Explorers look toward better remote sensing data

Remote sensing for oil and gas exploration seems on the brink of the availability of much more precise spectral data. The U.S. Navy has let a contract for development, launch, and early operation of a satellite that could materially aid oil and gas exploration companies, among other objectives. The Geosat Committee has proposed a "group shoot" to begin obtaining and interpreting highly sensitive hyperspectral data from an aircraft. Mining companies have been using hyperspectral data since about

March 9, 1998

5 min read

G. Alan Petzet
Exploration Editor

Remote sensing for oil and gas exploration seems on the brink of the availability of much more precise spectral data.

The U.S. Navy has let a contract for development, launch, and early operation of a satellite that could materially aid oil and gas exploration companies, among other objectives.

The Geosat Committee has proposed a "group shoot" to begin obtaining and interpreting highly sensitive hyperspectral data from an aircraft.

Mining companies have been using hyperspectral data since about 1984. These data are only now finding acceptance and implementation in the oil industry.

Hyperspectral data are valuable in oil and gas exploration because they provide information on rock properties that permits inferences about geologic processes that relate to the presence of oil and gas in the subsurface. Other applications include environmental baseline studies, loss reduction, and emergency spill detection and response.

Satellite project

The Office of Naval Research let the contract in December 1997 to Space Technology Development Corp., Arlington, Va.

The satellite, to be launched in 2000, would carry hyperspectral and panchromatic sensors in a near polar, low earth orbit that would enable it to overfly each point on earth at least once every 7 days.

Dubbed NEMO, for Navy EarthMap Observer, the satellite will deploy hyperspectral imaging technology, data analysis and compression, and other technologies developed by the Naval Research Laboratory. The Navy will use the satellite for scientific and logistical needs that relate to the world's coastal areas. It will be able to map the ocean floor in as much as 100 ft of clear water.

Science Applications International Corp., San Diego, will build the sensor system from NRL concepts and specifications. Cost of the NEMO program is $128.9 million, including 3-5 years of satellite operation, STDC said. STDC and its industrial partners and suppliers are to provide $58 million of that with the rest coming from the government.

NEMO concepts

STDC said a modified Globalstar rocket would be used to launch NEMO. As designed, NEMO would add spectral and spatial dimensions to the Landsat information much used by exploration companies.

STDC was set up in 1994 to commercialize space related technologies developed in government laboratories. Part of that effort is to obtain hyperspectral imagery of land areas for sale to users that include mining and oil and gas companies and environmental and agricultural entities.

The NRL will build two main subsystems: the imagery onboard processor and a data analysis and compression system called Orasis. Compression of the data will make possible its management and transmission to earth processing stations.

The hyperspectral sensor will have 30 m spatial resolution, 0.4-2.5 microns spectral range, 210 channels, and bandwidth of 10 nanometers/channel. The panchromatic sensor will have one channel, 0.5-0.7 microns spectral range, and 5 m spatial resolution. Both sensors will have a 30 km swath width.

Risks, payoffs

Oil and gas exploration people welcomed the prospect of the availability of hyperspectral data but noted the unsuccessful launch attempts of Landsat 6 and several other satellites (OGJ, Feb. 1, 1993, p. 51). Another satellite, Landsat 7, is pegged for launch this fall.

Traditional remote sensing data come from visible and near-infrared light, said David Koger of Koger Remote Sensing, Fort Worth, and president of the Geosat Committee. Hyperspectral data sample a greater range of light and capture it in more discriminate samplings, or narrower bands.

Thematic Mapper, the most popular existing satellite, yields data in seven bands. Using this multispectral (MS) data, explorers can detect and map the surface indication of structural effects and subsurface events, Koger said. Using MS data it is sometimes possible to detect apparent alterations of rocks, soils, and vegetation caused by microseeping hydrocarbons.

Hyperspectral data, on the other hand, have upwards of 200 bands and can differentiate the chemical composition of rocks from their color spectra. For instance, differences in clays can be an indication of hydrothermal alteration that takes place in minerals formation processes.

High end personal computers are capable of processing the data, said Tod Rubin of Geophysical and Environmental Research Corp., Millbrook, N.Y. GER manufactures hand-held and airborne hyperspectral units and leases and flies two survey aircraft worldwide.

Koger said a geologist can make better geologic maps with hyperspectral data than a geologist in the field could make. The field geologist observes a light range of 0.4 to 0.67 microns, while the typical hyperspectral instrument sees a range of 0.4-2.5 microns. More subtle differences in clays will be apparent.

Research projects

The Geosat Committee, a research partnership of companies, government agencies, and universities, is seeking 8-12 industry partners to support a hyperspectral "group shoot" project. Geosat recently moved headquarters to the University of Texas-El Paso from Norman, Okla.

The combine would probably use the TRW imaging spectrometer to replace data that would have been provided by the failed NASA-Lewis satellite, Koger said.

Test sites will be in the western U.S., probably where Geosat has libraries of data from the Landsat series, SPOT, the Jet Propulsion Laboratories/U.S. Geological Survey Aviris hyperspectral scanner, and large amounts of ground truth, or surface confirmation of remotely sensed data.

The $100,000-200,000 project is to include mobilization; radiometrically, atmospherically, and geometrically correcting the data; and on-the-ground collection of spectrometer data coincident with the overpass of the airborne scanner.

Elsewhere, Robert D. Jacobi at State University of New York-Buffalo is involved in two projects that will compare hyperspectral and more traditional remote sensing methods. Both projects are funded by New York State Energy Research & Develoment Authority, Albany.

GER will fly an airborne hyperspectral survey in search of gas reservoirs in western New York in one project. In the other, Quest Energy Inc., Amherst, N.Y., will combine traditional satellite imaging with subsurface mapping to identify gas reservoirs in Cattaraugus County, N.Y.