DATABASE CONSTRUCTION
Data for the Fort Union Formation and equivalent units were obtained mainly from drill holes and subordinately from measured sections in outcrops used as control points. Data from 18,207 drill holes were collected from government and private industry sources for this assessment. More than seventy-five percent of these drill holes are from coal exploratory drilling and development, and the rest are from oil and gas exploration. Thus, the basic entity of the database is a drill hole bored vertically from a surface location. Information from this drill hole is either in the form of driller’s or geologist’s lithology logs from rotary drill cuttings or cores, and geophysical logs (for example, gamma ray, density, neutron, resistivity, and spontaneous potential). The digital files of the geophysical logs are not included in the database.

The geophysical logs, which make up a large part of the drill-hole information, require special analysis to measure the thickness of coal beds and related rocks. The precision and accuracy of measurement of thickness of coal beds and adjacent rocks on geophysical logs depend on the speed of logging, scale of the log, type of log, type of equipment, and instrument settings (Vaninetti and Thompson, 1982; Wood and others, 1983). Perhaps the most important part of the log analysis is the ability of the user to identify the top and bottom of the beds by using the points of inflection and mid-point of inflection methods (Wood and others, 1983, p. 55). The points of inflection method requires picking the top and bottom of the beds where the curves change directions. The mid-point inflection method requires picking the top and bottom of the beds at points midway between the points of inflection and the initial peak. Thus, the value and accuracy of the geophysical logs in measuring the thickness of coal beds and adjacent rocks varies with the experience of the users. However, because the Fort Union coal beds are very thick, the “operator’s error” with the use of either method is considered negligible. In addition, the mining practice for these thick coal beds is to discard some of the top of the bed and leave the bottom of the bed in place during mining, both of which consist commonly of a “dirty” coal or carbonaceous shale considered uneconomic.

The basic information required for the StratiFact database manager from each drill hole includes: (1) point identification and geographic location; (2) stratigraphy; and (3) depth-based coal analytical data. More information on file designs or configuration of the database manager can be found in the guide to StratiFact published by GRG Corporation (1996).

DRILL-HOLE LOCATION AND STRATIGRAPHIC DATA
The drill-hole identification is typically a unique hole number or a hole name. Drill-hole location coordinates were entered in coordinate systems (Universal Transverse Mercator system, latitude and longitude system, or state plane survey). Drill-hole elevation data were entered as part of the location format. The rock types (for example, coal, sandstone, siltstone, mudstone, etc.) were entered along with the top and bottom depth of the lithology and lithology name. A “rock” name was entered when the lithology between the coal is either undifferentiated sandstone, siltstone, and mudstone or unidentified. Additionally, the stratigraphic data contain formation names (for example, Fort Union Formation), subdivisions labeled as coal zones (for example, Wyodak-Anderson) consisting of one or more coal beds, and unassessed zones above and below this zone. The assessed and unassessed zones are identified according to the names of coal beds contained within the zones and/or members of the formation in which they exist. A drill hole may penetrate one or more stratigraphic formations or members of a formation that contain alternating coal beds and rock types. The coal beds are recognized by standard nomenclature and may exist in one or more members and in a formation.

COAL ZONE
Throughout a basin of coal deposition, individual coal beds may either thicken or thin, merge or split into thinner beds separated by rock units (for example, sandstone, siltstone, mudstone), or pinch out. Coal splits into two or more beds that gradually thin or pinch out or interfinger with clastic rocks. Thus, a “zone” or an interval of coal beds and interbedded rock units will contain coal in one or more beds, and this “zone” is more correlatable over a wide area than are the component coal beds.

The term coal zone is used in this investigation to define related coal beds that are in stratigraphic proximity to each other but may not be all as one unit. Coal-zone names are either adopted from accepted nomenclature or are selected from the coal-bed names that are inclusive, from bottom to top of the interval, of contained coal beds. Figure DB-1 is a diagrammatic representation of variation of a coal zone that reflects the study areas in the Powder River, Williston, Hanna, Carbon, and Greater Green River Basins. A coal zone in these basins typically contains from one to as many as eleven coal beds (for example, Wyodak-Anderson coal in the Powder River Basin). As exemplified by the Wyodak-Anderson coal zone in the Powder River Basin, the coal beds, from west to east, are split by fluvial deposits, merge into one thick bed, resplit, and remerge into another thick bed (fig. DB-1). In addition, the associated coal beds thin westward and pinch out or abut against “want areas” or fluvial channel deposits. In this example, separate coal beds may exist within the coal zone at some places, but some beds may be missing elsewhere. In addition, a coal zone in one area may be interconnected with the same coal zone in an adjoining area by either a coal bed at the top, in the middle, or at the bottom of the coal zone (fig. DB-2). Thus, the laterally juxtaposed coal zone forms either a series of “onlapping” or “zigzagging” patterns (fig. DB-2) throughout the basin of deposition. Consequently, because of very complicated correlation and non-correlation of coal beds within the coal zone, the coal resource assessment is more easily performed on the coal zone rather than on the coal bed. The continuity of the coal zone over a large area makes it more amenable to more accurate calculations of coal resources than would assessment of individual beds.

