1 of 39

HectoSPEC

HectoCHELLE

Positioner

Lives of the Instruments: Hectospec & Hectochelle at the MMT 2003-2026

2 of 39

Health Warning:

Ancient photographs may cause discomfort

3 of 39

At the MMT, the fiber positioner was designed

to feed one of two distinct bench

spectrographs: Hectospec and

Hectochelle.

4 of 39

5 of 39

Builders

6 of 39

Twilight sky,

Oct 19, 2003

7 of 39

First target,

Oct 24, 2003

8 of 39

Prior to the arrival of Hectospec, the MMT instruments such as Blue Channel Spectrograph had been

scheduled and used in a “classical” way. Each PI got specific assigned nights; if the weather was bad,

“try again next time.”

In 2004, Dan decided that data taking for this instrument would be run as a queue, to maximum efficiency.

There was some initial (~2004) resistance to this based on people’s experience at other places (NOAO)

at the time, so we chose to do two things differently from other queues at the time.

A) Try to get everyone something early on, no matter how bad the weather 

B) Try to ensure that everyone got the same completion ratio at the end of the period

(first as a trimester, then semester). Thus the scheduling process itself did not favor one project

over another, except for the natural effects of position and moon tolerance. 

It was always possible for the TAC to simply give some projects more time,

Operation

9 of 39

Prior to the arrival of Hectospec, the MMT instruments such as Blue Channel Spectrograph had been

scheduled and used in a “classical” way. Each PI got specific assigned nights; if the weather was bad,

“try again next time.”

In 2004, Dan decided that data taking for this instrument would be run as a queue, to maximum efficiency.

There was some initial (~2004) resistance to this based on people’s experience at other places (NOAO)

at the time, so we chose to do two things differently from other queues at the time.

A) Try to get everyone something early on, no matter how bad the weather 

B) Try to ensure that everyone got the same completion ratio at the end of the period

(first as a trimester, then semester). Thus the scheduling process itself did not favor one project

over another, except for the natural effects of position and moon tolerance. 

It was always possible for the TAC to simply give some projects more time,

After some concern, that scheme was implemented and run for the next 23 years.

Operation

10 of 39

Sample coding for a column of the spreadsheet scheduler:

=IF(ISTEXT(F3);DEGREES(ACOS(  (COS(RADIANS(24*$R3))*COS(RADIANS((HOUR($Q3)+MINUTE($Q3)/60)*15))*COS(RADIANS((HOUR($A16)+

MINUTE($A16)/60)*15))*COS(RADIANS(24*$A17)) + COS(RADIANS(24*$R3))*SIN(RADIANS((HOUR($Q3)+MINUTE($Q3)/60)*15))*SIN(RADIANS((HOUR($A16)+

MINUTE($A16)/60)*15))*COS(RADIANS(24*$A17)) +SIN(RADIANS(24*$R3))*SIN(RADIANS(24*$A17))) /   SQRT((COS(RADIANS(24*$R3))*COS(RADIANS((HOUR($Q3)+

MINUTE($Q3)/60)*15)))^2+(COS(RADIANS(24*$R3))*SIN(RADIANS((HOUR($Q3)+MINUTE($Q3)/60)*15)))^2+SIN(RADIANS(24*$R3))^2) / SQRT((COS(RADIANS(24*$A17))*

COS(RADIANS((HOUR($A16)+MINUTE($A16)/60)*15)))^2+(COS(RADIANS(24*$A17))*SIN(RADIANS((HOUR($A16)+MINUTE($A16)/60)*15)))^2+SIN(RADIANS(24*$A17))^2)  ));" ")

Spreadsheet scheduler for nightly observations,

made by hand each night.

11 of 39

Better ?

TCL scheduler, semi-automatic,

Each night was prepared in the afternoon,

but operators used the schedule only as a guide

Names of queue

observers

Perry Berlind

Mike Calkins

Kali Salmas

Ryan Howie

Skyler Self

ShiAnne Kattner…

Telescope Operators:

Erin Martin

Ben Kunk

Alejandra Milone

Michael Alegria

John McAfee

Brian Pinault

Scott Gotilla

12 of 39

2003

2026

1996

Operators, Fixers

13 of 39

We also initiated the idea that the SAO TDC would reduce the data quickly, and make that available to the PIs. A scheme now considered standard for monolithic projects, but we eventually did it for small projects as well as insuring that scientists didn’t have to waste time learning yet another reduction package.

14 of 39

We also initiated the idea that the SAO TDC would reduce the data quickly, and make that available to the PIs. A scheme now considered standard for monolithic projects, but we eventually did it for small projects as well as insuring that scientists didn’t have to waste time learning yet another reduction package.

Knowledge from the operation of Hecto and the running of its queue, to pass onto future generations:

  1. When the weather is good, everyone is happy with a queue. When it is bad, no one is happy.
  2. ~90% of the science can be done with ~70% of the requested data.
  3. All data is reduced at least three times.
  4. So, it helps to think a lot about naming conventions, for projects, fields and targets.

