JWST Wide Survye (Discussion with M. Franx @ Berkeley Oct 2016)

From today's discussion the following proposal emerged:

What if we did a 700 sqarcmin survey of R=100 only survey of the entire CANDELS
area; 1h per setting

This yields 2000 galaxies to m_AB=24.8 for which we get continuum S/N >10/pix;
good enough to do quite good age, FeH and a/Fe.
at z>2
We also get good enough H/alpha for 2Msun/year

AS4 stellar astrophysics concept

Science scope:
  broad spectrum of ground-breaking stellar astrophysics,
 telling us how stars work, live and die...

With Gaia & TESS there are transformational space missions 
that focus on imaging/photometry/astrometry (some spectra) of bright stars.
The proposed program is tailored towards a spectroscopic survey that
in many ways ideally complements this:
 -- it is all sky
 -- it is matched in apparent magnitude
 -- it is time-domain
 -- it is dust-penetrating (disk/young stars)

Natural operational meshing of stellar astrophysics & DISCO

It is focussed on
 -- age-chemo-orbital mapping of the entire galaxy 
 -- binary stars across the HR (as calibrators, stellar evolution, SF,SN)
 -- astroseimsology to understand stellar evolution (need spectra)
 -- planet hosts 

It is a full sky program, with APOGEE and BOSS (1 each) in both hemispheres,
2/3 is multi-epoch observations; 2/3 APOGEE; 1/3 BOSS; still based on a 10min minimal cadence.

"Surprise" element:
Instead of a ROBOT, a 4-5 fold-set of 300 APOGEE fibers, feeding 4-5 gang connectors,
still using plug plates would enable:
 -- simultaneous use of 300 APOGEE and 500 BOSS fibers 
     --> better use of instrument/detectors
 -- more flexibility in fiber placement

This would mean:
-- minimal new hardware development 
   (RV improvement; fiber cart upgrade with BOSS and more APOGEE fibers)
-- move 1 BOSS spectrograph to LCO
-- plug plate costs in two hemisphere continues
-- could start 2019

All science can be done in minimally 3 years; comfortably 4 years;
minimum 25 Million x 10min x spectrum
few million fiberx10min open (for BOSS)

Much of the same science would work within fiber-robot framework,
too, in comparable survey duration.

Exercises for the cosmology block course

Though on exercises for the Cosmology Block course:

Day 1: Newtonian cosmology:
   Exercise: things to do with the collapsing top hat:
   -- explain that (in three days) we'll see that overdose parts of the Universe
can be treated as Newtonian Universes
  -- the collapsing sphere:
     -- toy-problem
     -- "virializing"
     -- properties
     -- application to the Milky Way:
        we take v_circ --> v_virial --> mass, size, etc..


Day 2: Friedman-Robertson Walker Universe
    -- code up, and plot, the angular diameter distances; luminosity distances
       as a function of cosmological parameters;
   -- what are the uncertainties implied by current cosmol. parameter uncertainties
       (which we prod ad hoc at this point)
    -- application: what's size of a galaxy at z=0.5,2,7
    [TOO Early; move all of this to day 3?]

Day 3: Linear growth of structure etc...
    -- code up the solution of contrast growth for non-trivial (i.e. realistic) cosmological parameters
    -- compare the (linear) growth of structure between toy-cases of cosmological parameters
    -- do top-hat now in a cosmological context  
    -- or code up and plot Press-Schechter

Exercises for the cosmology block course

Though on exercises for the Cosmology Block course:

Day 1: Newtonian cosmology:
   Exercise: things to do with the collapsing top hat:
   -- explain that (in three days) we'll see that overdose parts of the Universe
can be treated as Newtonian Universes
  -- the collapsing sphere:
     -- toy-problem
     -- "virializing"
     -- properties
     -- application to the Milky Way:
        we take v_circ --> v_virial --> mass, size, etc..


Day 2: Friedman-Robertson Walker Universe
    -- code up, and plot, the angular diameter distances; luminosity distances
       as a function of cosmological parameters;
   -- what are the uncertainties implied by current cosmol. parameter uncertainties
       (which we prod ad hoc at this point)
    -- application: what's size of a galaxy at z=0.5,2,7
    [TOO Early; move all of this to day 3?]

Day 3: Linear growth of structure etc...
    -- code up the solution of contrast growth for non-trivial (i.e. realistic) cosmological parameters
    -- compare the (linear) growth of structure between toy-cases of cosmological parameters
    -- have students walk themselves through the argument why CMB+LSS basically rules out    
baryonic dark matter.

