Satellite imagery is a complement to GIS data in the forest industry for geospatial mapping and analyses, but new uses will continue to emerge as satellite imagery becomes more available and affordable. As a result, companies should be familiar with additional uses beyond the traditional mapping functionality that most forest companies are accustomed to today.
Categories of Satellite Imagery
For many years, earth observation satellites such as Landsat 7 and Sentinel 2 have provided multi-spectral imagery for private or commercial use. Although free to download and widely available, these satellite images are medium resolution (10-30m per pixel) and are not ideal for accurate mapping purposes. Nonetheless, they do work well for vegetation analysis and to detect broad landform changes. Low resolution images may provide value to forest companies from a property-level standpoint.
On the other hand, high resolution satellite imagery (1-5m per pixel) is sufficient for mapping purposes and provides excellent detail at the stand-level, although they may have limited use for applications requiring a high degree of detail such as individual tree surveys.
Very high resolution imagery (20cm – 1m) provides the most accurate mapping and tree-level detail. These images open up possibilities for individual tree detection, measurements and sampling.
Forest Resource and Activity Monitoring
Satellite imagery is a good data source for common remote sensing tasks such as object detection and classification. Deep Learning solutions that rely on Neural Networks can accurately identify features and detect changes in vegetation, landforms or structures over time. It’s uncertain whether the very high resolution imagery will perform better (provide more statistical significance) in this regard, but medium and high resolution images appear to provide adequate results for most use cases.
Today, high resolution commercial satellite imagery is available for purchase directly from companies like Maxar (DigitalGlobe) and Planet. These companies offer subscription-based services as well as archive imagery products and requests for new capture. Both companies operate multiple satellite constellations with high accuracy sensors that provide daily coverage of the earth’s surface.
As commercial satellite imagery becomes more widely available, new services and industries are emerging that could have a profound impact on the forestry profession. In particular, companies like Swift Geospatial, Rezatec, and 20tree.ai are offering services that will monitor forest health and forest activities over a large-scale area. Both Rezatec and 20tree.ai also provide forest mensuration metrics using satellite imagery and Rezatec provides additional estimates of biomass and stored carbon. Forest health may be monitored with satellite imagery using plant reflectance indices such as NDVI to monitor for drought and insect damage even before the trees show signs of visible stress.
These services appear to use both medium- and high-resolution imagery, combined with Artificial Intelligence (AI) to provide a full suite of products and services. The primary advantage of satellite imagery is its global availability and timely capture which further allows daily or weekly analysis. Compared to other imagery sources like aerial photography or drone capture, satellite imagery offers a clear advantage in this regard.
Processing large amounts of satellite imagery requires additional computing power and storage requirements. As more companies recognize the benefit of satellite imagery, coupled with cloud storage and processing, it’s my belief that these services will become all the more commonplace, meaning that traditional methods of monitoring forest resources and activities will be altered for generations to come.
Categories: Satellite Imagery
The Forest Technology Blog 
Leave a Reply