Gravity Interpretations Using Causal 3D Geology Models VS. the Single Density Complete Bouguer

AREVA Canada, now known as Orano Canada, hired and expected me to use gravity data targeting blind uranium deposits buried under the glacial overburden and lakes in northern Saskatchewan, Canada.

Illustrating my successful interpretation philosophy is my interpretation over the Midwest project blind uranium deposit demonstrating the practical use of causal but straightforward 3D geology models for gravity interpretation. Below, is the Residual Free-Air Anomaly (RFAA) contour map, yes contours, over the published uranium ore grade map. These anomalies are targets!

The residual free-air anomaly (RFAA) contour map over the official uranium ore grade map. These anomalies are targets! The ore grade map is the result of many years of drilling data and published in an official Midwest Project Technical report. The 3D geology interpretation I used to model the observed free-air gravity did not use the ore grade data or the Mackenzie dike cross-cutting through the ore zone. The blue outline is the shoreline of the Mink Arm of the South McMahon Lake. Note: there are gravity stations on the lake!

The previous gravity interpretation, the Complete Bouguer Anomaly (CBA), it is evident that textbook processed CBA map did not definitively target the blind uranium ore zones in the Athabasca Basin.

The CBA contour map over the ore grade map for the Midwest Deposit:

The published CBA contoured. The original gravity survey was conducted because early-on in the drilling program there were uranium intercepts and they believed the gravity data would help delineate the full extent of the ore zone. The underlying ore grade map is the result of many years of drilling data and published in an official Midwest Technical report. It is evident that the CBA did not come close to definitively delineating the uranium ore zone. The blue outline is the shoreline of the Mink Arm of the South McMahon Lake. Note there are gravity stations on the lake!

To remind everyone what the single density Bouguer geology model looks like:

This figure is a block cut from the Midwest project CBA geology model. The vast majority of textbooks use a single density for the CBA (YES, the Bouguer density is a surrogate for a single layer geology model) from the topographic surface to a reference elevation. Traditionally the Bouguer density is 2.67 g/cc, and the reference elevation is mean sea level, zero elevation.

The CBA single geology model includes the drumlins, lakes, muskeg, Athabasca Sandstone, and a portion of the basement, at least in the Midwest area. Yes, this TRULY is a realistic geology!?!?!? Moreover, all because it is written in the textbooks that this is the way all gravity data must be processed.

I personally believe the majority of geophysicist abdicate their responsibility to build 3D geology models for gravity interpretations by claiming “it's too hard” or “we don’t know what the geology is.” Yes, it takes work, but it's our job! Moreover, just because “it is written” that the CBA is the only way to perform interpretations does not absolve us of our responsibility to test the calculated free-air gravity effects of the known and expected 3D geology models against the observed free-air gravity.

Therefore, comparing the RFAA map with two distinct TARGET anomalies to the CBA map, without any distinct target anomalies, why would any exploration manager spend exploration budget SOLELY on any of the CBA “anomalies,” and given there are multiple anomalies of similar magnitude, and the “correct” sign of the anomaly should be in question.

Yes, I understand that in the 1970’s when the gravity survey was conducted there were VERY few digital computer programs to build or calculate the gravity effect of any 3D geology models or the limited extent of terrain models available in the WORLD. Moreover, at that time the terrain “correction” would most likely have been done by “hand” using topographic contour maps by a person applying their subjective eye estimating the gravity effect of the terrain.

Also, integrating multiple CBA datasets brings up the problem where there can be various, and potentially inconsistent methods used to calculate the CBA, using different topography models and human and software interpreters.

Well, it’s the 21st century, and what does the geology look like now? In reality, it’s pretty much as was in the 1970’s or late 1800’s. However, now we can build and calculate the gravity effect of simple causal 3D geology models of the area to target the unconformity hosted uranium ore zones.

Perspective aerial view of the Midwest project area. The picture shows a view of the surface geology of northern Canada! Spruce forest on glacial deposits with lakes, streams, and muskeg, aka bogs (not swamp) over the Athabasca Sandstone, over the unconformity, over the granite and meta-sediment basement. As I pointed out earlier, the gravity survey was acquired on the lake, meaning the lake was frozen!

The sheer number of lakes, ~36000, within a 132 km radius of the Midwest project is almost overwhelming. Moreover, in 2007 only 16 lakes in the Athabasca Basin had published bathymetry.

Saskatchewan Geological Survey’s model of the unconformity/basement topography with Natural Resources Canada’s aeromagnetic map overlaid. My 2008 basement topography map did not have the Pasfield Lake astro-blem elevation data in the eastern Athabasca Basin.

Then there is the unconformity topography. The unconformity is also the bottom of the Athabasca Sandstone and is the basement topography. Between late 2007 and early 2008, I produced a regional digital elevation model (DEM) of the Athabasca basin unconformity. I could also use the unconformity DEM as the bottom of the Athabasca Sandstone.

