project background 2 - origin and organization

There are rope bridges in Tibet, but our work isn't all romance and Indiana Jones-style bravado. Actually there's very little of that (some memorable exceptions aside). Mostly there's a lot of planning and paperwork.

page 1 - science       page 2 - origin and organization       page 3 - field work

Quite separate from scientific issues is the very practical matter of getting the science done. There are many ways to organize, fund, and foster scientific research. Different nations handle this in different ways, and different scientific disciplines have different cultures and styles. However, certain principles and realities are common to all systems. Modern science costs money for both the actual science and for infrastructure. Money being a scarce resource, some means has to be found to allocate it. And even basic research is usually expected to contribute to the common good, if not through direct research results, than through training and education.

We thought that people might find it interesting to learn a little more about how science is done, so we've summarized what's involved in getting a project like ours under way. Keep in mind that we're describing a large Earth-sciences project, based in the United States and supported by the National Science Foundation (NSF). Other disciplines, other countries, other granting agencies, and scientists with different philosophies and objectives may all do things differently.

Academic Context

A good deal of fundamental research in geology is coordinated by academics based in universities as either teaching or research faculty. For better or worse, a Ph.D. is required as the de facto "license" to do research. This involves a four-year undergraduate degree in the sciences or its equivalent, often a two-year Master's degree, and then a three- to five-year Ph.D. program. In many cases these days an additonal two- to three-year postdoctoral fellowship is becoming the standard for new faculity hires. So all in all, people are starting untenured junior faculty positions in their late 20's or early 30's.

Much as they try, many academics don't have the time to carry out complex, extended research solely on their own because of teaching and administrative duties. In addition, the training of future scientists and technical specialists is seen by universities and NSF as an integral role of researchers. So most research is performed by a blend of undergraduate and graduate students, postdocs, and PIs (jargon for Principal Investigators). In return for funding, NSF PI's are expected to disseminate the results of their work by writing or having their students write papers in peer-reviewed journals.

Preparing a project and a proposal

Ideas from projects and proposals usually stem from an unequal blend of insight, creativity, and mostly experience from previous work. Often it's the data that don't fit a model that are most interesting to follow up, because if they are not blunders, such data are symptoms of a mismatch between expectations and reality. Once you have an idea, the next step is to talk to some people who have the expertise needed to deal with the problem at hand, perhaps through a series of phone calls or over a beer or two at a professional meeting. If people think there is something concrete to work with, the next steps are to tighten up the idea, sketch out a timetable for action, and secure the go-ahead from department chair or dean if there are any resource issues or matching funds involved.

The origin of our indentor-corners project was tied to directly to the work some of us did at Nanga Parbat. We had developed the tectonic aneurysm model, and we had talked quite a bit about how unique this feature might be, whether it could be tested or at least shown to exist somewhere else, what it's regional tectonic setting might be like, and in general what the dynamics of an indentor corner would be. We were interested in continuing this work, and we learned that Jean Pierre Burg and his group had found geology very similar to Nanga Parbat in the area around Namche Barwa, thousands of kilometers to the east in a setting that was different in detail, but overall located at the heart of the eastern Himalayan syntaxis. We knew Bernard Hallet at U. Washington was interested in the dynamics of this part of the world, so we enlisted his participation and that of his surface-processes group, and we began to plot and scheme about what concrete problems could be addressed in southeastern Tibet.

To develop new field projects, it is nearly essential to experience the geology of the study area, and for foreign projects, it is completely essential to nurture solid foreign contacts and counterparts. Clark Burchfiel at MIT, who has worked in China and Tibet for many years, kindly made some introductions for us with colleagues at the Chengdu Institute of Geology and Mineral Resources. In 1998, with support from NSF, Hallet, Meltzer, and Zeitler made a reconnaissance trip to Namche Barwa along with colleagues from Chengdu. Such trips are not typical for NSF projects, but the Continental Dynamics Program recognizes that large, complex international projects require a fair bit of planning, and is willing to fund such trips if the PIs make a case that a reconnaissance is needed. Given that our planned work included installation of a large seismic array, and that the part of Tibet we were interested in was remote and little-studied, we felt we had a case. In retrospect, the experience with logistics alone was invaluable in helping us plan our proposed work.

Writing a proposal for a large group project is a special challenge, as the ideas, styles, enthusiasm, procrastination, schedules, and budgets of many different PIs need to be folded together into a seamless, compelling document. In our case we had five main PIs representing a total of 13 PIs and over ten disciplines, and we were proposing collaboration with overseas counterparts in a remote part of the world. The NSF program we were submitting to, Continental Dynamics (CD), is competitive and can only fund a few new projects even in a good budget year. Compared to a standard 10-page small-grant NSF proposal, CD proposals seem roomy with their 30-page limit, but these pages have to include not only the science description, but also enough details about each discipline to satisfy specialist reviewers, as well as details about project organization and a justification for why the project requires an integrated multidisciplinary approach. Needless to say, for CD proposals the four weeks before proposal deadlines are always desperate. It's hard to imagine pulling this off without fast internet connections!

If you're really, really interested in what's involved, we've made the the full text of our proposal available as a pdf file, with most of the figures available in a second pdf file. Note that the proposal has all of the budget paperwork, other forms, and personal data removed, so what you're seeing is still a streamlined version of the complete beast.

      proposal text (516 kb)       proposal figures (xxx kb) (not quite done!)

