Topics covered in this draft include the following:
Whenever appropriate, I strongly recommend using e-mail to communicate questions and (short) updates to me. That way, I can take time to think about my response without the pressure of someone standing in my doorway and without having to interrupt my current train of thought. Also, I will then have your message on file in case I need to refer to it again later.
If you run into a problem on the weekend or in the evening, particularly if it pertains to a computer failure or other crisis, I don't mind getting phone calls at home (see no. on blackboard) to at least hear about them. Either I will have some idea of what to do or else I can try to get hold of someone who does.
We will meet sometime during the first year to begin drawing up an informal plan of study which will specify which lecture courses you should take as part of your degree program. While the number and type of courses will depend on your background, your research project, and whether you are pursuing an M.S. or Ph.D., you should be aware of the following minimum credits for all AOS students:
Ph.D.: At least 15 credits of lecture courses, at least 600 level and above. Details subject to approval by Advisory Committee.
In the meantime, I encourage you to begin thinking as soon as possible about the following:
Ph.D. students who have established a Ph.D. committee must meet yearly with that committee to give a progress report. A written summation is then submitted to Connie for inclusion in the student's file. The meeting will normally be scheduled during the Spring semester.
Nowadays, this degree is regarded as a desirable qualification for most meteorological occupations wherein the employee is performing some kind of technical work, usually with at least some degree of supervision or direction. In my opinion, virtually any admitted graduate student who is sufficiently diligent and has already achieved good academic success at the undergraduate level is at least theoretically capable of successfully completing an M.S. program in our department (the same is not necessarily true for the Ph.D. degree; see below).
It is generally expected that the student's M.S. thesis research will
be completed with significant direct supervision by the major advisor.
Any journal publications that result from that thesis will therefore probably
have the advisor listed as first author, unless the student has shown unusual
initiative and competence in pursuing the assigned task. In short, an M.S.
student is not required to demonstrate that he/she is necessarily
well-suited to a career as an independent research scientist.
In my opinion, the awarding of a Ph.D. certifies that the student has demonstrated the intellectual ability, scientific maturity, and personal initiative and independence necessary to produce and publish original and significant scientific research with at most limited supervision. A Ph.D. program is specifically designed to prepare a student for a career as an independent researcher and/or scientific project leader in academia, industry or a government lab.
Key qualities that I therefore expect my Ph.D. students to demonstrate by the time they defend their dissertation include ALL of the following:
Even if a student succeeds in forming a Ph.D. committee, the committee
retains the right to conclude at a later date that the student does not,
and probably will not, satisfy all of the qualities outlined above, in
which case it is possible that the student may be advised to terminate
his or her studies in this department. Note that it is neither
in the student's interest nor in the interest of this department to award
a Ph.D. degree to someone is unlikely to function well as an independent
researcher.
In fact, the most likely result is that the student would end up
over-qualified (on paper at least) for most M.S-level employment opportunities
and yet unable to compete for, or perform satisfactorily in, most Ph.D.-level
research positions.
On the one hand, your primary purpose in being here is to earn an advanced degree in Atmospheric Science, and if you're like most people, you would like to accomplish this task as quickly and efficiently as possible.
On the other hand, your financial support from me is conditioned on your active contribution to fulfilling the specific research objectives spelled out in my grant proposal (you are encouraged to review that proposal as needed to keep its aims fresh in your mind!). Sometimes, the two objectives coincide completely; often, however, they do not. When they do not, it is not uncommon for students and advisors to have differing interpretations of their own rights and responsibilities.
To forestall possible misunderstandings, I would like to outline the principles that I think should govern our working relationship. To begin with, a standard half-time Research Assistantship is based on the assumption that you are employed by your advisor half-time, or 20 hours per week on average. In other words, I believe the faculty advisor technically has the right to request up to 20 hours per week (on average) of specifically assigned work, some of which may be unrelated to the student's thesis research and could, in fact, be pure drudgery (e.g., proofreading papers, copying tapes, photocopying articles, etc.).
On the other hand, you will soon notice (if you haven't already) that most successful graduate students expect to work at least 40 hours per week on average (often more like 50-60), counting course work, thesis research, unrelated work for one's advisor, and (at the appropriate time) writing one's thesis. This means that, even in a worst-case scenario, you should still be able to find a minimum of 20--30 hours per week of time spent exclusively in pursuing your degree (including course work), even in the unlikely event that your advisor has you mopping floors on his 20 hours.
In summary, I consider it my responsibility to make sure that I do not demand more than 20 hours per week of work (on average!) which does not directly further the completion of your degree. Nevertheless, if at any time you feel that the tasks I am giving you require consistently more than this, I encourage you to communicate this to me. It is often difficult for an advisor to accurately gauge the amount of effort or the number of hours required for a given task, taking into account the experience and skills of the student.
