This article was adapted from a presentation at the conference "X-ray
and Radio Connections" http://www.aoc.nrao.edu/events/xraydio/
Note that there was a companion piece on radio astronomy proposals by Ed Fomalont.
Each year there are 700-800 proposals for Chandra observations, archival work, and theory projects. Because of observing time and funding limits, only about 200 are accepted, one in four. The numbers for Cycle 5 were:
|Normal||606||173||70 ks (ave)|
|Large Project||68||10||300-1000 ks|
Table 3: Chandra observing proposals.
Chandra proposals are solicited once a year. The Call for Proposals contains schedule and rules. The ProposerŐs Observatory Guide contains instrument descriptions. Software for feasibility calculations, information, and help are available at http://cxc.harvard.edu/.Content
If a proposal is truly excellent, and the review panel recognizes this, it will quickly be approved. However, 85% of proposals submitted are considered "good" and reviewers spend most of their time ranking the "good" proposals. Your challenge is to write the proposal so the excellence of the project is clear to the panel; or, at least, produce a clear, pleasing document that will stand out in the pack of "good" proposals.
A proposal has three vital parts: Science goals, feasibility, and suitability of Chandra for the project. The panel must be convinced that the science derived from an observation will be interesting. The proposal must show that the observing time requested will produce enough signal (counts for X-rays) to do the job. It must also demonstrate that Chandra capabilities (e.g. arcsecond resolution) are needed.
You can write a single proposal (for example) for Chandra Time and VLA/VLB, NOAO, HST, XMM, RXTE, or Spitzer time. The Chandra review can award up to a few percent of the observing time on another instrument to such joint proposals (subject to approval of the appropriate Director). The proposal must show that both Chandra and the other instrument are essential to meet scientific objectives, and must demonstrate feasibility for each.Format Two sections comprise a Chandra proposal:
Take care to avoid mistakes. Misspelling investigator or institution names negate the use of the computer to find conflicts of interest. Every year at least one proposer puts first names in the last name boxes.
Errors in target coordinates can prevent the review organizers from
finding target conflicts. Every year there are ~25 gross errors
in target coordinates submitted; for example,
Target: MS0735+7421 @ RA = 07 41 50.3, Dec = -74 14 50.6, and
Target: Tololo 0109-383 @ RA = 13 35 22.1, Dec = -42 32 20.0.
A latex/pdf template is supplied which will produce an
easy-to-read Science Justification. Keep in mind that reviewers have
to read 60-70 proposals. Documents with tiny font, small margins,
or small figures with minuscule captions will not be
The organizers enforce page limits by removing any excess pages.
All proposals must be submitted electronically using the RPS software. Before the deadline, errors can be corrected and proposals resubmitted. No post-deadline submissions are accepted. Each year we get 1000 submissions for 800 proposals. There are 600 submissions in the last day and 100 in the last hour. People wait until the last hour, submit the proposal, then read it, then hurriedly resubmit or call to explain the special circumstances leading them to request a late submission. "I accidentally sent an early version of our proposal." is not uncommon. Late corrections to the Science Justification are not accepted.The Peer Review
Proposals are divided among 12 or 13 panels according to science topic. Each science topic is covered by 2 or 3 panels so proposals can be placed to avoid conflicts. There are ~65 proposals/panel and 8 reviewers/panel.
Before the review, each panelist reads all proposals and assigns a preliminary grade. These grades are used for Triage. The first task for each panel is to view the list of the lowest-ranked proposals. The bottom 25% (~15 for each panel) are then eliminated from further consideration. However, if a panelist thinks any proposal in this group should be considered further, it can be resurrected and discussed with the rest. The panel then discusses and grades the remaining 50 proposals - about 10 minutes for each! Each has been assigned a primary and a secondary reviewer who are responsible for a detailed reading of the proposal, presentation to the panel, and writing of a report communicating the grade and panel comments for the proposer.
On the third day of the review, the Large Projects, already graded by the topical panels are discussed and ranked by a Big-project panel which awards 20-30% of the observing time.Advice
We go to a great deal of trouble to find competent reviewers and to avoid conflict of interest in the review process. All proposers are treated equally. Reviewers are conscientious and fair but this is an intense process. There is a lot to accomplish in a limited time. Proposers should retain the perspective of Ecc 9:11 where, "the race is not to the swift, nor the battle to the strong, nor bread to the wise , nor riches to the intelligent, nor observing time to the writers of good proposals, but time and chance happen to them all."
Start early. DonŐt wait until the last minute.
Write some of the description for someone who is not an expert in the field. Remember some of the panelists do not work in your specific area and the proposal has to survive triage.
Print the proposal and read it before submitting.
Avoid unnecessary Coinvestigators. Reviewers cannot participate in reviews of proposals for which they are CoIs and sometimes for those which have CoIs from their institutions.
Volunteer to be a peer reviewer. Having seen the process you will be able to write better proposals.