The application of latitudes and longitudes to verbal locality data is called georeferencing. Latitude describes a position in degrees north or south of the equator. Longitude describes a position in degrees east or west of the Greenwich meridian. However, coordinates alone are of limited use without information on uncertainty and the coordinate frame of reference (datum).
Protocols for georeferencing natural history collections are described in the MaNIS Georeferencing Guidelines. Tools that automate these protocols include the Georeferencing Calculator, BioGeomancer, and GeoLocate. GeoLocate is called as a web service by applications within Arctos.
Georeferences are treated as determinations. That is, someone at some time determined these data, and more than one such opinion can be retained. Thus, a locality may have several determinations that differ from each other because the determiners have interpreted the verbatim locality differently, or because original data are potentially incorrect.
Original Units for geographic coordinates vary with the source of the data. Classically, latitude and longitude have been recorded in degrees, minute and seconds. Now, modern GIS applications generally use degrees only, including decimal fractions for all levels of precision. In Arctos, coordinates are stored and displayed in their original format. Except for UTMs, coordinates are also translated to, and stored as, decimal degrees for standardization and mapping. There is not yet programming to convert UTMs to decimal degrees. Include as many digits of precision as are provided in the original data. Forms in Arctos are customized according to which format is chosen, and the vocabulary and formats are defined and described by a code table.
Datum: The geodetic datum to which the latitude and longitude refer. A geodetic datum describes the size, shape, origin, and orientation of a coordinate system for mapping the earth. Latitude and longitude data referenced to the wrong datum can result in positional errors of hundreds of meters. Therefore, when providing latitude and longitude data, it is important to know from which datum those data are derived. Most GPS units allow you to select the geodetic data from which its coordinates will be determined (default usually set to WGS84, but this should be checked in the field). Maps and gazetteers generally provide this information as well.
VARCHAR(255) not null
Reference Source(s) refers to the source of the coordinates and not to the source of the error. Coordinates may be original data collected with the specimen, or they may be provided by after-the-fact georeferencing efforts. In the latter situation, data in Source(s) should be specific enough to allow anyone in the future to use the same resources to validate the coordinates, or to georeference the same locality. These data might be a list of maps, gazetteers or other resources used to georeference the locality. Examples:
- USGS 1:24000 Florence Montana Quad
- Dictionary of Alaska Place Names (Orth, 1967)
- Geographic Names Canada (NRC website)
- Geographic Names Information System (USGS website)
In cases where the coordinates are original data, a description of the original source should be provided. Again, these data should make the coordinates as verifiable as possible by referring to records associated with the specimen. Examples:
- collector’s notation
- preparator’s notation
- specimen label
- accession file
- global positioning system (download)
- global positioning system (transcription)
VARCHAR(40) not null
Georeference Method is the protocol used to obtain the values for the coordinates and the measure of precision. Different methods and tools have been produced, and are sometimes revised, and these differences can produce different results. The vocabulary for this field is controlled.
GPS Accuracy is the distance error reported by a Global Positioning System. Therefore, it a number plus distance units. The original units for this distance should be recorded in the GPS Units field unless the GPS does not report accuracy, in which case a default GPS Accuracy of 30 m should be used under normal reception conditions. GPS Accuracy can be null, but it cannot be zero. (See discussion at MaNIS.)
Extent of a named place is one component contributing to the value of Maximum Uncertainty. It is a measure of the size of the feature of origin in the specific locality – the distance (number plus units) from the point defined by the coordinates to the outer perimeter of the feature of origin (e.g., from the center of town to the farthest point on the city limits). If the coordinates are taken from a GPS, then the extent is the distance from the point where the GPS coordinates were taken to the furthest distance from that point in which collection occurred (e.g., the distance to the furthest trap in a trap line from where the coordinates were taken). Extent and Maximum Uncertainty must share the same units. Extent cannot be zero if GPS Accuracy is null. (See discussion at MaNIS.)
