSM Model comes from de Oliveira-Costa et al. (2008)
χ² Fitting:
𝜒2 𝜐 = Δ𝑇𝑚𝑒𝑎𝑠 𝑡, 𝜐 − Δ𝑇𝐺𝑆𝑀 𝑡, 𝜐2
𝑡
Daily Mean Subtraction:
Δ𝑇𝑚𝑒𝑎𝑠 𝑡, 𝜐 = 𝑇𝑚𝑒𝑎𝑠 𝑡, 𝜈 − 𝑇𝑚𝑒𝑎𝑠 𝐷𝐴𝑌(𝜐)
Δ𝑇𝐺𝑆𝑀 𝑡, 𝜐 = 𝑇𝐺𝑆𝑀 𝑡, 𝜈 − 𝑇𝐺𝑆𝑀 𝐷𝐴𝑌(𝜐)
24 Background Deployments Results
Instrument Analysis Improvements
KΔGSM calibration is most successful when a full day of data is collected.
25 Background Deployments Results
Instrument Analysis Improvements
Example KΔGSM Mean and Fit
After calibration, foreground removal was done using a simple polynomial fitting and subtraction.
26
𝑙𝑜𝑔10 𝑇𝐺𝑀(𝜐) = 𝑎𝑘 𝑙𝑜𝑔10(𝜐
70 𝑀𝐻𝑧)𝑘
𝑛
𝑘=0
𝑇𝑟𝑒𝑠 𝜐 = 𝑇𝑚𝑒𝑎𝑠 𝐷𝐴𝑌 𝜐 − 𝑇𝐺𝑀(𝜐)
Daily Fit:
Daily Residuals:
Background Deployments Results
Instrument Analysis Improvements
27
Results
Preliminary data collection in June 2013 gave good results, but further improvements are necessary.
KJNC Calibrated
KΔGSM Calibrated
Predicted 21-cm Signal Models
28 Background Deployments Results
Instrument Analysis Improvements
29
Improvements
Problem: DC battery is unreliable. Solution: Replace battery with an AC gas generator.
30 Background Deployments Results
Instrument Analysis Improvements
Problem: DC battery is unreliable. Solution: Replace battery with an AC gas generator.
31 Background Deployments Results
Instrument Analysis Improvements
Problem: Antenna frequency range is too small. Solution: Build multiple scaled antennae.
32
70 MHz Scale
100 MHz Scale
Background Deployments Results
Instrument Analysis Improvements
Problem: Self-generated RFI seen in data. Solution: Double Faraday cage for computer.
33
Double Faraday Cage
Computer inside inner Faraday Cage
Background Deployments Results
Instrument Analysis Improvements
Problem: Isla Guadalupe still not remote enough. Solution: Find more remote sites.
34
Time Series of the FM band for June 1st, 2013 data
Time Average of the FM band for June 1st, 2013
Background Deployments Results
Instrument Analysis Improvements
Problem: Isla Guadalupe still not remote enough. Solution: Isla Socorro/Clarion and/or Marion Island.
35
Guadalupe
Socorro Clarion
Marion Island
Background Deployments Results
Instrument Analysis Improvements
With these changes, the SCI-HI experiment will help us better understand the first stars.
Foregrounds
Current Best Results (KΔGSM)
Predicted 21-cm Models
36 Background Deployments Results
Instrument Analysis Improvements
S C I – H I Sonda Cosmológica de las Islas para la Detección de Hidrógeno Neutro
37