Transformations - Spatial Mathematics: Theory and Practice through Mapping - page 73

3.4.3 Raster data resolution: Considerations
Four types of raster data resolution compose the “quality” or “scale” of raster
data. The first of the four types is spatial resolution. Spatial resolution refers
to the amount of land area that is represented by each pixel. A data set with
a 0.5-meter spatial resolution is finer than a data set with 10-meter spatial
resolution. Higher spatial resolution means that there are more pixels per
unit area, and is a result of how much of the Earth's surface the sensor on the
airplane or satellite “sees” with each image captured. The part of the Earth's
surface being sensed is known as the Instantaneous Field Of View, or IFOV.
The size of the area viewed is determined by multiplying the IFOV by the
distance from the ground to the sensor. If a feature on the ground is greater
in size than the resolution cell, then that feature can be detected. Conversely,
if the feature is smaller than the resolution cell size, it cannot be detected.
Therefore, a 10-square-meter lake would not be detectable from a sensor with
a 30-square-meter spatial resolution.
Spectral resolution refers to the number of spectral bands a data set has and
thus is only applicable to rasters that are created from sensors capable of
recording more than one band. This is usually confined to some aerial imag-
ery and most satellite imagery. It depends wholly on the ability of a sensor on
a satellite or aircraft to distinguish between wavelength intervals in the elec-
tromagnetic spectrum. A higher spectral resolution corresponds to a narrower
wavelength range for a particular band. Where a grayscale orthophotograph
has a low spectral resolution because it captures data from across the entire
visible range of the spectrum, hyperspectral sensors collect data from up to
hundreds of very narrow spectral bands.
Temporal resolution or the “revisit period” has to do with the frequency
with which images are captured over the same place on the Earth's surface.
Therefore, a sensor that captures data once every three days is said to have
a higher temporal resolution than multispectral scanners on the first Landsat
satellites, which could revisit the same place only every 18 days. Temporal
resolution depends on the satellite's latitude, its physical capabilities, and the
amount of overlap in the image swath.
Finally, radiometric resolution represents the ability of a sensor to discrimi-
nate between very slight differences in reflected or emitted energy. Image
data are generally displayed in a range of gray tones, with black repre-
senting a digital number of 0 and white representing the maximum value.
Image data are represented by positive numbers from 0 to one less than a
selected power of two. This corresponds to the number of bits used for cod-
ing numbers in a computer. All numbers in a computer are stored as binary
numbers—0s and 1s. The number of levels available depends on how many
computer bits are used in representing recorded energy. If a sensor used
8 bits to record the data, there would be 2 to the 8th power, or 256 avail-
able values, from 0 to 255. However, if only 4 bits were used, then only 2