Exercises for the cosmology block course

Though on exercises for the Cosmology Block course:

Day 1: Newtonian cosmology:
   Exercise: things to do with the collapsing top hat:
   -- explain that (in three days) we'll see that overdose parts of the Universe
can be treated as Newtonian Universes
  -- the collapsing sphere:
     -- toy-problem
     -- "virializing"
     -- properties
     -- application to the Milky Way:
        we take v_circ --> v_virial --> mass, size, etc..


Day 2: Friedman-Robertson Walker Universe
    -- code up, and plot, the angular diameter distances; luminosity distances
       as a function of cosmological parameters;
   -- what are the uncertainties implied by current Cosmol. parameter uncertainties
    -- application: what's size of a galaxy of 1kpc size at z=0.5,2,7

Dear All,
   after the group meeting on Wednesday, I played around just to get
a more quantitative sense of what kind of stars (at what distances,
and what velocities, [Fe/H] etc..) we should expect, I played around
with GUMS (=Gaia Universe Model Snapshot;
http://arxiv.org/pdf/1202.0132v2.pdf ).

This e-mail contains
a) a pointer to a roughly-DR1-TGAS-like mock catalog:
    https://www.dropbox.com/s/a1hiqrrdg1euqfs/GUMS_Gmag_11.5.fits?dl=0
    [provided by Jan; see http://www2.mpia.de/GC-NEW/wiki/GC/MWGroupMeeting ]
b) some plots and thoughts on the expected astrophysical
    properties of the sample members
c) if you take a however-simplistic model for the errors
    (e.g. 0.4mas, and 0.4mas/yr for TGAS<10.5). With that,
   I see no reason why you could not proceed, tune the code and make the plots
   for whatever Gaia-day1 paper you have in mind.

Details below.
HW


For technical reasons (travel web connectivity), I only downloaded the 1.5M
stars G<10.5; not the G<11.5-ish that may be most appropriate for;
Jan R. will put the full catalog on the MW@MPIA Wiki soon.
Let's look first at "all" sample members, and then at the
10^5 nearby ones, within 200pc. The catalog includes kinematics, binarity,
but is (largely) mute on white dwarfs, etc..

If we look at all 1.5M stars, we get the following distributions:




.. counts dominated by stars in the 3700k-5000K range.




.. giants and MS stars have approximately equal portions; the red clump
(at logg 2.3) sticks out a bit.




most stars are within 1kpc, but there's a LONG tail of distant stars (see below)...



at ~0.4 mas/year error, the DR1 measurement is precise and accurate for most stars




Now, let's look at the sky distribution, as a function of distance:

the most distant ones (>5kpc) are super giants in the Galactic plane


as we consider closer samples; they (of course) become increasingly
more homogeneously (yeah, a Mollweide projection would show that more nicely)







Just as a specific example of what to do with this,
here's a Galactic-top-down view of
the spatial distribution of some low-latitude tracers:
G-dwarfs & red clump stars in TGAS:



If you can pick out those two types of stars, you can make an instantaneous
map of the vertical motions of the Galactic disk (ideas on how to do this in a
subsequent communication).


Now, let's look at the very nearby stars; those may be good to calibrate
stellar physics:
There are 150.000 TGAS stars within 200pc.



their distribution is more dominated by warm/hot MS stars:

-- 

A vertical motion map of the Galactic disk with TGAS (Gaia DR1)

In a separate post, I have sketched out what "Gaia DR1 (TGAS)" contains in astrophysical terms,
here is a zoom-in on a possible Gaia-day1 project:

Idea: can we make a map of the vertical motions of the Galactic disk, using only
proper motions and parallaxes; no spectroscopic information, to get spectrophotometric distances.

1) no spectroscopy --> take low-latitude stars (|b|<5deg),
    where all the vertical motion is in the plane of the sky.

2) we need stars with good photometric distance:
       these could be
      a) RC stars (can we pick them photometrically; I think we can?);
          there should be 60.000 RC stars at |b|<5deg in TGAS, with <D>~800pc
      b) MS stars (can we pick MS stars)? the plot below shows that we can eliminate non-MS stars by their parallax measurements (of we have a parallax accuracy of ~0.5mas

 

Note:  there are only 100 G-MS stars at |b|<5deg in TGAS, with <D>~100pc; that pins down the solar velocity?

3) that would lead to the following spatial distribution of the two tracers

which should be good enough to make a vertical disk corrugation map (by simple binning of the vertical proper motions).