CORRELATION
The coal zone is established in order to correlate in detail the various coal beds and associated rock types. Correlation of coal beds is guided by the regional and local geologic structures and lithofacies association or depositional settings of associated rock types. That is, differences in depths of coal beds may be explained by the structural dip of rocks and/or by structural faults and folds (Whitacker and others, 1978; Broadhurst and Simpson, 1983; Weisenfluh and Ferm, 1984). In addition, occurrences of coal beds near the surface may be controlled by ancestral (for example, glacial) and modern (for example, river) erosion. More importantly, the environments and related processes at the time of deposition or accumulation of the coal beds (as peat deposits) influenced their lateral extent or continuity (Wanless, 1955; Ferm, 1970; Ferm and Staub, 1984; Flores, 1986). That is, highly dynamic environments such as deltas and rivers, which laterally switched back and forth (avulsion process) during their existence, are prone to develop very discontinuous associated peat swamps that form coal. This process of avulsion makes correlation of coal beds difficult unless it is aided by very closely spaced drill holes. In general, inactive areas associated with these environments, such as floodplains, interdeltas, and abandoned deposits, are prone to development of associated laterally extensive peat swamps and subsequent coal beds. This allows uncomplicated correlation of coal beds requiring less closely spaced drill holes. However, when these environments were overrun by rivers and/or deltas, the accumulation of associated peat deposits was interrupted, resulting in nondeposition of coal beds. Once these rivers and deltas shifted elsewhere (for example, topographic low areas) the same area may have been reoccupied by peat swamps and the coal deposition may resume. These processes may explain the variable continuity of the coal beds as well as their splitting and merging to form coal zones within the same area of deposition. When these processes are repeated contemporaneously basinwide, the result is “onlapping” and “zigzagging” of coal beds and zones. Thus, continuity of coal beds and/or zones and associated rock types is a function of their depositional environments, and the degree of reliability of correlation is determined by the spacing of drill holes that indicate the lateral variations imposed by these geological factors.

Palynology has been applied throughout the assessment region to provide a biostratigraphic (“palynostratigraphic”) framework for correlation of coal beds and zones. A palynostratigraphic zonation was developed from reference sections in selected outcrops, coal mines, and cores that are correlated to subsurface drill-hole data. The Fort Union Formation and equivalent rocks were divided into six palynostratigraphic zones designated P1 (lowermost Paleocene) through P6 (uppermost Paleocene) by Nichols (1994; 1996). The palynostratigraphic zonation is the basis for age determinations of individual coal beds and zones, and for correlations of coal-bearing rocks between basins in the assessment region.

COAL ANALYTICAL DATA
Each coal sample locality and/or drill hole is identified by a unique number and name. When possible the original location number and name were retained from the drill hole where the coal sample was collected. Location coordinates were recorded in Universal Transverse Mercator, latitude and longitude, and state plane coordinate systems. The coal analytical data consist of proximate and ultimate analyses from drill holes and mine locations. Guidelines for sampling to ascertain the rank and chemical, mineralogical, petrographic, and geophysical and physical properties of coal are discussed by Swanson and Thompson (1976). The proximate analysis includes the moisture, volatile matter, fixed carbon, and ash contents (on an as-received, moisture-free or dry, and ash-free basis). The ultimate analysis consists of the hydrogen, carbon, nitrogen, sulfur, oxygen, and ash contents (on an as-received, moisture-free or dry, and ash-free basis). The calorific or heat value (Btu/lb), forms of sulfur (sulfate, pyritic, and organic), and chemistry/mineralogy of the ash (for example, aluminum, calcium, manganese, potassium, silicon, and/or sodium oxides) are also included. Prescribed methods for analyses of these physical and chemical properties of the coal are discussed in American Society for Testing and Materials (1997).

The geochemical dataset includes the 12 trace elements of environmental concern named in the 1990 Clean Air Act Amendments (antimony, arsenic, beryllium, cadmium, chromium, cobalt, lead, manganese, mercury, nickel, selenium, and uranium). These data were entered in a spreadsheet format and each data entry includes the name of the coal bed or zone from which the coal sample was collected. The coal quality and geochemistry data from each sample location was linked to either the actual drill hole from which the sample was taken or to the nearest stratigraphic drill hole using the point identification number and the name of the coal bed/zone. This method relates the coal quality and geochemistry data to the coal stratigraphy, permitting spatial analysis and ultimately the use of coal quality (for example, rank, total sulfur, SO 2 per million Btu) in reporting the coal resources.