Authors: Drew Baustian and JP Metcalf(NV5)

When it comes to selecting geospatial data for a project, there are numerous factors to consider. Each project has unique needs, and the type of data chosen can significantly impact the outcome. For this blog, I won’t cover every consideration, but I want to highlight the use of data from Capella Space for a specific application—open-pit mining.

For this particular project, Synthetic Aperture Radar (SAR) data, such as that produced by Capella, is a perfect fit. SAR data is uniquely suited to monitor changes in terrain and track movements with exceptional precision, even under challenging conditions.

Capella’s SAR data has good spatial resolution with a resolution of up to 25 cm, and consistent temporal coverage with revisit rates as frequent as every 2-4 hours. Capella’s high-resolution data can be used to detect even the smallest changes in topography and material movement, capturing intricate details that other sensors might miss. This distinction is critical when monitoring dynamic environments like mining sites, where understanding minute changes can make a big difference in operational planning and risk management.

For context, here are two SAR images of the same location at the West Angelas Iron Ore Mine in Western Australia. The one on the left was captured with Sentinel-1, the one on the right from Capella.

The level of detail from Capella data unlocks new use cases for SAR data.

In this blog post, I will walk through several methods for detecting and analyzing changes that have occurred over the course of about a week at the West Angelas mine site.

SAR data in its originally processed format can be challenging to interpret. Consider two Capella SAR datasets of a mine site, captured on July 28 and August 5, 2024.

Spotting the differences between the two images manually is difficult. However, by performing Coherence Change Detection (CCD) in ENVI SARscape, these datasets can be systematically compared, and a new map generated that displays coherence (sameness) from dark (little to no coherence) to bright (high/total coherence). The CCD output from these two datasets is shown below:

Dark areas on the CCD output indicate potential changes, such as excavation or vehicle movement, from the first data capture to the second. However, due to the SAR sensor’s off-nadir angle and the mine’s topography, abrupt elevation changes can create radar shadows that appear black and can be misleading. While CCD simplifies visual comparisons, it doesn’t fully explain the cause of changes or differentiate radar shadows from actual low-coherence areas. ENVI SARscape does have a shadow calculation, however for this site, we were unable to find a DEM with a high enough spatial resolution to produce a shadow map.

The ILU RGB Workflow in ENVI SARscape enhances coherence data by integrating amplitude values into a Red-Green-Blue composite image. When analyzing SAR data, amplitude is used to measure the strength of radar signals reflected from the Earth’s surface. This measure varies based on surface type with rocky terrain producing high backscatter due to its rough texture, while smoother surfaces like water produce weak backscatter. By incorporating amplitude into an RBG composite, the ILU workflow allows for a more comprehensive visualization of surface characteristics, improving the interpretation of changes over time.

The ENVI SARscape workflow assigns coherence (sameness) to red; amplitude average is assigned to green (no amplitude change), and amplitude variation to blue (a change in amplitude), creating the following output:

The brown areas (a mix of red and green), which dominate the image, show high coherence and little change in amplitude, indicating stability. Blue areas highlight changes in amplitude and coherence, suggesting recent activity or material movement. For example, blue lines on storage piles indicate new material deposits, while the brown-red hue on the piles shows no change. The green streaks on the north side of all the mine pits are likely due to sensor angle and pit topography, rather than actual surface changes.

In the image above, note the southwest corner of the dataset where a prominent blue hue reveals recent excavation work.

In the image above, the green lines on the north side of the mine’s pits likely do not represent actual ground characteristics.

The spatial resolution of new SAR satellite constellations, such as those of Capella, make it possible to extract insightful information about changes that have occurred. When this data is analyzed with ENVI SARscape, high levels of precision can be brought to monitoring storage sites, identifying excavation activity, evaluating topographical changes, and much more. This capability is critical for organizations that want to optimize operations, manage risk effectively, and make more informed decisions based on accurate and timely information.