Keeping pace with the rapid advancements of remote sensing data

We can use the properties of light to extract a great deal of information from any material.

All light can exhibit similar behaviors when interacting with the material. Still, the wavelength of light and the composition, thickness, shape, and size of the material depends on how and if this light reflects, absorbs, transmits, scatters, refracts, diffracts, or polarizes.

Spectroscopy is the study of how light interacts with materials at different locations of the electromagnetic spectrum. Years of thorough research have built a sizable library of various materials used as references for image analysis in remote sensing.

Camera sensors are designed and used as imaging spectrometers — by taking multiple images of the same scene. Each image collects light at a different wavelength of the EM spectrum. In that case, a single pixel at the same location within each of those images is its point spectrometer. The recorded spectra correspond to the spectral behavior of that area within the scene. Depending on the sensor and platform — a single pixel could represent a large area. The area might range from the size of a city block down to the size of a fingernail.

Remote sensing has taken advantage of imaging spectroscopy to provide a dizzying array of applications for Earth Observation. The tradeoff between design, cost, and purpose when designing a sensor for use on a satellite means several spatial and spectral resolutions are available, each sacrificing spectral resolution, bandwidth, and spatial resolution to keep costs reasonable, producing usable data for a specific application.

This article aims to unpack some of the primary use cases for remote sensing data and the applicable regions in the EM spectrum with available satellite data, airborne, and UAV data.

Visible to Near-Infrared

Visible and near-infrared (VNIR) are being grouped here since imaging methods often use both ranges to detect certain features. Another reason for grouping the two is related to camera sensor technology. The same silicon sensors that capture images on your smartphone are sensitive to wavelengths beyond the visible range and detect infrared but coated in an IR-blocking filter. We can also remove the physical filter on digital cameras to capture full VNIR images.

Fields such as biology, ecology, forestry, and agriculture witness the usage of VNIR imaging, but there are also some practical geological applications:

Leaf Area IndexCrop vigor/canopy healthEarly identification of pest damageLichen mappingIron oxidesOutcrop identificationRare-earth element enrichment (Neodymium)Sediment transportGeneral object detection

• Leaf Area Index

• Crop vigor/canopy health

• Early identification of pest damage

• Lichen mapping

• Iron oxides

• Outcrop identification

• Rare-earth element enrichment (Neodymium)

• Sediment transport

• General object detection

Shortwave Infrared (SWIR)

The SWIR wavelengths are more extended than VNIR and require more sophisticated sensor technology and internal cooling mechanisms to reduce noise. SWIR bands are particularly useful for detecting clays and clay minerals, which can, in turn, be valuable indicators of precious metal deposits, risk sites for landslides, and materials for tailings construction. SWIR also has some cloud and smoke penetrating abilities and can monitor active burning and burn scars during forest fires.

Clay mineralsSulfatesCarbonatesIndustrial methane emissionsForest fire/burn scar monitoringSoils

• Clay minerals

• Sulfates

• Carbonates

• Industrial methane emissions

• Forest fire/burn scar monitoring

• Soils

Thermal Infrared

Thermal infrared can accurately measure surface temperatures, temperature variation over time, and if the resolution is sufficient, heat loss from buildings.

Some minerals, particularly silicates, only exhibit characteristic absorption bands in the thermal infrared region. Depending on the platform and resolution, it is possible to identify some carbonate minerals by combining SWIR and thermal imagery.

Thermal imagery has a drawback as well — the spatial resolution is very coarse from satellite-based platforms on 100m per pixel. Such spatial resolution makes thermal imaging useful for regional surveys unless they use airborne or UAV platforms for collection.

Suomi NPP VIIRS Image

Radar/Microwave

Radar and microwave remote sensing are synonymous, and these data are quite different from the previous EM data. As wavelengths get longer and longer, the wave energy is reduced to the point that radar waves are more sensitive to surface materials’ textural characteristics than the spectral features.

Two additional aspects of radar data separate it from other sources — wave polarization and wave phase. Many radar platforms can send vertical waves or horizontal waves and record the direction (polarization) of the return signal, allowing for additional inference of surface materials, either natural or anthropogenic. The wave phase records and measures displacement between two images taken at the same location at different times.

One of the most significant draws for using radar is that it penetrates all cloud cover. No matter the location or cloudiness, the image will always be precise.

Here are a few examples of radar applications:

Flood monitoringDeforestation monitoringGround subsidence trackingSoil moisture levelsForest biomassCanopy densitySurface texture measurementsWetland detection

• Flood monitoring

• Deforestation monitoring

• Ground subsidence tracking

• Soil moisture levels

• Forest biomass

• Canopy density

• Surface texture measurements

• Wetland detection

LiDAR

LiDAR uses an entirely different collection method and is capable of producing incredibly detailed reconstructions of an area. LiDAR sensors send tens of millions of laser pulses in all directions at a single wavelength that record the distance, location, and reflected intensity of the material they interact with, making them ideal survey tools. The most common applications are:

Digital elevation modelsDigital surface modelsVegetation canopy structureBathymetric surveysUrban monitoring3D reconstructionsGeomorphology and landslide risk management

• Digital elevation models

• Digital surface models

• Vegetation canopy structure

• Bathymetric surveys

• Urban monitoring

• 3D reconstructions

• Geomorphology and landslide risk management

Conclusion

The range of use cases for remote sensing data is rapidly expanding. The use cases listed here are only a small, generalized portion of how to use Earth Observation data. When choosing the best option for remote sensing data, we must find answers to the following questions:

What is the acceptable pixel size required to distinguish my features of interest?

What spectral range identifies those features?

What is my budget? For every project, we must find a balance between the three factors mentioned earlier.

For more information, reach out to our team of professionals.