This one is built from Wyoming's and Montana's own filings and nothing else. Where there are no wells there is no colour, because a smooth surface drawn across an empty township is an invention that becomes invisible the moment it is shaded in. Every cell carries the number of wells behind it.
Survey grid and surface ownership from the Bureau of Land Management, provided as is. In BLM's words, these data are neither legal documents nor land surveys and must not be used as such: the grid is drawn for reference, not to determine anybody's corner. Sections and quarter-quarters load for the townships on screen, so zoom in to see them. Surface ownership shows which agency administers the surface, and says nothing about who owns the minerals underneath.
Filled circles carry a number. Dashed circles are wells whose decline fit was refused, so we will not claim a recovery for them. Plain hollow circles are wells with a sound fit that simply cannot be normalised, because the perforated interval was never filed. All three are drawn: a map showing only the wells that behaved is a map that has quietly chosen its own evidence.
Drag to rotate, scroll to zoom, hold shift to pan. Click a wellhead to isolate that well and its pad. Each mark is a formation top at the depth an operator filed for it; each line is a wellbore, following its filed survey where one exists and vertical where none does. The ground is the USGS bare-earth elevation model on a … grid, with the township lines laid on it, and it can wear a basemap: aerial imagery, the USGS topographic map, or OpenStreetMap. Hover a wellhead or a top for what it is. Every switch that changes what is drawn is in the legend below.
This is a structural and performance model. It is not a reservoir model, and the difference is most of the list below.
Everything above comes from the Wyoming Oil and Gas Conservation Commission and can be checked against the same filings. The written study sets out the method.
Surfaces are available and are off until you ask for them, because a surface joining one well's Niobrara top to the next well's is a statement about ground nobody drilled. Switched on, each one is a weighted average of the tops around it, and it is drawn only where there is well control: at least three tops within eight kilometres, the nearest within four. Everywhere else stays empty, which is why the surfaces have ragged edges and holes. Across ten formations, 17,251 cells could be computed and 42,269 could not.
Colour fades toward the page as the nearest well gets further away, so thinly supported ground looks thin. Inverse distance weighting is used rather than anything smoother on purpose: smoothness is the quality that makes an interpolated surface look like knowledge.
Two other limits are worth stating plainly. of the wells here have no directional survey on file and are drawn as vertical lines, which is wrong for any horizontal well but is what the filings support. And the surveyed paths are thinned before they reach your browser, so a path is the shape of the filed survey rather than every station in it.
A formation in the legend governs two things at once: the tops filed at that horizon, and the wells completed in it. They are the same subject seen twice - what a well passed through and what it was perforated in - and having them on separate switches meant a layer could be turned off while the wells producing from it stayed on the screen. Completions come from the filed producing intervals: … wells are matched to a formation drawn here, … of them completed in more than one, … have no producing formation on file, and … produce from something this view does not draw because too few tops were filed for it. The last two groups are not folded in with the rest; they have their own row, so no switch claims to control more than it does.
The ground surface is the one surface here that is not an inference. It is the U.S. Geological Survey's bare-earth elevation model, built largely from airborne lidar and served at one metre, resampled onto a … grid because what it is wanted for is a datum under a column of formation tops rather than a terrain analysis. Bare earth means the ground, not the trees or the buildings on it, and anything needing the real resolution should go to 3DEP rather than to this page.
Having a measured ground makes the filings checkable against it. Well by well, across the … wells with both a filed elevation and a grid cell under them, the two agree in the middle and not in the tails: the median difference is …, the quartiles are … and …, and … wells differ by more than 100 m. A hundred metres is not a survey error. In those wells either the filed elevation or the surface coordinate the well is mapped at is wrong, and this site does not yet know which; the disagreement tracks neither survey coverage nor lateral length. They are drawn anyway, because leaving them out would be choosing the evidence.
Each wellbore hangs from the elevation filed with its own tops rather than from this grid, even where the two disagree, because that is the datum those tops were computed from and a well has to be internally consistent before it is consistent with anything else. … wells had no usable filed elevation and hang from the 3DEP ground instead.
The township lines on that surface are the Bureau of Land Management's survey grid, draped by sampling the ground under each step along the line. In BLM's words these data are neither legal documents nor land surveys and must not be used as such, and putting them on a three-dimensional ground makes them look more like a boundary than ever, so it bears repeating: this is the survey grid drawn for reference, not a determination of anybody's corner.
Vertical exaggeration is on by default because the basin is about a hundred kilometres across and a couple of kilometres deep, and at true scale the layers collapse into a line. Exaggeration distorts dip and thickness. The figure is shown above and can be set to 1.
Gossans is a research project built on Wyoming's public records, and the company is not formed. If you have found something on these pages that is wrong, that is the email worth sending.