Button to scroll to the top of the page.

Hints for Writing Successful Observing Proposals

From: Tom Barnes & Chris Sneden
To: Graduate Students, Faculty, Research Staff
Subject: Dissertation Research with McDonald Facilities; Hints for Writing Successful Observing Proposals 

Some questions have arisen about the McDonald TAC policies for observing proposals that are related to graduate students' PhD research. It has been some time since these policies have been broadcast widely. Now is also a good occasion for all of us to be reminded of how to write proposals to which the TAC will respond well. 


First, here is the relevant sentence in the description of TAC procedures: 

"Members are asked to recognize the importance of graduate education at UT Austin and other institutions by adjusting rankings for dissertation proposals as appropriate (usually done in the TAC meeting)."

The TAC will be reminded of this and other scheduling considerations when considering each round of proposals. 

We would like to re-emphasize that McDonald wants to encourage student research! But proposals from students, like anyone else, should conform to common sense. For a rumored historical (bad) example, how about this one: a prominent extragalactic researcher is alleged to have requested the entire amount of KPNO 4m dark time for a year when that instrument was newly commissioned. Get the point? Be realistic. 

To assist students in assessing the realism in dissertation observing time requirements, first and foremost the students' PhD committees should be consulted. But the McDonald Director and Associate Director will also be happy to discuss students' observing needs, and can often make good guesses as to the likelihood that those needs can be met within other scheduling constraints. 


Getting observing time at McDonald has not been much of a problem recently, but as the HET ramps up its operations we expect a higher competition level to emerge. Thanks to Inese Ivans, who surfed the web for proposal guidelines from major observatories around the world, we offer below three sets of hints that might be of use to you. Notice the general agreement among the different guideline statements, and pay heed. 

Here are the web addresses for various observing proposal guidelines:


And below are parts of three of the most helpful suggestions: 


Historically, the TAC has focused on a few major areas of concern, virtually independent of changes in committee membership. The first and foremost of these is, of course, scientific merit. Because the committee is rather small and must cover a wide range of astrophysical expertise, it is essential that proposals make very clear just what scientific questions will be addressed by the proposed observations. The committee often wishes to see these goals placed in a larger context; i.e., how these questions relate to major unresolved astrophysical questions.

Vague generalities such as "increasing our understanding" is not sufficient; the committee looks for specific questions that will be unambiguously addressed by the observations. Another issue of concern is the description of a clear path from the taking of the data through the reduction and analysis that will permit answers to be obtained. Other questions that often arise are whether or not similar programs have been done or are being done elsewhere, and if previous observing programs have resulted in the data being published in a timely manner. 


The scientific case for observing time should establish two things.

  1. It should outline the scientific problem(s) or question(s) toward whose solution the observations are requested, and place these questions in the larger scientific context.
  2. It should show how the measurements requested will be used to illuminate these questions or problems.

The technical case should demonstrate that the proposed measurements are technically feasible, given the performance of the proposed instrument(s), in the time requested. The number of target objects required should be justified.

Applications must include complete lists of the objects to be observed, their magnitudes and their approximate equatorial coordinates. Applications without such lists will be rejected.

Specific points which must be addressed:

  • How the proposed observations relate to the applicants previous work, and to other work in the field.
  • Why the Keck Observatory and its site on Mauna Kea, are particularly important or even essential for the proposed observations.
  • The spatial, spectral and temporal range and resolution required.
  • Estimates of the signal-to-noise required and expected, and justification for the number of nights requested for the entire program.
  • If new or unusual techniques are to be used, make clear how observations and calibrations will be obtained.
  • Brief description of the status of large telescope time that has been awarded during the past 2 years, such as progress with data reduction and publications.
  • Any other information which may assist the TAC in evaluating the scientific merits of the proposal and its suitability for the Keck Telescopes.

The presentation should be aimed at someone who is not a specialist in the area of astronomy under study. A specific scientific case, rather than a broad general one, is usually more successful. 


The members of the panels must make their decisions in a limited amount of time, and based on a limited amount of information. It is incumbent on the proposer to provide the panel with as much relevant information as possible, in a form which can be digested rapidly. Write clearly and concisely. Check that the order of the text follows the logic of the proposal. Make sparing use of emphasis, bold face, and exclamation marks. 

Scientific justification 

A possible structure of an observing proposal is as follows. The first few sentences sketch the general area in which the proposal belongs, and the important questions of current research in this area. As the panel members will be knowledgeable in this area, this sketch can be very brief, and should serve mainly to indicate to the panel the general interest of the proposer in his/her field. For example, if you write a proposal on M dwarfs, X-ray binaries, or an S0 galaxy, you may safely assume that the panel members know what these are, but you should indicate what you think are the important questions that current research on M dwarfs, X-ray binaries, or S0 galaxies tries to answer.

The second part of the proposal zooms in on the specific subject of the proposed observations. Which specific question do you want to tackle with your proposal? The third part gives a brief non-technical description of the proposed observations (the technical part is addressed later) and explains how these observations answer the goal of the proposal. It is important here to be both specific and realistic. Thus to state that a measurement of the B-V color of the target will determine the amount of dark matter in the Universe will not impress the panel unless you are able to specify through which steps you can do this. It is this part of the proposal which will to a large extent determine your standing with the panel. A realistic assessment of the expected results will show the panel that your scientific judgment is sound; an exaggerated claim will not. 

If the method of analysis and/or interpretation has been described in an article, you may limit yourself to a brief outline, and provide a reference for the full detail. If the article only exists as a preprint, provide a web site or an ftp address where the preprint in question can be obtained. The panel should be provided with enough information to assess how reliable your proposed methods are. 

Points of interest 

When writing a proposal, it is advisable to keep in mind the following points. 

  • Who will be interested in your results?
    Maybe only the proposer... but maybe all astronomers working on the same object, on the same class of objects, and indeed maybe many astronomers in different fields of interest. It is useful if you assess this in the proposal. The wider the possible interest in your results, the higher the ranking of your proposal.
  • Do you have theoretical backup?
    If so, the value of your proposal is enhanced. For example, if your proposal aims to determine the element abundances in an M dwarf or an S0 galaxy, it will be useful if you can show that you have the ability to obtain theoretical line strengths as a function of abundance. If you wish to determine the amount of dark matter in a cluster of galaxies, it helps if you can show that you have simulations which indicate that the proposed data will suffice to do this.
  • Are you capable of handling the data?
    An indication of this is always useful; it is mandatory if the data analysis is complicated and not straightforward. The best way to show that you can do the job is to refer to your track record in the form of earlier publications. If this doesn't exist, because you are embarking on a new field, you may wish to propose a pilot study first so as to be able to show your mettle.
  • Statistics
    You should show awareness of the statistical aspects of your proposed observations. For example, if ten similar systems have been observed, and you propose to observe number eleven, you must assess how much can be learned from the one new system that hasn't been learned yet from the previous ten. If a hundred systems have been observed, and you propose to observe ten more, you cannot claim that the statistics will be improved dramatically, unless you show that the proposed targets cover a range of parameters that is not well covered by the earlier observations.
  • Target selection
    It is useful to explain why you selected your specific target(s). For example, if you wish to determine the amount of dark matter in the galaxy cluster Abell A2218, explain why this cluster is a better target than other clusters. Possible answers to explore include
    • the galaxies in this cluster are bright and can be observed in relatively short exposures
    • many velocities have been measured already, to which your data can be usefully added
    • simulations show that the galaxies in this cluster are well distributed to determine the dark matter distribution
    • the amount of dark matter has already been determined with another method, to which our results can be compared usefully.