Likewise, I consider it your responsibility to give me my money's worth 20 hours a week, on average. This time INCLUDES carrying out specific assignments I give you, performing calculations and preparing programs and graphics for the research project, reading relevant research papers, writing, etc. It EXCLUDES homework and studying for your courses, time spent on personal web surfing and e-mail, and similar activities. If you notice that you are spending a lot of time on the latter activities, there's a very good chance that I will have noticed too!
In any case, it bears repeating: RA pay is always subject
to the availability of funding and is also contingent on satisfactory progress
by the student.
What this means is that many M.S. students, and most Ph.D. students, will have the research grant that initially supported them expire before their studies are complete. Continued research support then depends on the advisor winning another three-year research grant prior to the expiration of the first grant.
Regrettably, it is impossible to predict in advance whether a grant proposal will be successful, and no amount of proposal writing can completely guarantee success. Thus, one must always live with the knowledge that a student's funding may dry up at an inopportune time.
I am therefore extremely cautious about accepting new students unless I can be reasonably confident of being able to supporting them long enough for them to complete their degree. In practice, this means that I won't even offer you an RA position unless I already have at least one year's guaranteed support, with a strong likelihood of new funding thereafter. But as I said, there are never any guarantees, and I sometimes have to take risks in order to even get the most qualified students when they apply.
What if your funding DOES run out? If this should ever happen, I would do my best to find alternate sources of support for you. This could include a TA position or possibly some kind of temporary project work with someone else in the building.
Side comment on Teaching Assistantships: On the one hand, TA stipends are often viewed by the student as less than optimal sources of financial support because the effort of teaching conflicts with, rather than overlaps, the effort of completing a thesis. Note however that there is an increasing trend among graduate programs to require some minimal level of teaching experience of their Ph.D. students. I tend to agree with this view. Therefore, if you are my Ph.D. student, there is an excellent chance that I will ask you at some point to switch from an RA to a TA assignment for one semester.
As regards the 20 hours that you officially owe me as an RA, I don't
particularly care if you work one 20 hour day per week or seven three-hour
days, as long as you are awake and reasonably productive during that time.
The only exceptions to this policy might be when deadlines arise that require
a task to be completed within a certain time frame --- for this reason,
I need to have a reasonably good idea in advance of when you plan to be
available so that we can iron out any potential timing problems well in
advance. The most important thing is communication. Therefore, please let
me know in advance if you plan to be away. I will in turn let you know
as much as possible in advance if I will need your help preparing a poster,
conference preprint, or some other project by a certain deadline.
I may meet with you individually more often. If we have not already
done so, let's meet as soon as possible to establish a regular meeting
time during which you can get me up to speed on the work you are doing
and give me an opportunity to comment, critique, and/or congratulate. The
above scheduled meetings would of course be in addition to regular interaction
over the details of the current project. In addition to planned meetings,
my door is usually open if you just want to drop by to discuss something.
If there is any doubt about my availability, then e-mail is usually the
best way to get my attention long enough to at least make an appointment.
Furthermore, it is surprisingly common to discover that a seminar speaker is working on a problem, or using methods, that hold the key to a problem of interest to you, even if it's not in the same general area.
Finally, having a large, engaged turn-out at seminars and at departmental social functions reinforces the image of a strong, active, and stimulating department in the eyes of our visitors as well as in the eyes of our own graduating students. Such impressions are self-fulfilling, because they contribute to our favorable standing among the best and brightest prospective students and faculty, reviewers, funding agencies, etc. This is all good for the long-term future of our department.
In view of all of the above benefits, I expect my students to regularly attend AOS and CIMSS scientific seminars and colloquia and to pursue opportunities to interact both socially and scientifically with other students and researchers in the building. It is also expected of my students, especially Ph.D. students, that they will occasionally present progress reports on their research in an informal seminar.
This is particularly important for M.S. students, since they have a much tighter timetable for completing their degree requirements. Also, my ability (or willingness) to provide adequate supervision or resources may suffer if you end up choosing to focus on something which deviates too far from the objectives of my funded projects and/or my other areas of interest.
Apart from these constraints, I encourage you to explore scientific problems or questions that happen to catch your interest in the course of your research. Though many may be dead ends, one might end up being the central focus of your thesis.
Recall that M.S. thesis work often falls in the category of ``directed research''; that is, the work is largely executed and written up by the student, but the gross objectives and methods of the investigation are largely those specified by the supervisor. However, even an M.S. student should not be content with a passive role in the research, merely following the advisor's instructions to the letter. If you are contemplating continuing in the Ph.D. program, the quality and originality of your M.S. thesis will have a huge impact on how you are perceived by potential Ph.D. committee members.