15 of 39

And the winner is,

16 of 39

Total nights reported from 2004-2026 (not necessarily scheduled):

951 for HectoSPEC

825 for HectoCHELLE

That’s about 15% of the total nights for the active period (11% for Chelle). The most active years used about 1/3 of the time.

Total program completion  - hard to say. The worst in a trimester was about 30% , the best was 85%. Typical was probably 70%.

Total spectra at the end is

866,114 object spectra for Hectospec,

464,303 for Hectochelle (longer exposures).

17 of 39

18 of 39

19 of 39

20 of 39

Spec 552 papers, 33099 citations

Favorite topics were

AGN studies (Hickox, Stern, Kochanek, Eisenstein, Murray)

Large Scale Structure and Galaxy Clusters (Geller, Rines, Egami, Sohn, Hwang)

Dwarf Galaxies around larger ones (Weiner, Park, Fang)

Massive stars in nearby galaxies (Massey, Humphreys, Sokal, Drout)

Young Stellar Objects (Kim, Allen, Espaillat, Sicilia-Aguilar, Winston)

Star clusters, integrated and resolved, young and old (Caldwell, M.G. Lee, Chandar, Wright)

Chelle 117 papers, 7481 citations

Topics were:

Individual stars in stellar systems: dwarf galaxies and clusters (Furesz, Olszewski, Walker, Caldwell, Strader, C.Johnson, Meibom)

MW halo structure, streams(Conroy “H3”; Bonaca; Caldwell)

Young Stellar Objects (Allen, Kim, Lim, Espaillat, Hsu, Covey, Curtis)

21 of 39

Highlights of well-cited papers:

Spec (low res)

 Mid-Infrared Selection of Active Galaxies, by Stern et al 2005

“find evidence for a large population of optically obscured active galaxies”.

Host Galaxies, Clustering, Eddington Ratios, and Evolution of Radio,

X-Ray, and Infrared-Selected AGNs" Hickox et al 2009

propose a “model of AGN and galaxy evolution, whereby a ‘quasar’ phase and the

growth of the stellar bulge occurs when a galaxy's dark matter halo reaches a critical mass”

"Black Hole Masses and Eddington Ratios at 0.3 < z < 4"  Kollmeier et al 2006

“Fueling rates in luminous AGNs are ultimately determined by BH self-regulation of the accretion flow”

"Low-Resolution Spectral Templates for Active Galactic Nuclei and Galaxies

from 0.03 to 30 μm" Asset et al. 2008 –that’s it.

"The Universal Relation of Galactic Chemical Evolution: The Origin of the

Mass-Metallicity Relation" Zahid et al. 2014

“Turnover in the mass–metallicity relation occurs when

the gas-phase oxygen abundance is high enough that the amount of oxygen locked up in low-mass stars is an

appreciable fraction of the amount of oxygen produced by massive stars”

Chelle (high res):

(see Andy S and Matt W’s talks)

22 of 39

Personal Favorites

23 of 39

The Virgo Cluster,

Strader et al.

Peng et al.

M.G. Lee et al.

(more than 1000 GCs)

24 of 39

25 of 39

HGVC-1, aka ”Strader 1”

Vel = -1026 km/s

(note the minus sign)

Globular Clusters in M87 and the Virgo Cluster

26 of 39

Projected distance to M87 is 85 kpc

Caldwell, Strader et al. 2014

JWST time allocated to determine location in Virgo !

Ejected to high velocity via a 3-body encounter with a binary BH ?

27 of 39

M31, The Andromeda Galaxy

28 of 39

29 of 39

M31, The Andromeda Galaxy

30 of 39

M> 13,000 objects of various types with Hectospectra

Caldwell

Humphreys

Huchra

Lada

Massey

MG Lee

Neugent

31 of 39

Young star clusters,

Color represents radial velocity

(red is 0 km/s, blue -550 km/s)

(Caldwell et al, 2007-2016)

Lots of rotation, disk-like

32 of 39

Old star clusters (age > 8 Gyr)

Just a little rotation, mostly in the metal rich group

Some rotation, more than MW clusters

33 of 39

Mg b lines

Unusual H3

Hectochellle Stellar

Spectra

Conroy et al,

2019

34 of 39

(Building up to an anecdote)

Gaseous

Nebulae

in M31

35 of 39

M31 Supernova Remnants

36 of 39

Perhaps similar to Eta Carinae –

high velocity Nitrogen

enriched Circumstellar Material

excited by supernova shocks.

~3000 years old

Caldwell and Raymond, 2024

Ratio of ionizing

Flux to density in a HIIr

Is such that most S is doubly

ionized (hot star), so SII is

Low.

37 of 39

All reduced Spec data in this archive

https://oirsa.cfa.harvard.edu/

Chelle archive coming.

(recall item 3 above about number

of times data needs reducing)

38 of 39

Last Image

March 16, 2026

39 of 39

Last Image

March 16, 2026

Thanks for the memories,

Of little specks of light,

Shining dimly in the night

Those millions of spectra you gave us, with just a little fight.

How lovely it was.