USGS - science for a changing world

Open-File Report 03-236

National Geochronological Database

Revised by Jan Sloan1, Christopher D. Henry2, Melanie Hopkins3, and Steve Ludington3

Original database by Robert E. Zartman, Charles A. Bush, and Carl Abston

This report is preliminary and has not been reviewed for conformity with U.S. Geological Survey (USGS) editorial standards or with the North American Stratigraphic Code. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government.

1U.S. Geological Survey 11450 Joanie Court, Reno, NV 89509
2Nevada Bureau of Mines and Geology, Reno, NV 89557
3U.S. Geological Survey, Menlo Park, CA 94025

Introduction and Background

The National Geochronological Data Base (NGDB) was established by the United States Geological Survey (USGS) to collect and organize published isotopic (also known as radiometric) ages of rocks in the United States. The NGDB (originally known as the Radioactive Age Data Base, RADB) was started in 1974. A committee appointed by the Director of the USGS was given the mission to investigate the feasibility of compiling the published radiometric ages for the United States into a computerized data bank for ready access by the user community. A successful pilot program, which was conducted in 1975 and 1976 for the State of Wyoming, led to a decision to proceed with the compilation of the entire United States.

For each dated rock sample reported in published literature, a record containing information on sample location, rock description, analytical data, age, interpretation, and literature citation was constructed and included in the NGDB. The NGDB was originally constructed and maintained on a mainframe computer, and later converted to a Helix Express relational database maintained on an Apple Macintosh desktop computer. The NGDB and a program to search the data files were published and distributed on Compact Disc-Read Only Memory (CD-ROM) in standard ISO 9660 format as USGS Digital Data Series DDS-14 (Zartman and others, 1995). As of May 1994, the NGDB consisted of more than 18,000 records containing over 30,000 individual ages, which is believed to represent approximately one-half the number of ages published for the United States through 1991.

Because the organizational unit responsible for maintaining the database was abolished in 1996, and because we wanted to provide the data in more usable formats, we have reformatted the data, checked and edited the information in some records, and provided this online version of the NGDB.

This report describes the changes made to the data and formats, and provides instructions for the use of the database in geographic information system (GIS) applications. The data are provided in *.mdb (Microsoft Access), *.xls (Microsoft Excel), and *.txt (tab-separated value) formats. We also provide a single non-relational file that contains a subset of the data for ease of use.

Links to the Other Sections of this Report:

| Data Characteristics | Editing Changes | Data | References |

| Appendix 1 (PDF file; 72 KB) | Appendix 2 |

Data Characteristics

The data are organized into 6 tables (Location, K_Ar, Rb_Sr, FT, Sm_Nd, and U_Th_Pb). Location is the master table, and gives information about the sample itself, its location, and its geologic context. The other files provide information about the geochemical measurements and resultant calculated ages, and much of the information bears a many-to-one relationship to the records in the Location table. Figure 1 shows the way the tables can be organized into a relational database, using RecNo as the key field.

Most of the previously published documentation (Zartman and others, 1976a, 1976b; Marshall, 1993) applies to the data in this report. However, we made minor changes in the names of many fields to conform to requirements for Microsoft Access format, and eliminated some fields. In addition, the labeling of the data in successive versions of the database has varied. The tables (1 through 6) in Appendix 1 (PDF file; 72 KB) illustrate the name changes that we have made and track the labeling variations. These tables will be useful to anyone who wants to compare our information with earlier versions.

Additional information about the NGDB is available in Appendix 2

diagram showing the structure of the NGDB database

Editing Changes

In this section, we provide details of the changes made to the database for this release. Tables were exported from the Helix Express database to Excel, where most editing was done. The information in many fields was stored as two-, three, or four-digit numeric or character codes. Most of the codes were decoded into complete information (specific actions are described below). For fields not mentioned, no editing was necessary. This release restores those entries that were truncated at 255 characters in the DDS-14 release.

In testing of the files in GIS format, we identified a few hundred records in the location table that plotted in a location that conflicts with the information in the database, either in the oceans, in Canada or Mexico, or in the wrong state. Many of them accurately reflect offshore samples, or samples from small islands. Many appear to be only a few meters into the "wrong" state. But about 100 presented more serious location problems. By consulting the original publications and topographic maps, we were able to correct 88 of these. A few locations that were truly ambiguous were not modified.

Location file

LatDec and LongDec — In the RADB, latitudes and longitudes were stored in degree, minute, second format; for this release we converted all entries to decimal degrees and report these in separate fields in addition to the original format.