Maximum Uncertainty Distance is the upper limit of the horizontal (as opposed to elevational) distance from the reported latitude and longitude. It describes a circle within which the whole of the described locality lies. Leave the value empty if the uncertainty is unknown, cannot be estimated, or is not applicable (because there are no coordinates). Zero is not a valid value. Maximum Uncertainty cannot be less than the sum of the Extent and the GPS Accuracy converted to the Uncertainty Units.
This is a simple concept, but there are several sources of error which, when correctly combined, often demonstrate a value larger than intuition might suggest. These sources of error are enumerated in the MaNIS Guidelines, and are combined into estimates of total error by the Georeferencing Calculator.
In some circumstances the greatest source of error is the behavior of the collector and/or any intermediary sources of the data. For example, if a locality names a village, the collector may have obtained specimens from a resident who forages over a large area near the village. The collector may even have provided coordinates for the village, often from some standard source, implying specificity equal to the extent of the village. Estimating error can therefore be subjective, and conservative interpretation demands large values for Maximum Uncertainty. To avoid ambiguous or misleading locality descriptions, see MVZ’s guidelines.
INT not null
Determiner is the agent (usually a person) who determined that these coordinates and measures of uncertainty apply to this locality. Often, this is the collector or, dear reader, you. The form will load with the currently logged-in user as the default Determiner for new records.
Sometimes, a determination is developed by two or more successive agents. For example, one agent might locate a named place and provide the coordinates, but little or no information about the uncertainty. A second agent might then evaluate the determination (mapping it and comparing it to the Verbatim Locality) and then develop a Maximum Uncertainty. In this case, we assume that the second agent has re-evaluated the coordinates, and the determination is considered to have been made by the second agent (i.e., the agent who last modified the determination). If there is a need to maintain the identity of the first agent, then the second agent should create a second (separate) determination.
Determination Date is the ISO8601 date that the determination was made. Entry/editing forms load with the current date as a default for new records.
VARCHAR(40) not null
Verification Status: A categorical description of the extent to which the georeference has been verified to represent the location where the specimen or observation was collected. Vocabulary is controlled.
“Verified by collector” indicates that the person who removed the specimen from nature has looked at the coordinates and uncertainty represented on an appropriately scaled map, and believes that these data are accurate and that the represented uncertainty is as small as possible.
TINYINT not null
Accepted? There can be more than one georeferencing determination per locality but only the accepted determination is routinely displayed. You can revert to an earlier determination by changing its accepted flag from “no” to “yes.”
Remarks about the spatial description determination, explaining assumptions made in addition or opposition to the those formalized in the method referred to in Georeference Method.
- Click the select icon () to add a spatial query box to the map. Click the X to start over. Click and drag the box to select an area on the map. The northeast and southwest corner coordinates of the selected area are indicated on the line above the map.
- Use the search box above the map to locate features or places. This searches the map, NOT Arctos specimens.
- You may click the zoom icon in the select tool to zoom the map to the approximate size of the select tool. Scrollwheel and map control zoom work too.
- Spatial query can be used alone, or in combination with any other search parameters.
- Not all records can be found with a spatial query. As of March 2013, 78% of the records represented in Arctos have been documented with geographic coordinates. Within individual collections, this proportion ranges from all to none.
- Localities are not simply points, but circles indicating precision. (Some have radius 0, which should be interpreted as “NULL” rather than “infinitely precise,” and some are very large.) Choose the “query by points and error” option to find specimens by error (the match is “any overlap”) rather than only by coordinate points. Error is represented as a simplified rectangle and may be especially inaccurate at high latitudes and large errors.
- All records are plotted on the geodetic datum WGS84, irrespective of the datum on which they were recorded. Such untransformed non-WGS84 records may be misplaced by as much as half a kilometer. The original datum is documented for individual records.
For example, this query:
(note the “query by points only” option)
produces these results:
Note that BerkeleyMapper does not by default include precision or remember your settings, so you must enable it every time:
Feel free to stop by https://code.google.com/p/berkeleymapper/issues/detail?id=44 and leave a comment!
Change the original query to “points and error”….
To remove the huge-error specimens, we can add a maximum coordinate precision value:
which results in: