GSA home

Log In | GSA Community | GSA Store | Join GSA | Donate | Contact Us

GSA home

| GSA Community | GSA Store | Donate | Contact Us

About GSA

Career Resources

Connected Community

Divisions &
Associated Societies

Education & Outreach

GSA Foundation

GSA International




Public Policy



Find Your Science at GSA

Return to Main
Discussion Page

Discussion of GSA Time Unit Conventions


Add a comment

Time Units

David T. King commented at 9/21/2009 11:16:25 AM

An abbreviation is supposed to be a shortened version of the original words. For million years, one would expect m.y., for example. The abbreviation Ma is for mega-annum. If the original sentence said "mega-annum," Ma would be correctly used. Until we all use "mega-annum" in everyday speech and writing, m.y. should be used when "million years" is meant. One other thing: Abbreviations should not be used at the end of sentences. So, if Ma is at the end of a sentence, it should be written out, according to the rules of most style manuals.

Anthony Douglas Pollington commented at 9/21/2009 11:53:39 AM

It is clearly important for those studying rates and durations to have two distinct conventions for absolute age and durational age. Adopting a clearly defined convention of "a" (Ma, Ga etc) for absolute age and "yr" (Myr, Gyr etc) for duration will allow for the continued distinction between two importantly different topics (age and duration) while avoiding the ambiguity of Gy (grays) and b.y. (English-centric language).

Raymond V. Ingersoll commented at 9/21/2009 1:27:15 PM

Arguments for and against both conventions have merit. One aspect that should be stressed, however, is that the present designation of Ma, Ga, etc. that indicate years before present represent negative numbers (years before present, with zero at present, and positive numbers for the future or time intervals). Therefore, Ma and Ga should always be plotted with larger (more negative) numbers on the left and zero at the axis. Making a clear distinction between negative numbers (years in past) and positive numbers (intervals) may be the way to resolve the issue. In other words, -100Ma indicates a time in the Cretaceous. The time interval of -100Ma to -90Ma is represented by 10Ma. Using negative values would make plots conform to convention with larger negative numbers to the left, and they would be read from left to right as time progresses.

Mark T. Brandon commented at 9/21/2009 2:45:26 PM

Renne and Villa made their original argument in a letter in 2004 to GSA Today. I accepted their recommendation then and have found no problem in applying it in my writing. It is also easier, in my opinion, to explain this usage to science students, who have already been exposed to the SI system, but have not been steeped in the traditions of our "geoscience culture". Renne and Villa make a very simple point: They argue that the age of an event in the time scale is nothing more than the duration of time since the present. Thus a "geologic age" should use SI units for time. The SI system of units is widely accepted as the standard in the sciences. It seems silly for geoscientists to invent or retain a different set of units. The notations kyr. and Myr. evolved in an ad hoc way, along with other variants, such as kybp, Mybp, etc. My reading of the geoscience literature indicates that these ad hoc units are on the decline, and I suspect that they will be gone within the next decade. That said, I am confident that this issue will remain controversial for some time to come. My rule as an editor is that exceptions in writing styles should be tolerated, but the burden is on the writer to explain why an exception is useful or preferred. Thus, I suggest that GSA editorial guidelines should leave it as a recommendation that authors use standard SI units, including units of time a, ka, Ma, and Ga. There is no need for a absolute ruling on this issue.

Mark T. Brandon commented at 9/21/2009 2:52:25 PM

One additional thought: Some geoscientists have argued that we should start using the geons for geologic time, in much the same way we use numbered centuries. (See for details.) As an example, the earth was formed during geon 45 (which includes 4.5 Ga). We presently live in geon 0 (and in the 21st century). I see nothing wrong with this proposal, but it would be wrong to consider this scheme as a replacement for the SI units of time (including the derived unit of annum).

Gary J. Axen commented at 9/21/2009 3:23:47 PM

There are two main problems with the use of “Ma” for time before the present, but m.y. (or something similar) for durations of time. The greatest problem is that units of measure and frames of reference should be separate, as they are in every other unit of measure in common usage. Imagine the laughing stock that physics would be if kilometers were always measured from some arbitrary (but changing!) reference point, say the center of the Earth, and one had to use “t.m.” (thousands of meters) for all other distances. There is nothing sacrosanct about measures of time that puts them in a different category. For some reason, many geologists believe wholeheartedly that geological time is a sacred, special measure of time. It is not! It has associated errors and a variety of valid reference points for different purposes. The argument that “everyone knows that 10 Ma is in the Miocene” is specious, given that (a) many good scientists do not know that fact (ask your average chemist) and (b) “Miocene” is defined on the basis of fossils and stratigraphy but “10 Ma” is measured using radiogenic isotopes, so the numerical age of the boundaries of the eons, epochs, etc., are always changing. Given that “the present” changes on a second-by-second basis, the unit Ma (with “before present” implicit) does not even have a well-defined frame of reference! It makes us Earth scientists, as a group, seem to the rest of the scientific world like fossil throwbacks. The second problem is that the current usage causes confusion and miscommunication within a complicated field where clarity should be our goal. As Earth sciences become increasingly interdisciplinary, we should strive to open our communications to those with other specialties. This is best done with simple prose--i.e., by adding short words like “ago”--rather than by adopting a confusing, mostly unstated, insiders-only definition.

