5.12.3. Radiometric Resolution

Radiometric resolution is determined by the number of discrete levels into which reflected radiations may be divided (quantization) by a sensor. With a given spectral resolution, increasing the number of quantizing levels (radiometric resolution) will improve the clarity/identification of the objects.

Radiometric resolution refers to the sensor's ability to distinguish differences in the intensity of the received signal from reflected radiation. A higher radiometric resolution means the sensor can capture a greater range of intensity levels. This allows for better detail and differentiation in the images captured by the sensor. For example, if a sensor has a radiometric resolution of 8 bits, it can record 256 levels of brightness. If it has a resolution of 16 bits, it can record 65,536 levels of brightness. The more levels a sensor can capture, the more gradations of shade it can exhibit, leading to clearer images.

Think of radiometric resolution like the shades of color on a painter's palette. If an artist only has three colors, their painting will have limited shades and will look flat. However, if the artist has a larger palette with more colors, they can create depth and subtlety in their work. Similarly, a higher radiometric resolution allows for more subtle differences in light intensity, leading to richer and more informative images in remote sensing.

Radiometric resolution depends on the wavelengths and the type of the sensor used. If the radiometric resolution is higher, the small differences in reflected or emitted radiations can be measured accurately, but the volume of data storage will be larger.

The significance of radiometric resolution lies in its ability to enhance the quality of data that remote sensing systems capture. A sensor's design and the wavelengths it operates in determine its radiometric resolution. Sensors that differentiate subtle variations in light intensity are crucial when examining complex objects or surfaces, such as vegetation types, soil properties, or water quality. However, capturing more data requires increased storage capacity and processing power because the larger number of intensity levels results in larger image files.

Consider radiometric resolution like the resolution of a photograph. A high-resolution photograph will show fine details like textures on a person's face or the subtle colors of a landscape. Conversely, a low-resolution photo may blur these details together. In remote sensing, high radiometric resolution makes it possible to observe intricate features on the Earth's surface accurately.

Most images used in remote sensing are 8 bit (i.e., 2^8 = 256), so in many examples that are often cited are related to 8 bit images. Generally, higher the bits, better is the image quality for interpretation.

Common radiometric resolutions include 6-bit, 7-bit, 8-bit, 9-bit, and higher. Each bit increase in resolution implies a doubling of the quantization levels of intensity. For instance, while a 6-bit image can depict 64 shades of grey, an 8-bit image can display 256 shades, and a 10-bit image can represent 1024 shades. This means images can show more subtle differences in features based on slight variations in reflectance, making the final interpretations more accurate.

Imagine you're cooking and trying to quantify the sweetness of your dish. If you only have a coarse measuring cup (like 6-bit resolution), you might only add three levels of sweetness: sweet, medium sweet, or not sweet. But if you have a fine measuring spoon (like 10-bit resolution), you can measure out very small amounts, allowing for a perfect level of sweetness that enhances the dish. Similarly, higher radiometric resolution helps create a more refined image that captures the subtle differences between objects.

Example of various images at different bits include Landsat-MSS [Landsat 1-3 provided 6 bits (64 grey values)], IRS-LISS I-III: 7 bits (128 grey values), Landsat-TM (from Landsat 4-5), SPOT-HRV: 8 bits (256 grey values), Landsat-ETM & ETM+ (from Landsat 6-7): 9 bits (only 8 bits are transmitted), IRS-LISS IV: 10 bits (only 7 bits are transmitted).

Different remote sensing instruments have varying radiometric resolutions. For example, Landsat-MSS in its earlier models used a resolution of 6 bits, whereas newer systems like the Landsat-ETM+ utilize higher resolutions of up to 9 bits. This difference impacts the levels of detail a satellite can capture and transmit back to Earth. Higher bit resolutions are particularly important for applications such as agriculture, where distinguishing between plant health levels is crucial.

Consider a TV with standard definition versus high definition. A standard definition TV may only show limited detail, making it difficult to see individual features, akin to lower radiometric resolution. In contrast, a high-definition TV can show every wrinkle in an actor's face or every leaf on a tree, similar to how higher radiometric resolutions can provide clearer and more detailed imagery in remote sensing.

--