Additionally, for individual project areas, I developed regionally consistent but project specific detailed 3D geology models from the surface topography to the unconformity and basement by integrating project geology data from each project geologist. Yes, all of the project geologists have their 3D geology interpretations, but they don’t necessarily share unless you ask and are ready to integrate their model into the gravity interpretation. I believe the lack of openness with me at the beginning was due to outright and justified animosity toward the geophysics team in addition to wanting to “see” what I could produce with the geology data they did share. After all, until I had asked for geology data, none of the geophysics team had ever asked for geology models from the geologist.

So, how did I go about building the 3D geology interpretation that showed the ore zone targets in the RFAA?

The complete 3D model from the topographic surface to the Athabasca Sandstone to the unconformity/basement to the bottom of the model at zero elevation.

The first of three 3D geology model layers consists of the lakes and overburden down to the top of the Athabasca Sandstone. The lakes and overburden are distinct models at multiple densities. The lakes are placeholders that are reasonably sized, and the inversion returns the equivalent mass necessary to accommodate each lake, and the overburden is considered to be a single density layer also supplying the mass essential to minimize the misfit between the observed and estimated free-air gravity.

Glacial overburden, lakes, and muskeg down to the top of the Athabasca Sandstone.

The second layer is the Athabasca Sandstone. The Athabasca Sandstone is assumed to be a single density layer at 2.54 g/cc.

The model of the top of the Athabasca Sandstone is also the bottom of the overburden.

The surface topography of the top of the Athabasca Sandstone uses reasonable yet straightforward geologic assumptions. The first assumption is that larger lakes are on average 10 meters deep, as is Mink Arm, and the bottom surface of the lakes are at the top of the Athabasca Sandstone, the second assumption, the streams interconnecting the lakes are two meters deep, and their base is at the top of the Athabasca Sandstone. Then gridding these expected model elevations of the bottom of overburden elevation combined with diamond drill hole intercepts at the top of Athabasca Sandstone are used to model a regional top of Athabasca Sandstone surface.

The bottom of the Athabasca Sandstone is the unconformity and is also the top of the granite/meta-sediment basement.

In 2008 there were a fair number of diamond drill-hole intercepts of the basement but only at uranium exploration sites, along with seismic refraction surveys from the 1960’s. By integrating these different datasets, I produced a reasonable regional DEM of the known and expected elevation of the unconformity surface. Along with a simple two geology, granite and meta-sediment basement geology model based on the total magnetic anomalies first vertical derivative zero crossing. Since 2008 the Saskatchewan Geological Survey have published their unconformity elevation model of the Athabasca Basin unconformity. There is a significant improvement between my elevation model and the SGS model at the Pasfield Lake area where they now have elevation data to model a second major astro-blem in the Athabasca Basin where I did not have definitive sub-surface data.

The unconformity surface, this is the bottom of the Athabasca Sandstone and the top of the basement. The basement model uses granite and meta-sediment geology model (two density.) The aeromagnetic data define the boundaries between the granite (reds) and meta-sediment (blues.) The bottom of the basement model is at zero elevation.

The base of the Athabasca Sandstone is the unconformity and is also the top of the granite and meta-sediment basement. The basement geology model consists of granite bodies, reddish color, and meta-sediment bodies, blueish color. The contacts between the granites and meta-sediments are the first vertical derivative zero contour.

The bottom of the model is a horizontal surface at zero elevation and has the same granite and meta-sediment boundaries as the top of basement surface. Yes, the exact topology of the contacts between the granite and meta-sediments is unknown, but the density inversion recovers each granite and meta-sediments body mass the same it recovers each lake’s mass.

The geologic cross-section through the ore zone at the red line shows the unconformity uranium deposit and the alteration zone in the Athabasca Sandstone.

The 2D cross-section through the Midwest uranium deposit reveals the reason for the large residual free-air anomalies over the ore zones due to the un-modeled low-density alteration zones above the Midwest uranium deposit. Also, the break in between the two anomalies is due to an un-modeled Mackenzie diabase dike cutting through the ore zone and extended more than 500 km to the northwest.

The takeaway; building one or more consistent but straightforward 3D geology models of the known and expected geology, then inverting the model for the estimated rock densities that minimize the misfit between the observed and estimated gravity data at the data locations. Also building the interpretation should include enlisting the help of one or more knowledgeable geologist and if possible testing their 3D geology models. Then by making these consistent yet straightforward 3D geology interpretations, you will be continually surprised with what shows up in the interpretation results that you didn’t know was there. Additionally by enlisting contributions from the geology team and recovering the subtle geologic effects that are not in the geologic model, but the geologist know are there in their geology helps with your credibility. These subtle effects, and sometimes not so subtle, shows that the gravity data has more information embedded in it than has previously been illuminated by dogmatic rote textbook data processing steps.