Andrew V. Okulitch commented at 9/21/2009 3:52:33 PM

I submitted the letter below to the GSA in February 2005 and it was published not long afterwards. My opinions remain unchanged. A further complication likely not appreciated by chemists and physicists is that the astronomical year changes over time. The astronomical year governs the cyclicity of events preserved in the geological record and is thus more significant in geology than an arbitrary year defined by arbitrary seconds. Direct linkage to SI units remains beyond our grasp at present and using non-SI unit nomenclature might be a good reminder of that limitation. Dear Editor, In their letter to the GSA in October last year, Paul Renne and Igor Villa urged GSA journals to conform to the SI regarding units of time. They conceded that the SI unit of time, the second, is impractical for astronomical and geological time intervals and noted that SI "tolerates" other units, in geology, the "annum". They then deplored the apparent inconsistent dual usage of millions of anna (Ma) versus millions of years (m.y.) presently common in all geological literature and urged adoption of a single usage in conformity with the SI. With due respect to their abilities as geochronologists of international reputation, Paul and Igor are confused about a couple of things. Firstly, annum is Latin for year. Since English is the international language of science at present, the latter unit is quite acceptable and readily understood. There is no reason to abandon it. Secondly, they are mixing up mileposts with distances, ie. points in time (ka, Ma, Ga) with durations (k.y., m.y., g.y.). Just as the number on a milepost denotes a location in a particular direction from a chosen datum, a number chosen to be conveniently equal to the distance in miles (or any SI unit) along the path to be traveled, so does the designation of the age of a geological event or object position it along time's arrow a specific duration from the present (our usual chosen datum). Again, the number selected is equal to the number of years that have passed between the event and the present. When writing of ages, one uses "mileposts"; when the duration of an event, or the duration of time between two events is being described, one uses years. Admittedly, the milepost labels could be ky, My and Gy in keeping with modern language, but retaining anna in the abbreviation emphasizes the difference between points in time and lengths of time, thus is a clearer form of presentation. In this, I expect we all agree that the goal is clarity and precision in communication. Adherence to inadequate rules is merely mindless bookkeeping, and neither good writing nor good science. Cheers, Andrew V. Okulitch, Emeritus Geologist Geological Survey of Canada

Thomas M. Scott commented at 9/22/2009 5:47:33 AM

The short and sweet of it is that geological time is special. To avoid confusion between time spans and points in time, the use of different abbreviations or symbology is necessary. To use a single series of abbreviations, as is proposed, invites confusion.

Lucy E. Edwards commented at 9/22/2009 8:08:28 AM

Fortunately or unfortunately, the abbreviation “Ma” means “millions of years before the present.” This is how it was defined. The abbreviation Ma (Mega-annum) refers to units of yr x 10(superscript)6 measured from the present (1950 A.D. by international agreement) pastward. It means the same as the cumbersome "million of years before present" and it is a fixed chronology analogous to calendars tied to historical events. Berggren and Van Couvering (1979, p. A506). It simply isn’t a unit of time in the International System of Measurements (SI).

Brian Romans, Ph.D. commented at 9/22/2009 1:35:46 PM

I appreciate arguments for abandoning dual units for date and duration with respect to the correct usage and origin of 'annus' and how it applies to geologic time. I think Renne and Villa, among others, make valid points. However, an example that comes to mind is reporting rates of geologic processes. For example, Holocene researchers commonly compare measured historical rates to millennial rates (e.g., cm/yr, cm/kyr). Under this new convention wouldn't yearly rates have to be reported as "cm/a" to be consistent? I foresee the usage and editorial enforcement of "cm/a" as problematic. At this point I think the community should continue to debate and discuss the numerous examples and instances where a major change like this might introduce more confusion than it solves.