If you are a Ph.D. student, you should be aware that the primary distinguishing characteristic of a successful Ph.D. dissertation is that it represents an original scholarly contribution to the field by the student, based on an investigation largely designed and executed by the student (please review the section on what I expect of Ph.D. students). In fact, if you are working toward a Ph.D., you are expected to become at least as much of an expert in your chosen specialty as your advisor!
In keeping with the above, one hallmark of a successful Ph.D. dissertation is that it describes research which is of sufficiently high quality to be published in a respected peer-reviewed journal, with the student as first author. In fact, my personal policy is that at least one article derived from the student's Ph.D. dissertation should be accepted for publication in a respected journal prior to final approval of that dissertation. In my experience, students who complete their dissertation and then leave for a new job find it very difficult to find the time to finish up the preparation and submission of journal manuscripts. This hurts both the student, whose future employment may depend in part on his/her publication record, and the advisor, whose own career, not to mention future research funding , depends heavily on his having something worthwhile to show for the time and money invested in his/her grad students.
You should plan on using the text processing package LaTeX to prepare your thesis. Not only will this simplify formatting (we have some customized style files for this purpose), but it will greatly simplify the creation of journal manuscripts from your thesis when the time comes. More discussion of LaTeX is given later in this document.
Note also that you will undoubtedly, at one time or another, do work for me which is unrelated to your thesis but which contributes to the publication of a paper. This brings me to the traditionally sensitive issue of authorship:
To appear as an author on a scientific paper in a respected journal is one of the most important things you can do to enhance your standing as an atmospheric scientist. Particularly in the academic world, one's publication record is traditionally (for better or for worse) the single most heavily weighted criterion for promotion, pay raises, etc.
Historically, the issue of who gets to be an author on a paper, and in what order, has often been the source of hurt feelings and resentment on the part of graduate students, who may feel that they do ``all the work'' and the advisor takes all the credit as author of the resulting paper. I would like to try to head off any such misunderstandings by spelling out in advance my views on authorship.
In my opinion, being an author on a scientific paper implies a substantial contribution to planning the research, collecting new data in a laboratory or in the field (as distinguished from merely reading data tapes), selecting or designing the analytical techniques to be employed, and/or interpreting the results. In other words, there is a significant intellectual and/or creative investment in the project, in addition to mere grunt work like programming.
By tradition the major professor almost always appears as an author (not necessarily the first author, however) on a paper simply by virtue of having written the grant proposal and obtained the funding to conduct and publish the research in the first place, though he/she (hopefully) will have made other contributions as well.
As for you, the research assistant, it may or may not be appropriate for you to be an author on a publication in any given instance. In general, if your primary contribution was simply to carry out specific technical tasks that I have assigned, such as crunching data in a prescribed way, plotting graphs and maps, proofreading the text of a manuscript, etc., then I do not believe this in itself qualifies for co-authorship on the resulting paper, especially if you have no role in actually writing the manuscript which gets submitted. There has to be a more substantial contribution of the type cited above.
Assuming there is such a contribution, and that you therefore have earned the right to be an author on a paper, there is also the issue of who should be first author. Again, the deciding factor should be who has taken primary scientific responsibility for the work which is to be documented. However, a couple of generalizations can be made here:
A journal paper derived directly from my student's Ph.D. dissertation will almost always have the student listed as first author and me as second author. [Note: In many other fields (e.g., Chemistry) it is traditional for the advisor to be first author even on papers derived from Ph.D. dissertations!]
This is because an acceptable Ph.D. dissertation, by definition, is an original research contribution whose design, successful execution, and interpretation is largely in the hands of the student, with the faculty advisor serving mainly as, well, an advisor. This does not mean that all publications that a Ph.D. student contributes to fall into the above category, even if the work involved eventually gets incorporated into his/her dissertation.
In the case of a M.S. thesis, it can in principle go either way, depending on the degree to which the advisor personally directs the execution of the research. A few M.S. students who are unusually capable and who have the necessary background quickly assume a great deal of creative control over the path of the research described in their thesis; in such situations, the student will have earned the right to be first author. More often, however, both time and experience preclude an M.S. student from doing much more than executing and writing up a project under the advisor's supervision. In such cases the advisor will normally be first author.
Usually, in order to justify the cost of travel to a meeting the student must be presenting a paper, either orally or as a poster. In most cases, the primary source of support for such travel is expected to be the faculty advisor's research funds. However, there are other possibilities as well.
