-
Notifications
You must be signed in to change notification settings - Fork 7
Home
Alex Ji edited this page Aug 14, 2020
·
3 revisions
Welcome to the smhr wiki!
Link: Frequently-Asked-Questions
by Kaitlin Rasmussen, Alex Ji
This is Kaitlin's notes from a Skype session with Alex. There's some bias towards cool, carbon-rich stars in the discussion. We will fill in more details as needed.
- Begin a new file with Cmd+N
- Open an old file with Cmd+O
- Select a spectrum template with the ‘...’ button. Defaults are in ~/smhr/smh/data/spectra/
- Use HD122563 for giants, HD140283 for subgiants, and G64-12 for dwarfs. If the evolutionary status of the star is not known, choose any and go back after determining stellar parameters to redo.
- Choose a wavelength region to test. The Mg triplet is recommended (more symmetric than the Ca triplet) and requires a new region list--click ‘Edit list’, type in 5150 and 5200 into the lower/upper wavelength (A) boxes and click ‘Save to list’, then ‘Save as default’ if desired. Click ‘exit’.
- Click ‘Cross-correlate’ and then ‘correct’ or ‘cross-correlate and correct’ to perform the radial velocity correction and application.
Some standard good values
-
Function: Spline
-
Order: 3 (can be changed but Alex would stick with 3)
-
Maximum iterations: 5
-
Low sigma clip: 1.0
-
Upper sigma clip 0.2
-
Knot spacing: 20
- For larger numbers of lines in the order, increase KS
-
To zoom:
- Right-click and drag
- 2 right-clicks to un-zoom
-
To fit:
- Mask features you don’t want to include in the fit by left-clicking and dragging over them
- Adjust sigma values
- Adjust KS
- Use up/down keys to rescale the continuum (not recommended if you can adjust the sigma values)
- Double-click adds a point, but highly recommend only using for quick-and-dirty normalizations not final analyses
-
As a general rule, the fit is good when the continuum goes through the median of the noise values
-
To undo changes, click c
-
To set a flat line to an already-normalized spectrum, click d
-
To move to the next order: left/right keys
-
When satisfied with the fit of all orders, press the ‘Normalize and stitch orders’ button
- Go to the Line Measurements tab
- To add a new Fe line list (or any set of lines to measure equivalent widths):
- In the ‘edit’ tab, click ‘Transitions’ and then the ‘Add EQW List’ button at the bottom.
- To add non-Fe lines for synthesis:
- Still in the ‘Manage Measurements’ window, click the ‘Add synth list’ button and select the line list you want
- A periodic table will appear in a new tab, select the corresponding element, then click ‘Ok’. The new line will appear the bottom of the table. Cmd+a to select all lines.
- Do NOT click the Import/Export Spectral Models buttons, it's broken.
- Click the ‘Import master list’ button, then the ‘Ok’ button
- The tab directly above the table should be set to ‘All’ and will indicate how many lines you have loaded in.
- Click ‘Fit all EW’ (it can be helpful to look at the terminal output)
- A check mark in the far left column indicates a successful measurement has been made.
- The type of line should be ‘Gaussian’ for all lines but the Mg triplet (use Voigt if too saturated).
- Use caution with C-rich stars, the blue Fe lines are typically heavily blended; if they must be used, use a carefully synthesized line
- To fit Fe lines:
- Set quality control under the ‘Stellar parameters’ tab at the bottom. (‘Quality control’ button)
- For red giants, a good lower limit is 3800-4000 A. Use blue lines when dealing with a star so metal-poor that Fe II lines become scarce.
- A good upper limit is 6300 (where the telluric lines begin), or 8000, where fringing effects become noticeable (MIKE spectra). Change in the same location as above.
- Set an upper limit for REW at -4.5 to -4.7
- Click ‘Fit all EW’
- Mask unwanted lines with click+drag
- Anti-mask good areas of the continuum with Shft+click+drag
- Click+drag will undo anti-masking
- “Use anti-mask” is always greyed out but it’s checked when it’s on
- “Clear masks” to get rid of masks.
- To change size of view, enter numbers into ‘View window’ (Angstroms). Hit Enter to refresh.
- hotkeys
1
,2
,3
,4
zoom the y-axis
- Look at the Summary Plot by going to the very top under the ‘Plot’ tab. Estimate the expected temperature and metallicity.
- Examine and possibly remove bad Fe outliers. These are often unknown blends, especially in carbon and s-process-rich stars. Other times it's just bad atomic data. You can and should curate your own Fe list for every class of star (temperature, metallicity, carbon-richness, etc)
- For a first-pass set of parameters, you can click the ‘Solve’ button
- Consider only calculating the parameters with red Fe lines if possible for higher quality
- Change the temperature, logg, [M/H], and vt with the end goal of having the Fe I/II fits (red and grey lines) as flat and close together as possible. Don’t change alpha.
- [M/H] (METALLICITY) If this is far from your visual estimate, set it to that first. Re-measure with ‘Measure abundances’.
- LOGG: Set this one second; get Fe I average as close to Fe II average as possible; increase logg to increase the fit line
- Teff (EFFECTIVE TEMPERATURE): Set this one third, Increase temperature to decrease slope of Fe I fit line toward flatness
- VT (TURBULENT VELOCITY) : This one needs to increase as logg decreases; a dwarf/subdwarf can be around 1.5, an evolved giant can have as high as 3
- Alter [M/H], logg, Teff, and vt as necessary until the best fit is achieved. Individual Fe lines can be adjusted within the bottom window similarly to the previous step,
- To fix several lines at once, click, Cmd+click others to highlight, then go to the Line Measurements tab where they will also be highlighted. [APJ: I didn't realize this worked; either way in the Review tab you can set the user flag] There is the option to flag lines to look at later (especially for synthesized lines).
- Click on the line in the left hand table you want to look at
- The ‘Use poly[nomial] for cont[inuum]’ button, if clicked, is fit; otherwise if unclicked, it is set as a straight line at 1.0 (desirable if you did an excellent job normalizing).
- Similar for smoothing and radial velocity.
- Smoothing, relative to the last iteration of SMH, is now the standard deviation instead of FWHM. Multiply by 2.355 to get the ‘old’ smoothing value
- Click ‘update abundance table’ and then ‘ fit model’ (have to press fit model at least once)
- Like Fe lines, you can mask and anti-mask to adjust the fit.
- You have to set all the isotopes in the isotope tab
- Note: find Sc abundance before carbon CH
- You can set the abundance manually by clicking in the window to change and pressing synthesize
- Calculate upper limits by checking the ‘upper limit’ box below the abundance table, then clicking ‘Fit Model’ and also ‘Upper Limit’.
- Here is where you can check the by-wavelength scatter to see if there are any outliers (typically occur in blue especially in cool stars)
- Export various things under the ‘Export’ tab at the top
- Note: Wavelength is no longer linear for the whole spectrum compared to old SMH that interpolated to fixed linear dispersion