Donald E. Owen commented at 9/22/2009 7:00:16 PM

I am appalled that you would consider adopting a proposal from chemists who never deal with geologic time that would result in loss of the distinction between date (Ma) and duration (my). This would be confusing and would result in a loss of clarity. The following is quoted from a portion of my paper published this week in the journal, Stratigraphy: "NUMERICAL AGE/DURATION DATA In papers that deal with isotopic age or other quantifiable age data, a few minor, but troublesome, terminology problems regarding numerical ages can occur. First, according to the Code, use of the nouns "isotopic age" or "numerical age" instead of the nouns "date" or "absolute age" is recommended. Second, the term calibration should be used for the special form of designating chronostratigraphic boundaries in terms of numerical ages. Authors need to be aware of conventions for abbreviating numerical ages. The Code and Guide recommend the following abbreviations for numerical ages in years before the present: ka = l03; Ma = 106; Ga = 109. The prefixes k, M, and G are borrowed from the International System of Units (SI); however, time in years is not an SI unit, and time in years ago is not an SI concept. Incidentally, the duration of an annum is a modern year and the present refers to AD 1950. Qualifiers such as "ago" or "before the present" are redundant after the above formal abbreviations, because duration from the present to the past is indicated by their use. Avoiding the use of m.y.a./m.y.b.p.-type abbreviations for ages is recommended. However, authors should remember that the formal Ma-type abbreviations are not used for the duration of an interval of geologic time that does not extend to the present; in such cases, the informal abbreviations, y., k.y., m.y., and b.y. (or yr., k.yr., m.yr. and b.yr.) are used. The Code (Article 13c) and Guide (Salvador, 1994, p.16) specify this distinction. For example, the numerical age boundaries of the Late Cretaceous Epoch are calibrated at 99.6 Ma and 65.5 Ma (Gradstein et al.; 2004), but the duration of the Late Cretaceous Epoch is 34.1 m.y. (99.6 – 65.5), not 34.1 Ma (which was during Late Eocene time)--many authors make this mistake—don’t join them. Think of it this way: I may have a class that meets from 3:00 to 4:00 PM, but the duration of that class is 1 hour." If you persist in this unwise action, you will find that most stratigraphers, journals, and geological societies will not follow your capitulation. Please, don't do it!

Greg Balco commented at 9/23/2009 11:25:30 AM

One important aspect of this discussion is that the issue of the relationship between geological and SI units is a straw herring. A glance at the NIST document that serves as the US reference for correct application of SI units, or any similar document, makes it extremely clear that years, or any variant thereof, are specifically not envisioned as an SI unit. In addition, "a" overlaps for all practical purposes with the preexisting SI unit of "A," meaning ampere. So the geological community is pretty well off the leash with regard to SI already. The relationship or lack thereof between various units of years and the SI is neither here nor there, and doesn't belong in this discussion.

Leslie R. Fyffe commented at 9/23/2009 11:55:56 AM

Never occurred to me that we would need two time conventions. And I still can't see why we would want to be so pedantic. Surely the context will make it clear when one is referring to a time interval vs an absolute time.

Dr. Glenn Borchardt commented at 9/30/2009 3:13:56 PM

Dates and durations must have differing abbreviations. Thus Ma and ka are specific points in time, whereas My and ky are durations. In addition, we should follow the suggestions of Coleman and others (1987) in which ka is used only for calendar-corrected data. Thus: yr B.P. Uncorrected radiocarbon age expressed in years before present, calculated from 1950. Calendar-corrected ages are expressed in ka, or, if warranted, as A.D. or B.C. Soils example: to = age when soil formation or aggradation began, ka tb = age when the soil or stratum was buried, ka td = duration of soil development or aggradation, ky Reference: Colman, S.M., Pierce, K.L., and Birkeland, P.W., 1987, Suggested terminology for Quaternary dating methods: Quaternary Research, v. 28, p. 314-319. Glenn

Peter Eichhubl commented at 10/1/2009 11:31:48 AM

Scientific conventions are changed for many reasons, not all of them because the new convention offers a tangible benefit. Sometimes old ones fall with the head of the king. The designations of Pluto as a non-planet or the Tertiary as a non-Epoch come to mind here. While justifiable on some theoretical grounds, these changes clearly benefit the textbook industry more than its readers. Same here: Abbreviations like m.y. for million years (and that’s what m.y. is, not a unit!) may seem strange to the non-geologist, but so is our habit of licking pieces of shale. Even if we change the convention now, we have to train students in the use of the old convention. What is the gain? The high priests of SI can’t even define the kilogram in a way that does not depend on a slowly varying chunk of metal. Until they have figured that out, m.y. is o.k. with m.e.

Nicholas W. Hayman commented at 10/19/2009 3:56:31 PM

I used to use the annum abbreviation (Ma, Ka, etc..) universally, but at some point was corrected. The correction is to NOT use "annum" as a substitution for "years ago". In other words "MegaAnnum" means Million Years, and not Million Years ago (which is, strictly speaking, correct, I think). So, I now use m.y.-ago in the sentence structure of when something happened (e.g. rifting began 2 m.y.-ago), and Ma when there is actually a date (the ~2 Ma basalts). In other words, "Ma" is the unit of time, but does not have greater grammatical significance. Following this logic, I suppose that I am OK with the 159Ma-150Ma=9Ma usage, whereas I am less comfortable reading "the basalts were erupted 9Ma"). I'd be happy to dispense with all of this in favor of a universal "Ma", but please can I keep using "Tertiary" and "Sphene".

Travis Michael LeRoy commented at 12/9/2011 7:56:11 PM


Travis Michael LeRoy commented at 12/9/2011 7:56:46 PM


Travis Michael LeRoy commented at 12/9/2011 8:05:41 PM