Commercial satellite imagery plays an essential role across multiple fields, including agriculture, urban development, defense, and disaster response. Its usefulness is heavily dependent on its quality.

Several important characteristics must be evaluated to guarantee the imagery effectively serves its purpose. This article outlines the main factors that affect the quality of images typically ensured by commercial satellite imagery providers.

Spatial Resolution

Spatial resolution, or ground sample distance (GSD), defines the smallest object detectable in an image, impacting the level of detail.

High-resolution imagery (less than 1 meter) reveals fine details, making it ideal for tasks like urban planning or vehicle detection. According to scientists, “30cm is the best spatial resolution option available today for remote sensing using high-resolution commercial satellites.”

Medium-resolution imagery (1-10 meters) is helpful for environmental monitoring and land use classification, while low-resolution imagery (greater than 10 meters) suits large-scale applications such as weather or climate analysis.

“We’ll then often use imagery… whose higher spatial resolution gives us a better chance of seeing smoke. If we see smoke on the imagery at the same location, then it’s a pretty sure bet it’s an active fire.” — Gordon Seymour, GIS and Wildfire Data Technician; Canada’s Department of Environment and Climate Change.

Choosing the proper resolution for the current satellite imagery depends on your project needs, as higher resolution increases detail but also raises costs.

Spectral Resolution

Earth's physical and chemical surface qualities interact with electromagnetic radiation by absorbing, reflecting, emitting, and scattering it. Satellites capture these energy wavelengths to create spectral images.

Spectral resolution is the capability of a satellite sensor to capture data at different wavelengths of the electromagnetic spectrum. This feature is essential for recognizing materials, noticing changes in plant life, and evaluating water quality. A higher spectral resolution means more bands, which enables a more thorough analysis.

Multispectral imagery captures data in 3 to 10 broad bands, often including visible (red, green, blue) and near-infrared wavelengths, useful for land cover classification and crop health monitoring.

Hyperspectral imagery gathers data across hundreds of narrow spectral bands, providing detailed insights into surface chemistry. It enables precise mapping of minerals, identification of plant species, quick detection of crop diseases, and other applications, and is crucial for industries such as precision farming, forestry, and geology.

Temporal Resolution

Temporal resolution refers to the frequency at which a satellite revisits and captures images of the same location. High temporal resolution, with satellites revisiting areas daily or every few days, is essential for observing rapid changes like crop health, disaster recovery, or urban expansion. This frequent data collection allows for near real-time satellite view analysis and timely decision-making. 

Conversely, satellites with lower temporal resolution revisit areas only every few weeks or months, making them more suitable for long-term studies such as climate change monitoring. 

However, high spatial resolution often comes at the cost of lower temporal resolution, requiring a balance based on project needs.

Temporal resolution depends on the satellite's orbit, its sensors, and the width of its swath. Even if the satellite doesn't fly directly above the target area, the swath width can still enable the capture of recent satellite images, enhancing the temporal resolution.

Balancing high spatial and temporal resolution is challenging, as higher spatial detail often results in slower revisit times. New microsatellite constellations are addressing this by using clusters of small satellites for frequent, high-resolution imaging, while freely available options like Landsat and Sentinel offer lower imagery resolution but more regular revisits.

Radiometric Resolution

Radiometric resolution measures a sensor’s ability to detect small variations in electromagnetic energy, typically expressed in bits. It determines how many levels of brightness (or grey levels) a sensor can distinguish, with higher resolution indicating greater sensitivity to subtle energy differences. 

The higher the bit depth, the clearer and more detailed the image. For instance, an 8-bit image displays 256 intensity levels (where 0 is black and 255 is white), while a 12-bit image offers 4,096 levels, capturing more subtle variations.

High radiometric resolution is important for tasks requiring precise detail, such as monitoring plant health and soil moisture levels or noticing early signs of drought stress. In contrast, lower resolutions are sufficient for broader applications like land use classification, where precise intensity detection is less critical.

Geometric Accuracy

Geometric accuracy refers to the degree to which a satellite view of the Earth aligns with the actual geographic location on Earth's surface. It is crucial for ensuring that the image data can be reliably integrated with other geospatial information, such as maps or GPS coordinates. Any distortions in geometric accuracy can lead to errors in applications that require precise location data.

Orthorectification is a process of correcting geometric errors in satellite or aerial images caused by terrain changes, capture angles, or Earth's curvature. By using a digital elevation model (DEM) and accurate sensor data, it adjusts the image to represent the Earth's surface accurately. This ensures that each point in the image corresponds to its true geographic location, making orthorectified images essential for precise spatial tasks like mapping, city planning, and GIS.

Cloud Cover

Cloud cover significantly affects the quality of optical current satellite imagery by obscuring the Earth's surface, which is particularly problematic for applications like agriculture that require complete visibility. Clear-sky imagery is essential for detailed visual analysis, while radar satellites, such as those using SAR sensors, can capture cloud-tolerant imagery, making them suitable for tasks like real-time flood monitoring or disaster response. 

SAR images can be captured both at night and through clouds, providing more flexibility. Weather and seasons may impact optical imagery, but SAR ensures reliable observation, regardless of conditions, enhancing the overall accuracy of satellite data.

Signal-to-Noise Ratio (SNR)

The signal-to-noise ratio (SNR) shows how clear an image is compared to background noise. A higher SNR means a clearer image with less interference from noise or environmental factors. A low SNR can lower image quality and make analysis less accurate, especially in dim light or with weak signals.

Optical remote sensing images are often affected by noise, making SNR evaluation challenging, especially in heterogeneous regions with stronger feature interference. New approaches improve SNR estimation by calculating residuals in homogeneous areas, showing more stable and accurate results compared to traditional methods.

The quality of commercial up-to-date satellite images relies on factors like spatial, spectral, temporal, and radiometric resolution, geometric accuracy, cloud cover, and signal-to-noise ratio. Understanding these features is important for selecting appropriate imagery for applications such as precision agriculture, urban planning, environmental monitoring, and disaster response. As satellite technology advances, these quality factors will be vital for improving the accuracy and usefulness of satellite observations, enabling users to make better-informed decisions.