CG::Remote Sensing

Remote sensing, in its simplest sense, is the viewing of an object that is located at some distance from the observer. In a sense, you are remotely-sensing this web page, as you are separated from the computer monitor by several inches.

Your eyes are the sensor, gathering electromagnetic radiation, in the form of light. Your brain collects and analyzes this data, and converts that information into images of the everyday objects we see around us.

In the scientific sense, a device, such as a satellite or airborne scanner, is used to obtain electromagnetic radiation information about the target object. Computers are then used to process the data.

NASA has a Remote Sensing Tutorial available for public use.

Remote sensing continues to be an arcane science to the general public for a number of reasons. People are not familiar with satellite or sensor technology. Secondly, we tend to forget that there is more energy around us than our eyes can perceive. Thusly, as our eyes cannot sense this other energy, training our brains to interpret the information contained within the other wavelengths of energy can prove troublesome.

The nature of the data is also problematic. Remote sensing data is generally termed, "raster," that is, an image that is comprised of a matrix or grid of little squares. Each square represents some uniform area on the ground. The area covered by one of these square is referred to as, "spatial resolution." The area covered by one of these squares varies depending on the sensor used to acquire the image. Landsat 5 imagery, for example, has a resolution of 30m, meaning the length of a side of one of these squares is 30m. Citing another example, the Quickbird satellite has a resolution of 2.4m, meaning that the length of one side of one of these squares is 2.4m.

To further complicate matters, remote sensing also requires knowledge of the "spectral resolution" of the data. Spectral resolution refers to the different wavelengths of energy that are measured by the sensor. The human eye is basically sensitive to the visible wavelengths of the electomagnetic spectrum. We cannot see into the infrared, ultraviolet, thermal, x-rays, or the vast portion of the electomagnetic spectrum that is outside the visible range.

Finally, remote sensing data formats are very numerous. Commercial and governmental remote sensing efforts may have their own data storage format. Many countries have their own remote sensing assets and as a result may have their own unique data storage format.

Recently, web applications have been introduced that take advantage of the available remote sensing data resources. Google Maps and Google Earth, Microsoft's Virtual Earth, and NASA's WorldWind connect to servers and satellite imagery that enhance the utility of online mapping applications.

While many tools exist to help people construct a map, few people have readily accessible tools to manipulate remotely-sensed data. To those of you that do want to move in that direction, please find information contained herein about hardware, software, imagery vendors, imagery data, and organizations that assist those interested in remote sensing.

Hardware Tips:

1. Lots of RAM: This will speed processing and display, preventing the constant hard drive access that results from caching processes in virtual memory.

2. Lots of Hard Drive Space: The amount of hard drive space taken by software (the footprint) gets larger with each release or version. Huge amounts of digital data are available online that require storage space. Space is also needed for the working space and the 'live' data, the data which represents the latest good copy.

Sophisticated GIS users may opt for more complex hard drive arraignments, such as SCSI hard drives, or RAID arrays. Data access speeds are a little faster with SCSI hard drives. RAID provides for data recovery in the event of a disk failure.

3. Processor: Get the fastest one affordable. Most GIS is not written to take advantage of dual processors. However, dual processors create an environment whereby the GIS software can run on one processor, leaving the other(s) to run other applications. Dual-core technology is recommended; 64-bit apps appear to be problematic as of this writing.

4. Graphics Card: GIS software is graphics-intensive, especially if one gets into 3D rendering. Purchase an OpenGL compliant graphics card with atleast 128MB of RAM. Some users may opt for cards that support multiple monitors. Ensure the resolution, refresh rate, CPU frequency, and memory configurations will allow for the type of application.