LatDeg, LatMin, LatSec, North, LongDeg, LongMin, LongSec, North, West — We divided the latitude and longitude into four fields each, one for degrees, one for minutes, one for seconds, and one specifying geographic position in relation to the equator or the Greenwich Meridian, respectively. All latitudes are north; there was one incorrectly labeled "S" (which we changed). All longitudes are west with the exception of records 4454 through 4456 for samples from Kiska Island.

State — We converted each to its two-character U.S. postal code.

County — We converted entries to mixed case and inserted "Unk" where blank.

QuadScale — We inserted "Unk" where blank.

QuadName — We converted entries to mixed case and inserted "Unk" where blank.

SampSour — We inserted "Unk" where "unknown" or blank.

RockName — We converted all entries to mixed case and wrote out all abbreviations in full to avoid inconsistencies.

Descrip — Many of these entries are still in upper case, abbreviations are used widely and inconsistently, and we have not assessed content for accuracy. Because editing would require attention to each record (of which there are more than 18,000), we did not do any editing.

Labs — We decoded all laboratory names from RADB, which were previously represented by a unique two-digit code. Some entries had more than one code, which we assume means that the same sample was analyzed at different labs, most likely by different methods; in this case we listed all of the labs. Several codes were not listed in the RADB documentation and required a literature search. Where followed by a (?), the code was inconsistently represented in the literature; we included the original code with the most questionable entries and inserted "Unk" where blank.

Compiled — We converted entries to mixed case and a consistent format.

DateComp — No changes were made.

LocPrec — In DDS-14, each entry was preceded by the phrase "Latitude/longitude"; we removed this as extraneous. Where cells were blank, we chose to insert "Location not given or only in terms of less accurate than 30 minutes" as this had been used elsewhere.

LocComm — No changes were made.

Ref and Ref2 — All data in this compilation have been published previously. In DDS-14, the Ref field contained the year of publication, which was represented by a two-digit number, the 19- prefix assumed. The Ref2 field contained a unique number assigned to each reference within a given year. We decoded and wrote out all references in the Ref field. Within the Microsoft Excel files, the references are completely contained within this one field and the Ref2 field was removed. A small percentage of these references contain more than 255 characters. Microsoft Excel files are able to store and display more than 255 characters in a single cell. Microsoft Access, however, cannot hold more than 255 characters and thus we have included the part of those references that exceed 255 characters in the Ref2 field in the Access database. In most cases, this portion consists of the journal (or other source), volume, and page numbers. We included this field (Ref2) only in the text files.

Geologic Unit — We deleted this field entirely, because there were very few entries, all of which were coded and imparted little additional information.

Lexicon Ages — We deleted this field entirely, because the time scale used is out-of-date, and such divisions can be easily calculated as needed.

Petrographic Code — This information is contained in the RockName field. Because it is redundant, we deleted this field.

Fields present in all Age Data Files

RecNo — We changed the name of this field from "Record #".
Mineral fields (KarMin, RbSrMin, UPbMin, FTMin, SmNdMin) — We decoded all entries and inserted "Unk" where "unknown" or blank. A few codes were not included in the original documentation and required a literature search. This information was not available for the following records. In the K_Ar table, record 12455 came from a basalt but the mineral analyzed (or whether it was whole rock) was not specified. The references for records 14675, 14782, and 7783 were unavailable. In the Rb_Sr data file, references for records 6863 (code ZA6), 6864 (ZA6), 6862 (ZA7), 6865 (ZA7), 6866 (ZB1) were unavailable. In the U_Th_Pb table, the reference for record 7731 (MC4) was unavailable.

Decay Constant fields (KarDecCon, RbSrDecCon, UPbDecCon, FTDecCon, SmNdDecCon) — We decoded all decay constants. Some entries are followed by an asterisk; because the significance of this symbol is unknown, we chose to retain these. The only file with questionable entries is that of the fossil track ages. The entered values appear to be eight orders of magnitude different from the decay constant per year value usually recorded. In all other age files, the entered values appear to be per million years (instead of per year as typically reported). We did not recalculate radiometric ages using more recently accepted decay constants. There are cases, however, where ages have been converted to modern decay constants by the original compilers and are now different from those originally published. In either case, this field only includes the decay constants reported in the original literature. To avoid misinterpretation, check for consistency between records; some recalculation may be needed where direct comparison is intended. Use of obsolete decay constants may be indicated in the OtherComm field as well.

SampSuit — These numbers are used only to differentiate groups of samples; thus no editing was necessary.

AddRef — This field contained codes similar to those in the Ref and Ref2 fields (two-digit year followed by unique number for all publications of that year). We decoded all references and entered them in shorthand (author(s) followed by year) in this field. All of these references are available in full in the Ref and Ref2 fields.