For example, the American Meteorological Society has often given travel and lodging support to Purdue students who agree to work for the AMS at a particular conference, usually by running audiovisual equipment, communicating logistical problems to the AMS leadership, etc. The main drawback is that you are usually obligated for the entire conference and may have little opportunity to attend the sessions of your choice. This has been a source of frustration for many of our past students. On the other hand, this arrangement makes travel to conferences possible for students who might not otherwise be able to attend at all, especially if he/she is not presenting a paper.
There are several kinds of local resources that you will be concerned
with in the course of your studies. These include (1) local expertise,
(2) local hardware, (3) local software, (4) data, (5) miscellaneous supplies
and services. Eventually I want to put a document together that summarizes
these local resources in some detail. Since this will require more time
than I have at the moment, let me just try to hit a few highlights here:
Remote access to computers: If you have an Internet Service Provider that you subscribe to from home, then consider installing a good telnet client , such as TeraTerm, on your PC. In the latest versions of most web browsers, a telnet client is already built in, but it's not always very good (e.g. poor terminal emulation). You can then point your web browser to a host ....
As a general goal, I would like to encourage everyone in my group to help document software and hardware resources as you become aware of their existence and/or become acquainted with their use. The ideal would be to have ``cookbook''-type explanations of how to do almost anything any one of us would ever want to do, whether it be related to plotting figures, running radiative transfer models and analysis programs, reading data sets, or what have you. Any assistance you can give me in putting together documentation of this type will be greatly appreciated. Incidentally, I think a web page might be a great way to organize this information.
Let me also refer you to some software tools that I would like to encourage you to become familiar with. Some of these have somewhat steep ``learning curves'', but the invested effort is probably worth it, since these tools are almost universally available in a UNIX environment and can be very productivity-enhancing when used intelligently.
TeX (and LaTeX): These typesetting languages can't be beat when it comes to preparing professional quality documents and preprints, especially those involving extensive mathematical equations. In the engineering and physical sciences, LaTeX is almost universally used; even many book and journal publishers prefer LaTeX files in place of (or in addition to) old-fashioned hardcopy manuscripts. Unfortunately, the meteorological community is way behind the curve on this and, as a whole, still uses WYSIWYG word processing programs like Word or Wordperfect, which are admittedly easy to learn initially but rather ill-suited for complex scientific documents. And they don't even exist on UNIX machines. For a brief rundown on the advantages of LaTeX, click here and here.And if you get REALLY ambitious, read a rather lengthy (though somewhat out-of-date) exchange about pros and cons here.
EMACS: Distressingly many people get used to editing with ``vi'' on UNIX systems and become so comfortable with this DECIDEDLY CLUNKY editor that they never realize what they're missing. EMACS is a widely available, full-blown editing environment that is exceedingly intelligent about various programming languages, including FORTRAN, C, and TeX, among other things. Unlike vi, which is basically a juiced-up line editor which unnaturally requires one to jump back and forth between ``input'' and ``command'' modes, EMACS is a true full-screen editor with a tremendous range of functionality. Like vi, EMACS is available (or can be installed) on all UNIX systems (and even on PCs running Windows). In summary, I can't think of many reasons why anyone would want to stick with vi as one's primary text editing program, but I can think of plenty for taking the effort to learn EMACS.
Once you know both EMACS and TeX/LaTeX, there are also rather few convincing reasons I can think of for wanting to use Microsoft Word to prepare scientific documents, such as your thesis.
In fact, as a matter of policy, all journal papers and other extended documents prepared with any involvement from me must be prepared using LaTeX. The only exceptions will be when we are secondary authors on a paper prepared somewhere else, using Word or something similar.
IDL: I am just learning IDL myself, but I have already found this commercial package to be exceedingly useful for both numerical and graphical work, including preparation of publication-quality figures. I have found that a lot of things I used to do with FORTRAN and NCAR Graphics, I can now do much faster and more easily with IDL. Please get to know this package. You'll be glad you did. By the way, I have a copy of a book called IDL Programming Techniques by David Fanning which does a much better job of explaining the essentials than the official IDL documentation does.
There's a lot more I could say about our local resources, but it will have to wait. If you have any questions, please ask one of your fellow students, Bob Green, or me. And we will all appreciate your sharing your own expertise with others as well.
Supplies, services, and related expenses
(under construction)
There is no way that a (relatively) short document like this can
cover every aspect of your program of study and your work as a Research
Assistant. I will undoubtedly revise and expand it as issues arise that
I haven't anticipated. The most important thing is to bring any questions,
problems, or concerns you may have to my attention so that we can work
them out before they begin to hinder your/our progress.