Remote sensing facilitates the collection of high-frequency spatially disaggregated data in a consistent manner across entire countries and has become a crucial tool in a farmer’s arsenal. To boost productivity and profitability, farmers need to carefully manage where they invest their time and resources. Advancements in remote sensing technologies have made such precision agriculture practices possible at a scale never before seen. Improvements in computing power and machine learning algorithms have also made it increasingly feasible to automate data processing, thus facilitating the provision of spatially downscaled data to farmers to inform their resource allocation and other management decisions, including application of precision agriculture practices.
“This data was always in the farmer’s mind. But this is a tool that makes it easier for that information to become digital. It’s easier to remember, track, and compare to other data from the past,” said Nathan Faleide, strategic initiatives manager, AgIntegrated (AGI). “It’s really helping [growers] not only become more efficient but helping [them] see what they’re doing beyond just what they think in their head. They can actually see the data in front of them and know it’s not lying to them, data is data.”
Historically, the agriculture industry has had challenges when accessing relevant imagery data. Information needs differ considerably, in terms of temporal and spatial resolution relevant for decision-making, across different components of the agricultural supply chains and between large farms and small ones.
However, according to James Rineer, director of Geospatial Science & Technology, RTI International, barriers to accessing imagery are decreasing while imagery resolution and sensor variety continue to increase. “The resulting precision agriculture products will impact farmers’ ability to better plan when to plant, what to plant, how to best manage and maintain their crops, including allocation of inputs, and when to harvest,” he said.
“Technological advances in remote sensing, the reduction in cost for data access, and increasing numbers of capable in-county UAV operators capturing detail at the individual plot area are all factors likely to have a significant impact on the productivity and profitability for farmers,” said Robert Beach, senior economist and fellow, RTI International.
Many research institutes and industry organizations have begun to offer insight into the application of precision agricultural practices and cutting-edge technology to improve agricultural resilience.
Skywatch and AGI Offer Satellite Imagery for Precision Agriculture
On June 23, SkyWatch Space Applications and AGI announced that they will offer satellite imagery for precision agriculture. AGI will offer customers access to SkyWatch EarthCache, a software platform offering access to multiple sources of satellite data, through its Field Data Exchange, an application program interface (API) for multiple data sources.
For AGI, according to Mike Santostefeno, business development vice president, AGI, it made sense to work with SkyWatch “to help solve some of the many challenges that have plagued agriculture when it comes to getting imagery data,” he said in a press release.
SkyWatch is supplying the first imagery for AGI’s Field Data Exchange, which is home to other types of agriculture-related APIs. In the future, AGI plans to add aerial and drone imagery as well as data analytics to its platform.
One major challenge this partnership would solve, according to Faleide, is the access and distribution of data—getting it from satellite to the source (client). “Not too long ago, in order to do this, a third-party company had to buy a lot of data from X satellite imagery provider and then parse out the data for distribution. To connect directly to APIs from a satellite source wasn’t even possible five years ago,” he said.
SkyWatch is focused on bringing different satellite sources together into one easy-to-use API. SkyWatch will also refer to AGI any agriculture-based clients interested in their EarthCache system to better facilitate data service and support around its imagery distribution services.
“We’re trying to connect to all the different data sources and platforms through one API so that it’s easier for others who want to connect to all those different platforms,” Faleide said. “We aren’t adding on anything extra beyond the ability to connect through one API to help reduce development time and costs.”
The Rwanda Project: An RTI Grand Challenge
Traditionally, the best way to gather data on land use was to send surveyors into the field, measuring crops by hand. This slow, expensive process did little to accelerate the kind of comprehensive insight into crop health and potential yield that Rwanda’s leaders were seeking.
Most of Rwanda’s crop production is from smallholder farms. Agriculture officials, historically, had insufficient data on where crops are cultivated or how much yield to expect. The government began to look to technology for a solution.
RTI has funded research and analysis using satellite images and artificial intelligence to improve agricultural resilience and promote precision agriculture practices in Rwanda.
“Satellite data can bring a more spatially disaggregated picture of current land use than is typically available from other data sources. Rather than accessing data on land use allocation at a political boundary level, for instance, land use can be identified at 10m x 10m resolution using the freely available European Space Agency’s Sentinel-2 satellite data or even higher resolution from a variety of commercial satellites,” said Rineer. “This can bring a more detailed and current understanding of land use than is typically available from other sources.”
The Rwanda project focused on identification of specific crop types at the smallholder farm plot level, spatial analyses of the impacts of alternative practices on yields, and potential climate change impacts on crop yields.
“We also developed information on agricultural planting dates, which is of great interest to the government for assessing where planting may be delayed and the implications for the growing season,” said Beach.
The algorithms implemented in the project use a blend of satellite and UAV information to increase the accuracy in accounting of country-wide crop inventory, better informing policy decisions and timely interventions to optimize productivity, profitability, and resilience. The RTI Team is currently looking to refine the models to incorporate additional sensor types including synthetic aperture radar as well as other information sources.
“The important thing to note is that we and many others already have working systems, but these systems will continue to evolve over time as technology changes,” added Rineer.
Featured Image: Image shows modified Copernicus Sentinel data captured in 2017. | Credit: SkyWatch