Intro to remote sensing
Remote sensing data refers to any data that are collected remotely using instrumentation that can range from a simple camera to high-tech sensors mounted on satellites. The majority of remotely sensed data used by Regrow provides measures of energy reflected or emitted from the land surface that range from UV to thermal to microwave wavelengths, including the red, green and blue wavelengths we see everyday. Indeed, our eyes could be considered remote sensing instruments collecting data across the visible spectrum.
Where does Regrow’s remote sensing data come from?
The remote sensing data implemented by Regrow are primarily sourced from government space agencies, such as NASA and ESA, using a legacy of sensors from the Landsat and Sentinel satellite missions (specifically Landsat 5 through 9 and Sentinel 1 and 2). These data provide complete coverage of our entire planet dating back to 1985 with multiple images per month and usually more. Many other satellite, airborne, and near-surface remote sensing data are available, from public and commercial sources, and Regrow capitalizes on these resources when needed to monitor agricultural lands to the best extent possible and advance our science capabilities.
Why use remote sensing data to monitor on-field management practices?
Remote sensing provides a wealth of information that would otherwise be impossible to collect across equivalent space and time. These data also enable unique insights from measurements across the electromagnetic spectrum (e.g. UV, thermal, microwave wavelengths) that go far beyond the standard RGB satellite images we are accustomed to viewing.
Monitoring farms and ranches on site can be expensive and logistically difficult as it requires teams of skilled and knowledgeable individuals, often with expensive equipment. And of course we can’t be everywhere all at once, so we clearly can’t collect ground measurements across every agricultural field on earth; or even, for example, a set of 1000 fields that may be in an area of interest. This is why scientists traditionally collect a smaller set of ‘samples’ across strategic locations that best represent those fields. However, we can’t just collect a sample and assume the practices we measure on one farm or ranch are the same as a neighboring farm or ranch. At Regrow we need to monitor each and every field to ensure that each farmer gets the credit they deserve for implementing resilient practices. Remote sensing provides the solution as sensors (particularly satellite sensors) provide land surface measurements every 8-10 days across every acre of the earth’s surface.
Not only do these data allow comprehensive monitoring that is otherwise logistically infeasible, they provide insight across a range of indicators critical to monitoring crops, ranches, and resilient agriculture practices. Using these indicators we can monitor commodity crop and cover crop chlorophyll levels, grassland productivity, drought effects, field level crop residue, crop responses to climate or management, and irrigation and water use, to name a few. And this is all possible at high resolutions (approximately 4 to 36 measurements per acre); consistently and objectively through time. This isn't to say that ground data plays no part - indeed it is extremely valuable and plays a critical role in calibrating or training the remote sensing models and verifying and validating our results.
Limitations of remote sensing data
Cloud cover and occasional sensor errors can limit satellite data collection over fields; field data to test or train the algorithms and models may be limited in certain regions; or the sensors measuring the surface may have limited sensitivity to the practice being monitored. We overcome these limitations as best as possible by implementing a suite of sensors to maximize data retrieval and source ground data from every possible option (government agencies, NGOs, grower coops, farmers). We also continually test and implement the most advanced machine learning and AI models, capitalizing on the power of technology to achieve more accurate results and ensure each farmer's practices are recognized and rewarded.