11.2 What Are the Capabilities of a Satellite Sensor?

A satellite sensor has four characteristics that define its capabilities: spatial resolution, radiometric resolution, temporal resolution, and spectral resolution. In Chapter 10, we discussed the concept of resolution in the context of spatial resolution, with the area on the ground represented by one pixel in a satellite image. A sensor’s spatial resolution will affect the amount of detail that can be determined from the imagery—a sensor with a 1-meter spatial resolution has much finer resolution than a sensor with spatial resolution of 30 meters (see Figure 11.5 for examples of the same area viewed by different spatial resolutions). A satellite’s sensor is fixed with one spatial resolution for the imagery it collects—a 30-meter resolution sensor can’t be “adjusted” to collect 10-meter resolution images.

If two bands (one with a higher resolution than the other) are used by the sensor to image the same area, the higher-resolution band can be used to sharpen the resolution of the lower-resolution band. This technique is referred to as pan-sharpening because the higher-resolution band used is a panchromatic band. As we discussed in Chapter 10, in terms of imagery, a panchromatic sensor will measure only one large band of wavelengths at once (usually the entire visible portion of the spectrum or the entire visible and part of the near-infrared spectrum). For example, a satellite (like Landsat 7) that senses the blue, green, and red portions of the electromagnetic spectrum at 30-meter resolution could also have a sensor equipped to view the entire visible portion of the spectrum and part of the near-infrared spectrum as a panchromatic band at 15-meter resolution. This panchromatic image can then be fused with the 30-meter resolution color imagery to create a higher-resolution color composite. Many satellites with high spatial resolution capabilities sense in a panchromatic band that allows for pan-sharpening of the imagery in the other bands.

As we mentioned in Chapter 10, a sensor scales the energy measurements into several different ranges (referred to as quantization levels) to assign brightness values to a pixel. The sensor’s radiometric resolution refers to its ability to measure fine differences in energy. For instance, a sensor with 8-bit radiometric resolution places the energy radiance values on a scale of 0 (lowest radiance) to 255 (highest radiance). The wider the range of measurements that can be made, the finer the sensor’s radiometric resolution is. The radiometric resolution (and the resulting number of levels) is a measure of the number of the sensor’s bits of precision. To calculate the number of levels, take the number of bits and apply it as an exponent to the number 2. For instance, a 6-bit sensor would have 64 levels (2⁶) ranging from a value of 0 to a value of 63. An 8-bit sensor would have 256 levels (2⁸) ranging from 0 to 255. An 11-bit sensor would have 2048 levels (2¹¹) ranging from 0 to 2047.

The finer a sensor’s radiometric resolution, the better it can discriminate between smaller differences in energy measurements. For example, a 2-bit sensor (2² or four values) would have to assign every pixel in an image a brightness value of 0, 1, 2, or 3, which would result in most pixels having the same value, and it would be difficult to distinguish between items or to create effective spectral signatures. However, an 11-bit sensor would assign a value anywhere from 0 to 2047, creating a much wider range of value levels and allowing much finer distinctions to be made between items. See Figure 11.6 for examples of differing effects of radiometric resolution on an image when performing remote sensing.

The sensor’s temporal resolution refers to how often the sensor can return to image the same spot on Earth’s surface. For instance, a sensor with 16-day temporal resolution will collect information about the swath containing your house and not return to image it again until 16 days have passed. The finer a sensor’s temporal resolution, the fewer days it will take between return times. A way to improve a sensor’s temporal resolution is to use off-nadir viewing capability, in which the sensor is not fixed to sense what’s directly below it (the nadir point). A sensor that is capable of viewing places off-nadir can be pointed to image locations away from the orbital path. Using off-nadir viewing can greatly increase a sensor’s temporal resolution, as the sensor can image a target several times during its orbits, even if the satellite isn’t directly overhead. This capability is a great help in monitoring conditions that can change quickly or require rapid responses (such as mitigating disasters resulting from hurricanes, tornadoes, or widespread fires). See Hands-on Application 11.2: Seeing What the Satellites Can See for an online tool that shows how much area on the ground a satellite can image (and the extent of any off-nadir capabilities).

The sensor’s last parameter is its spectral resolution, which refers to the bands and their wavelengths measured by the sensor. For instance, a sensor on the QuickBird satellite can measure four bands of energy simultaneously (wavelengths in the blue, green, red, and near-infrared portions of the spectrum), and the sensor onboard the Landsat 7 satellite can sense in seven different wavelengths at the same time. A sensor with a higher spectral resolution can examine several finer intervals of energy wavelengths (for example, the MODIS sensor onboard the Terra satellite—see Chapter 12—can sense in 36 different bands at once). Being able to sense numerous smaller wavelengths in the near-infrared portion of the spectrum gives much finer spectral resolution than just sensing one wider wavelength of near-infrared energy.