There are three basic means of researching a topic:
There is a starting list of references handed out in class . Some references will be relevant to your topics; this will give you a jump on (1) as you should be able to then look up these papers, read them, and note the references that describe things you would like to present in the field. Of course you won't be able to find more recent papers this way; nor can you find papers that were not cited but might be relevant. For this you need the other two methods of library research.
Keyword search is the most commonly attempted search by most researchers. You can start from scratch and hunt through the databases to find what you need. The downside is that it is frequently tricky to get keywords that yield the papers you really want. Searches often yield way too much or way too little. There are two main databases I will suggest; others can be found through the Chinook Electronic Resources page (http://ucblibraries.colorado.edu/research/findarticles/geology.htm): These two are only accessible from CU computers or through VPN from offcampus (see vpn.colorado.edu for info on CU's VPN).
Remember that these databases usually do your search as a phrase. Entering "Rio Grande Rift faulting" yields nothing because that exact phrase isn't in any papers (unlike Google, where you would get hits). Instead you have to use boolean operators AND, OR, and NOT. So a better search is "Rio Grande Rift AND fault*". The other key element is the wildcard *. This matches any word starting with the letters before the *.
Georef's advantage is mainly the keywords that are attached to articles, which are more extensive than in other databases. Also Georef includes a number of publication types that do not make it into Science Citation Index like field trip guidebooks, book volumes, and meeting abstracts. This last is a mixed blessing as you might find a blizzard of abstracts on a topic, but the tabs at the top of the search result page now make it relatively painless to separate abstracts out. Disadvantages include occasional lapses, especially of AGU journals (some issues of Tectonics are missing, for instance). The degree to which keywords are correctly applied can vary noticeably.
Keyword hints:
You can combine searches in many ways in Georef. For instance, if you were interested in finding out about the Mesozoic plutons in the Sierra Nevada, you could start with Sierra Nevada. This yields too many results, so you might then do something like "#1 AND Mesozoic" which uses your previous search (#1) and the keyword Mesozoic. (A shortcut is to just type a search starting with AND, OR or NOT, and Georef will default to appending this to your previous search--well, at least it used to. Now you would probably click the "Modify search " button and add the additional constraint).
Keywords are always a challenge. In Georef, state and county are frequently present, while some more familiar placename (XYZ Peak) might not. Similarly, an orogeny name might not be put in as frequently as the time (Jurassic). If you aren't picking up what you want, try to use more general terms and then winnow down. State names and often county names are included accurately in the descriptor field (see below). What is nice in the current version of Georef is that the Descriptors list for a particular record gives you an idea of the keywords Georef will use--you can check terms from one record that capture the aspect you are interested in and launch a new search from there.
If you find a record that fits well, look at the full record and see what keywords or descriptors are there. It might help you find similar papers.
More arcane in Georef is the ability to cut out some reference types or search within specific fields. There is a field in Georef, publication type (PT), that contains the type of document. Almost any search will turn up a huge number of hits that are abstracts. These are often not of interest. You can modify a search to exclude abstracts by adding "NOT (abstract in PT)" or, more simply in the current incarnation of Georef, click the "not abstract only" button . The one time abstracts can be useful is in finding somebody who was working on something of interest--you might look to see if they subsequently published something of interest that didn't meet your original search criteria (or if a preprint is on their homepage). Another useful field is the descriptor field. This field will accurately have the state and usually the county (if localized enough) of interest. Thus the Rio Grande Rift search might become "Rio Grande Rift AND fault* AND (New Mexico in DE) NOT (abstract in DT)". (You will probably be using the pulldown menus in the current version of Georef to set whether your search is limited to the descriptors). These extra fields are now more easily accessed through the Search Tools tab.
Following authors can be a challenge in GeoRef, as GeoRef preserves the various abbreviated forms of an author's name. Hyphenated names in particular can be a challenge.
SCI is one of the biggest database out there. We have electronic access to it back to 1965, which is often good enough for our purposes. This is a general science database and includes articles in many other disciplines, which can occasionally make a search quite unwieldy. The keywords are not as comprehensive as Georef (especially placenames) but the database is generally more up to date. You might want to unclick the non-physical science databases on the intro screen.
Access is pretty straightforward from the Web of Science homepage. You will want to choose the Science Citation Index to search and often you'd probably be better with a full search than the more limited regular search.
Citation Search is a very powerful, often underutilized way of finding recent papers that address a topic of interest. It often occurs that you will be able to quickly identify a key paper on some feature, but it might be quite dated. Obviously the old references in it are not much of a help, but it is the perfect paper (had it only been published last year). In SCI you can find out who else cited the paper.
In the full search of SCI, click on the "Cited Reference" button. This will let you search for things well before the 1970 start date of the online SCI that were cited after 1965.
Citation search is the only place you will be able to use non-journal literature in SCI. Things like GSA Special Papers and AGU monographs are not indexed as primary records, but citations to those papers will appear in journal articles.
Cited Ref hints: (SCI)
You can sometimes find out how best to enter an old paper by first looking at a post 1970 paper that cites the old paper. Find the new paper (author and year, for instance) and then click on the full record for that paper. You can then uncheck all but the desired reference in the lefthand box and click on related records.
You can pyramid things this way too. You search for citations to a 1920 paper that is cited by a "perfect" 1950 paper you had found, which in turn leads to a more recent paper, etc.
Another trick is the "Related Records" button. When you find a paper in the database, you will find this button towards the right. It finds other papers that cite much of the same literature. You can trim the list of references to be compared by clicking on the "Cited papers" link, then unclicking all but the references that really pertain to your topic (if you know them). This focused search can be pretty enlightening.
SCI also now has a tool to help follow a particular author through various combinations of initials. It is not perfect but can be a help in those instances where you want to find what a particular author has published.
Although the interface is familiar, the underlying database is more erratic in coverage. Google Scholar relies on publicly accessible databases, which in the earth sciences are kind of odd collections of things like physics and medical databases that happen to overlap onto journals. Because you are getting everything, you might have to sift through a lot more irrelevant information. The good thing here is that the net is wider: you might uncover something unusual. This is not the highest probability path to enlightenment at present.
CU has a subscription to this, so for the most part it works best if you are on a CU computer or otherwise connected as described above for Georef and WOS. Geoscienceworld is an initiative led by GSA to have online versions of journals available under one roof; it has a lot of journals but most notably is missing AGU journals. The search engine was supposed to be full text, which in theory means that it should catch papers without the proper keywords. However, this requires that the version of the text that is uploaded not be a simple scanned image, which is not always the case, especially for older articles, and the present search engine claims to only be searching on keywords. Citations from journals that are also in GeoscienceWorld are found on the abstract page for an article, so you can go directly to desired texts. One of the more novel tricks in GeoscienceWorld is the citation map, which is a link on the right side of the abstract page. This shows both citing and cited articles in a graphical manner.
Wikipedia, like any encyclopedia, is not a primary source of information. The articles on geologic locales are often the result of one interested individual and so can have a decided bias. (For instance, the page on the Rio Grande Rift in 2007 was based nearly entirely on work done at New Mexico Tech). However, the articles are usually footnoted, and those can point you at useful parts of the professional literature.
Return to Field Seminar Homepage
Please send mail to cjones@cires.colorado.edu if you encounter any problems or have suggestions.
C. H. Jones | CIRES | Dept. of Geological Sciences | Univ. of Colorado at Boulder
Last modified at September 23, 2008 3:57 PM