All in all, it took us three submissions before our project was finally funded, with a start date in 2001. The process also included a preproposal that we submitted to the CD Program, which is an optional step that allows the program to prefilter ideas for projects and either give PIs some guidance and feedback, or save them the trouble of preparing a complex proposal for an idea that the panel feels is unlikely to be successful. In addition, Zeitler and Meltzer made a planning trip to China in order to develop and sign a memorandum of understanding between the project and the Chengdu Institute.

The proposal review process

The National Science Foundation is organized into programs and initiatives which solicit research proposals in various disciplines and cross-disciplinary fields. These solicitations can be targeted, or just be open to basic research proposals. The Continental Dynamics Program seeks to fund projects aimed at understanding continental evolution, with the understanding that most such proposals will be larger multidisciplinary group efforts.

Although the details differ between NSF programs, proposal evaluation involves the same basic series of steps. Proposals are now required to be submitted electronically through the excellent FastLane management system, which integrates proposal submission, handling, and review. The Program director selects mail reviewers for each proposal who are asked to submit peer reviews involving a text narrative as well as numerical scores. Each program also has a panel of experts who typically serve terms of about three years. These panelists independently read all proposals, and then gather at NSF headquarters in Washington for a meeting at which the submitted proposals are discussed and their merits debated. The panel assessment, mail reviews, and scores are then used by the program director to make final decisions about which proposals to fund, depending on the availability of funds in the current budget and on obligations in future years. Generally, there are more proposals that would merit funding than there is money, so part of the job of the program director is to field calls from disappointed PIs. PIs receive copies of the anonymous mail reviews, as well as a written panel summary. PIs who wish to resubmit a proposal can do so after a one-year wait.

For funded proposals, the program director can make happier calls giving PIs the good news. However, in an attempt to spread funds as widely as possible, most program directors wheel and deal, using the advice of their panels to suggest budget cuts to the funded proposals in order to eke out a few more dollars. After a period of negotiation, pleading, and counter-proposals, a revised budget is agreed upon that can still support the science goals of the project, and then the formal award paperwork is processed.

Funding: where does all the money go?

If you actually went and downloaded the text of our indentor-corner proposal, you may have noticed that the budget was a healthy-sounding $2.1 million. If you're a US taxpayer, you may wonder how we're managing to spend that much on yak feed and grad student vittles.

The first thing to realize is that a portion of that budget is what are called "indirect costs," or overheard. Research institutions negotiate with the federal government on rates for this item, which is meant to reflect the costs of maintaining research infrastructure (lab space, libraries, internet, utilities, hazardous waste disposal, health and safety, etc.). These are real costs, and the presence of this infrastructure benefits students, so this actually amounts to an investment in the US higher-education system. Of course, PIs and reviewers grumble about overhead when they see budgets grow, and we always suspect we're not getting our fair share of this levy. It's human nature! Anyway, our project indirect-cost rate is about 30% because most of our work is being performed off-campus. So that takes the award down to about $1.6 million.

This leaves what are called "direct costs," which cover all aspects of our project: PI summer salaries and benefits, grad student support and tuition, foreign travel and field work, sample and equipment shipping, use charges for advanced lab equipment and sample preparation, travel to professional and group meetings, telecommunication, computing, and publication costs, and subcontracts for specialized professional services. There are about eight main PIs on our project, and the project is five years in duration. That equates to average funding per PI of only $40,000 per year. Given that support for just a graduate student might run around $25,000 per year in tuition and stipend, you can see that despite the impressive-sounding total budget, we're not rolling in dough, flying business class, or renting limousines (and to make you taxpayers feel even better, we can't use grant money for entertainment or alcohol, and we have to fly US flag carriers unless the routing is desperately inconvenient). We're also not complaining in any way, as this budget has been carefully thought through to get our work done.

Project coordination and communication

It's an overused image, but managing a dispersed science project really is like herding cats. Our project, like most CD projects, functions as a loose collective, with a project leader (in our case, Zeitler at Lehigh) serving as contact with NSF and overall coordinator. Each participating institution has its own directly award budgeted, so other than cajole and suggest and bluster, the project coordinator relies on the good will of the other PIs in getting things done on time, and in a coordinated fashion. This actually works pretty well, but to paraphrase what The New Yorker says about artists, academics lead complicated lives.

Fortunately, the web and email make it pretty easy these days to stay in touch, exchange ideas, and disseminate information. Every year our project holds a group meeting in conjunction with the fall San Franscisco meeting of the American Geophysical Union, which most of us would be attending anyway. And, given the long hours spent traveling to the field, there's usually a chance to catch up then. This web site is another means of keeping the group informed, as there is a members-only section that posts news, updates, data, requests, and papers in preparation.

Project outcomes

The goals of this project are the generation of new ideas and new data, and there is an expectation by both NSF and the scientific community that these will be reported in the formal, peer-reviewed scientific literature. However, there are other ways in which results can be shared. These include presentation of talks and posters at scientific meetings, participation in specialized workshops, outreach to the public and school groups, participation in documentaries (like the one made about our work at Nanga Parbat), and development of web sites like this one.

page 1 - science       page 2 - origin and organization       page 3 - field work