Comment fields — All entries were processed with a spelling checker, and obvious corrections were made, however errors in the spelling of proper names, locations, and stratigraphic names may remain.

Additional fields discarded — The following fields were not preserved because of no, or very few entries, or because the information can now be easily calculated: in the Rb_Sr table, percent radiogenic strontium; in the Sm_Nd table, age comments and other comments.

Fields unique to specific data files

RbSrAnlTyp — This field contained some codes that had not been translated previously. The entry for records 5513 and 6139 was "10", a code that should indicate conventional K-Ar. These were each one of a group of samples analyzed using conventional Rb-Sr methods (assumed initial 87Sr/86Sr); we assumed that these samples were analyzed by the same method as others in the group and edited the "10" accordingly. The entry for records 621, 622, 627, 656, 670, and 674-685 was "24", which does not exist in the coding tables. All records listed as "24", however, were duplicate records of the same samples. The only information these duplicate records contained that the first record did not was an assumed initial ratio 87Sr/86Sr and an additional age calculation; thus we consolidated the information into one record, and deleted the additional age calculation.

Within the Sm-Nd file, we removed the following fields because they had no entries: SampComm, 143NdRad, NdRadPct, 143_144Init, AnalType, AnalComm, SampSuit, and AddRef.

Norm (Rb_Sr table) — Although we did no editing to this field, we include these comments, as the meaning of this field is not adequately conveyed in preexisting documentation. This is a flag that indicates whether the original publication reported a Sr analysis that was for the total amount of strontium present (i.e., both original and radiogenic, indicated by T), or for only the normal, non-radiogenic strontium (indicated by N). If the original publication didn’t report this information, the field is blank. The information may be important for geochemical abundance studies that might not want to include radiogenic strontium. However, radiogenic strontium only forms a significant part of total strontium in minerals that are both old, and contain large amounts of rubidium; an example might be a Precambrian lepidolite.


These data are reformatted versions of the data originally presented in Zartman and others (1995, U.S. Geological Survey Digital Data Series DDS-14). The database contains age information from more than 1,300 published sources, found primarily through GEOREF searches. The sources of the data are described in DDS-14 and 3 additional open-file reports (Zartman and others, 1976a, 1976b; Marshall, 1993). Comments about data accuracy can also be found in those sources.

The data are provided here in 3 formats, as a Microsoft Access database (.mdb), as individual Microsoft Excel tables (.xls), as tab-separated text files (.txt). In addition, a query was constructed in Access that assembles all the age determinations, and produces an Excel table (allages.xls).

Download NGDB files

NGDB2.mdb (Microsoft Access database, 31 MB)
Location.xls (Microsoft Excel, 7.9 MB)
K_Ar.xls (Microsoft Excel, 2.8 MB)
Rb_Sr.xls (Microsoft Excel, 1.6 MB)
Sm_Nd.xls (Microsoft Excel, 16 KB)
FT.xls (Microsoft Excel, 272 KB)
U_Th_Pb.xls (Microsoft Excel, 744 KB)
allages.xls (Microsoft Excel, 5.9 MB)
Location.txt (tab-separated value, 9.1 MB)
K_Ar.txt (tab-separated value, 2.1 MB)
Rb_Sr.txt (tab-separated value, 1.3 MB)
Sm_Nd.txt (tab-separated value, 4 KB)
FT.txt (tab-separated value, 156 KB)
U_Th_Pb.txt (tab-separated value, 600 KB)


Marshall, Brian D., 1993, Conversion of the Radiometric Age Data Bank (RADB) to the National Geochronological Data Base (NGDB): U.S. Geological Survey Open-File Rep., 93-336, 76 p.

Zartman, R.E., Cole, J.C., and Marvin, R.F., 1976a, User's guide to the Radiometric Age Data Bank (RADB): U.S. Geological Survey Open-File Rep., 76-674, 77 p.

Zartman, R.E., Cole, J.C., and Marvin, R.F., 1976b, Reporter's guide to the Radiometric Age Data Bank (RADB): U.S. Geological Survey Open-File Rep., 76-675, 73 p.

Zartman, R.E., Bush, C.A., and Abston, Carl, 1995, National Geochronological and Natural Radioelement Data Bases, A, National Geochronological Data Base, B, Natural Radioelement Data Base: U.S. Geological Survey Digital Data Series DDS-14. [CD-ROM.]

Accessibility FOIA Privacy Policies and Notices logo U.S. Department of the Interior | U.S. Geological Survey
Page Contact Information: Contact USGS
Page Last Modified: Friday, April 04, 2014, 10:09:30 AM