How Are Geospatial-Temporal Insights Playing a Role in Better Crop Yields

 

These days, big data is affecting everything from the way companies deliver advertisements to crop yield prediction. IBM Research recently introduced one such offering called IBM PAIRS, or Physical Analytics Integrated Data Repository & Services), which is an analytics service in the cloud that can provide faster and more accurate insights regarding crops. These insights can then be used to boost crop yields significantly.

What Is Geospatial-Temporal Data?

The term “geospatial-temporal data” may sound complicated, but it is actually quite simple. The word geospatial indicates that data is influenced by a geographical location or component, and the word temporal indicates that data is influenced by a specific state in time. By putting the two together, IBM has come up with a cloud analytics service that makes the best of both worlds. Farmers can analyze information about crops in specific places at specific times, which can provide them with more insight about yields.

Overcoming the Problems Posed by Geospatial-Temporal Data

While it is simple to understand the meaning of geospatial-temporal data, the act of collecting it is quite difficult. In fact, in order to facilitate accurate data collection and analysis, PAIRS requires information from a variety of sources. These sources include:

  • Global models. Things like the ocean, climate, and even weather pattern models on a global scale must be analyzed to provide the best and most accurate data.
  • Images play another important role in the cloud-based PAIRS service. These range from aerial imagery captured by drones to satellite imagery.
  • Sensor networks. PAIRS also utilizes a variety of internet-enabled georeferenced sensors that measure everything from the soil’s moisture content to the plant’s growth.
  • Big-event data. This is a collective term used to describe the host of information available via sources like Facebook, Twitter, and the Global Database of Events. Though these sources have limitless data that could improve science’s understanding of crop growth and yield, filtering that data has proven nearly impossible in the past.

The Development of Pairs: IBM & E&J Gallo

The PAIRS project arose after IBM paired up with E&J Gallo Winery to help grape growers discover the best possible way to increase the yield of their crops while conserving as much water as possible. Engineers from the two companies paired up and created a very precise and innovative irrigation system. This system utilized sensors, actuators, satellite images, local weather models, and water loss estimation models all shared over a cloud-based network, and it was used for two growing seasons on a 10-acre grape ranch.

The crops on these 10 acres thrived. Researchers documented a 50% increase in crop uniformity and 26% more yield, which surprised growers and engineers alike. The most important part of these results was the fact that the crops needed 22% less water than before.

These days, big data is everywhere. The internet puts information at our fingertips, and with the right tools, including IBM pairs, it is possible to filter out the necessary data, analyze it, find relationships between important factors, and then tweak the farming process in such a way that crops produce more (and more uniform) yields with significantly less water.