Yes. I typically take 2 new students each year.

It makes essentially no difference as far as the research you'll do, the admissions process, the stipend you'll get, the office you'll be in, the students you'll hang out with, or the general requirements for the Ph.D. The only differences are in (1) the courses you'll be asked to take (even there the difference is slight), (2) the teaching requirement (see question on teaching requirement below), and (3) the label of your Ph.D. EPS requires you to take eight semester courses of which four have to be in your area of concentration (Atmospheric Science), two have to be Applied Math or Statistics at the 100 level or higher, and two have to be in another area of EPS (such as geophysics, geochemistry, geodesy...). SEAS also requires eight semester courses of which six have to be in your major and two have to be in a minor; you get to define your major and minor as you wish, and to tailor your course curriculum accordingly, but the courses have to represent a coherent ensemble and all but three have to be at the 200 level. Both SEAS and EPS can bend their rules upon petition (for example as to what defines "breadth" in EPS). SEAS is more flexible at granting credit for graduate courses taken elsewhere. EPS will give you a Ph.D. in Earth & Planetary Sciences while SEAS will give you a Ph.D. in the major you have chosen (for example Environmental Engineering, Atmospheric Science...). The current graduate students in the group are split almost 50/50 between SEAS and EPS. Browse through the SEAS and EPS web pages and decide where you'd feel most at home.

Not necessarily. Atmospheric chemistry is an interdisciplinary science. My graduate students come from a diversity of backgrounds from applied math to chemistry and everything in between. Some students come in with zero chemistry; others come in with little physics. Many have had little or no exposure to atmospheric science, while others have had quite a bit. All that's OK. There's no "ideal" preparation. Gaps are expected, and the purpose of your courses at Harvard will be to fill these gaps. I do expect applicants to have had at least two years of college math.

No. Most incoming students have little or no scientific programming experience. This is best learned on the job, like laboratory skills.

The application deadline is sometime in December for admission to the following fall semester starting in September. Check the departmental web sites (see links above) for the exact date. The application folders get to the faculty by mid-January (I don't see them before then!). Decisions on admissions are made in early February and applicants are contacted immediately after the decision.

The ideal applicant has very good grades from a well-known institution, and some research experience with a strong letter of recommendation from their research advisor. However, I often take students with weaker grades if they have demonstrated talent for research, and I will take risks on students with no research experience if their academic record is stellar. Although the Harvard application form allows you to be ambivalent about which professor you wish to work with, in practice I get so many applications that I only look seriously at the ones that express a clear intent to work with me (your statement of purpose is important for this). I also strongly encourage you to email me before your application to express interest, so that I can give you feedback on your preparation and also look out for your application.

You are always welcome to visit Harvard. Just drop me an email and I'll also arrange for you to meet with some students in my group and with the graduate study office. However, visiting prior to admission is really not needed. Admitted students residing in North America are invited to visit at Harvard's expense. Unfortunately we cannot do this for admitted students from outside North America (too expensive). In those cases we can communicate by email and phone.

Our admission process tries to match each promising applicant with a sponsoring professor, based on the explicit interest expressed by the applicant in working with that professor and the willingness of the professor to take that student. Because of the large number of applications, I only look at applicants who expressly state that they want to work with me. Once admitted you get assigned your sponsoring professor as academic advisor, but this doesn't actually commit you to doing your Ph.D. research with that professor. All first-year graduate students are supported by University Fellowships so that they are not financially beholden to any professor. You can use your first year to freely explore your options and you can easily switch advisors - all you need to do is find an other willing professor. In practice this happens rarely. By June of your first year you should have settled on a Ph.D. research advisor because this is when serious research starts and your stipend then begins to be charged to your advisor's research accounts.

You will have a lot of freedom to choose your own project within the general sphere of activity of the group. The group web site should give you a good idea of our research directions. Check out in particular our current research web page. I generally encourage students in their first year to take just 3 regular courses a semester (a full load is 4 courses) plus a "reading and research" course with me which is an opportunity to start thinking about research. The summer after the first year is an important time in which to get started on what will become your Ph.D. research project. In the second year you typically take 0-2 courses a semester and can begin to really spend time on your research, and after that you're 100% research.

Honestly, it's not very productive to discuss future projects with much specificity at the time of your application, because I generally don't have projects "on ice" - if something is interesting we get moving on it right away, we don't wait! I recommend that you browse through the current research web page to get a sense of the general research areas that we are engaged in, as it is likely that your future project will build on those. We'll get serious about defining your project when you actually start. That being said, if you need to discuss specifics of future projects for a fellowship application or for your own agenda then sure, let's talk.

All Ph.D. students admitted to SEAS or EPS are guaranteed tuition and stipend, i.e., a graduate fellowship, for the normal duration of their time at Harvard. During the first academic year the funding comes from Harvard, after that it comes from my research grants. I encourage my students once at Harvard to apply for external fellowships, because they look good on your CV and of course it brings some relief to my research grants.

If you want to work with me you need to apply to the Ph.D. program. SEAS has a small 1-year MS program (courses only - no thesis expected) and a ME program requiring in addition a small research thesis. However, there are no fellowships available for MS/ME students and we don't consider these programs as conduits for the Ph.D. Ph.D. students can pick up their MS degree at the end of their coursework if they so wish. Students who enroll in the Ph.D. program and decide after 1-2 years that this is not for them have the opportunity to leave with a MS or ME degree.

A requirement of your graduate fellowship is that you serve as Teaching Fellow (TF) during your time at Harvard. SEAS requires that you do it for one semester, EPS for two. TFing a class means teaching a section, grading homeworks, and having office hours. It is expected to take no more than 10h/wk. You can satisfy the requirement by TFing one of the undergraduate atmospheric classes, for example EPS133 which I teach. Once you've fulfilled your requirement you can teach more classes if you wish and get paid for this (on top of your stipend). Most of my students do the requirement and no more, but a few have taught for a number of semesters because they liked it and liked the money. This is not a problem with me as long as you maintain productivity in your research and that's always worked out fine.

You have to take a qualifying exam in the spring of your second year (at which point you'll typically have completed all or almost all your coursework). This exam consists in an oral presentation of your Ph.D. research proposal (EPS also requires a 10-page written version of the proposal). This is an excellent opportunity for you to get feedback on your research direction from a faculty committee. Just betweeen you and me, I have never seen anyone fail the exam - sometimes a student may be asked to take an additional course to address an apparent gap, and in rare cases a student may be asked to retake the exam. So it is really more an opportunity for the student than a selection tool for the faculty. We can afford not to be selective at the qualifying exam level because we are highly selective at the admission level. The duration of the Ph.D. is typically 5-6 years. We shoot for five.

I have graduated many Ph.D. students over the course of my career. Of these, most are now professors at research universities or research scientists in national labs. A few are research scientists in the private sector or staff scientists in non-research organizations. You can get more detailed information by clicking on their names on my alumni page (former Ph.D. students are identified by red circles).

They'll be happy to help. See the group list. I encourage you to check out their web pages to see what they're working on.