Successfully utilizing satellite imagery in remote sensing applications, such as the gathering of intelligence information for agriculture, environmental and resource management, is driven not only by the technology to extract and analyze data but also the ability to choose which satellite or sensors are the best fit for a particular application. While many vendors are limited to using a single low resolution source, GDA is the only commercial vendor with the technology expertise to utilize moderate, medium, and high resolution imagery on a global scale and also the ability to utilize multiple sensors and resolutions simultaneously. This is critical when the goal is to maximize accuracy while controlling costs. As shown in the graph on the right, the available medium and high resolution satellites are plentiful and having the ability to determine which satellite or combination of satellites provide the best resolution/cost ratio for a particular application is a major technology advantage for GDA.  

The satellite imagery delivered from the various satellites in use today vary in four key areas. The size of the footprint or swath as shown pictured left for the P6-AWiFS sensor. The resolution of the sensor, which defines the pixel size and therefore the detail and accuracy available. The number of spectral bands available to gather information and finally the temporal availability of the imagery (i.e. frequency imagery is available for particular location). Understanding and maximizing these four key areas are critical for success in remote sening applications utilizing satellite imagery.

A general categorization of satellite sensors by resolution is given below:

• Low resolution, 1km - 10km, suitable mostly for weather, typically free
• Medium resolution, 100m -1km, suitable for > 1:250,000, typically free to low cost
• High resolution, 10m -100m, suitable for 1:50,000 - 1:250,000, medium to high cost per scene
• Very High resolution, 1m - 10 meters, suitable for < 1:50,000, high to very high cost per scene

A categorization of satellite sensors by resolution for use in agriculture and environmental analysis is given below:

• Low resolution, 100m - 1km, suitable for large area NDVI and NDWI analysis, typically free
• Medium resolution, 30m -100m, suitable for area assessment and plant analysis on large scale, medium to high cost
• High resolution, 1m -25m, suitable for area assessment and plant analysis on small scale, high to very high cost

Satellite imagery, in general, is not suitable in its raw form, and must undergo processing for data extraction and analysis. As an example consider imagery from the IRS-P6 satellite which contains multiple sensors including the Advanced Wide Field Sensor (AWiFS), which has a swath path width of 740km and a resolution of 56m. This satellite sensor offers a unique combination of medium resolution with a large swath path thus providing a good resolution/cost ratio. However, due to the wide swath, varying angles from the sensor, varying angles of sunlight and even the curvature of the earth, the image becomes distorted the further a part of the image is from the nadir. A graphical representation is shown to the right. In order for imagery from this sensor to be used in agriculture or environmental analysis this distortion must be removed. Other challenges associated with processing satellite imagery include unpredictable natural occurrences such as cloud cover due to weather, shadows due to blocking of the sunlight that can cause distortion of the image/spectral data. 

This is where GDA's technology shines. While some of these distortion factors can be removed manually, this is a labor intensive process that is both costly and provides a high potential for error, even for a single image. Much of GDA's efforts under the Small Business Innovative Research (SBIR) awards were focused on the automation of this image processing allowing for the use of medium and high resolution on a global scale. GDA developed a library of proprietary image processing algorithms which are used to generate:

• SR: True Surface Reflectance Products
  
• SASR : Standard Atmosphere Surface Reflectance Products
• CASA : Cloud And cloud Shadow Assessment Maps
• TOPONORM : Topographically Normalized Imagery
• USFE : Under Shadow Feature Enhancement
• TIMESTAND : Radiometric Standardization of Multiple Scenes for Analysis and Mosaicing