Laser intensity mapping of outcrop geology.

 

Jerome Bellian

Bureau of Economic Geology

University of Texas at Austin

 

 

Laser outcrop mapping is being recognized as the future of outcrop field geology.  Laser mapping utilizes LIDAR (light detection and ranging) and various forms of image capture, laser intensity, traditional photography and digital imaging.  The concept is simple.   A laser is used to determine the position of points along an outcrop face in x, y, and z space.  If a sufficient number of laser points are gathered along an outcrop, the shape of the outcrop will be approximated by the point cloud.  The points can then connected into any type of mesh or 3-D grid to construct a surface in much the same way a seismic horizon is interpolated from interpretation on individual seismic lines.  Any image(s) can be applied to the surface as a texture.  The result is a 3-D surface model with a series of images fit on to the surface.  A “virtual outcrop” is born. 

 

In order to have a high-resolution virtual outcrop, first you need a high-resolution terrain model.  The tighter the point cloud is the better and bigger the terrain model will be (figure 1).  Some high-end laser scanners collect not only x, y, and z data but also the intensity of the reflected laser pulse.  The result of laser intensity plus x, y, z data looks like a black and white photo without sun shadows (figure 2).  The intensity “black and white” can be used to rectify and “cookie-cut” another image onto the model, a digital photo for instance that has been blended with laser intensity (figure 3).  All of these techniques can be done with little or now trouble as long as the data set in question is small enough to fit on your computer.  A standard texture-mapped VRML model works well across platforms in most web-browsers for mass consumption.

 

The “quality” of the model is relevant to the scale of the outcrop under investigation and the purpose for which the data are to be used.  If the desired output model needs to be at 1 cm point spacing and the outcrop is 3 km long and 100 m high, the file size of only the x, y, z and intensity data is 42 gigabytes binary!  The obvious solution then is to decimate your terrain data.  Our procedure extracts the full resolution data from the intensity, splits off the x, y, z data, optimizes the x, y, z data and then rejoins them together as a single data set.  This process can take a 42 gigabyte dataset down to 20 megabytes while preserving great detail of the original outcrop in full navigable 3-D.

 

One of the ways we have been utilizing the data at the Bureau of Economic Geology at the University of Texas at Austin’s Jackson School of Geosciences is to treat the data in much the same manner as we would seismic attribute data.  Examine multiple attributes simultaneously and look for patterns mathematically to aid in bringing out the subtleties of the geology that might be overlooked otherwise.  One example is this data set which used a combination of first derivative (slope) of the outcrop as well as laser intensity to perform a multi-component classification (figure 4).  This technique distinguishes between thin sands and shale as well as differences in sand grain mineralogy.

 

Feel free to contact Jerome Bellian and the Bureau of Economic Geology at the University of Texas at Austin at Jerome.Bellian@beg.utexas.edu for more information on the Jackson School of Geosciences Lidar Program.

 

 

 

Figure 1.  Laser point cloud data set at 1 cm xyz point spacing.  Human in front near outcrop is approximately 6 feet tall.

 

Figure 2.  Full resolution laser intensity point cloud.  Human for scale in foreground near outcrop is approximately 6 feet tall.

Figure 3.  Laser intensity image blended with a digital photo and applied as a texture to the xyz model.  Human for scale in foreground (removed from image but the outline is approximately 6 feet tall).

 

Figure 4.  This is a multi-component analysis of laser intensity data and first derivative (slope) map.  The dark red and dark blue bands running nearly horizontal across the image are two sand beds that look similar but have very different mineralogy.  Human for scale in foreground (removed from image but the outline is approximately 6 feet tall).

 

Oil IT Journal contributed paper 2003/001

January 2003

www.oilit.com

info@oilit.com