By Taylor Jaseph, Sept. 6, 2022
A research project received $299,911 from the United States Department of Agriculture to develop models that will help farmers by predicting how climate change will affect California’s specialty crops of grapes, strawberries and citrus.
This research will translate data collected over three years from unmanned aerial vehicles and satellite imageries into a map. Farmers will be able to access this map to aid in their decision-making of their farms according to how the climate changes.
“Instead of them (the farmers) just hearing ‘oh, the weather in California will change by that much,’ that doesn’t translate on how they manage their field,” said Gabriel Granco, assistant professor of geography and lead researcher. “But if they go to the internet and access a map and zoom in to their farm and see this model is predicting whatever will happen here — here being the farm — they can make a decision.”
The grant will fund the research for three years, where the focus will be on data collection. According to John Korah, assistant professor of computer science and researcher, this data will be used to develop initial models and see how machine learning techniques will help with predictions. Throughout the research, the models will mature as more data is collected.
This research is fed by multiple different disciplines, with each playing a necessary role in the project. Granco will work with satellite data to understand where production can take place. Korah is providing the computational infrastructure and helping connect all the data. Aerospace engineering professor, Subodh Bhandari is operating the UAVs to gather data. And plant science professor, Priti Saxena, is providing the necessary information on how the plants are faring.
By using existing bioclimatic variables, an ideal situation for the plants is being created — such as providing the proper nutrition, proper soil environment and irrigation techniques to have the plants maintained at their optimum health. Through the gathered data to understand the future bioclimatic variables, the models will help farmers find this ideal situation throughout climate change.
“In that scenario we can easily tell, OK the plants are doing great,” Saxena said. “So, any situation lesser than this, that means the plant is under stress. And that’s when we’ll see how stressed they are and what exactly the environment factor is affecting the plant health.”
To gather the data, Bhandari flies the UAV when the satellite passes over Cal Poly Pomona. This way, the data both systems collect are comparable. The UAV provides more precise data than the satellite, so the UAV can really define the conditions needed for the crops to grow. The satellite provides a wide coverage of the area. According to Korah, combining these two types of information isn’t really done at the present, so that is where some of the novelty of this project comes.
The model will be developed and applied under current climate and the climate conditions of 2021-2040, 2041-2060, 2061-2080 and 2081-2100. These periods were developed by the most recent scientific standard from climate models for the research. And from these periods, the team can gather information about how possible ways the bioclimatic variables will change over time.
“One of the points of climate change is that people think that that’s just in the future. Far, far future,” Granco said. “But we can see right now that even in 2040, the difference in climate will be meaningful. It can have a meaningful impact. So that’s what we want to dig in and find what will be the possible impact for these crops we are studying.”
By finding these possible impacts for California’s major crops of grapes, strawberries and citrus, the team can then find where the best location will be for the crops. So, the location of the crops right now is Spadra Farms at Cal Poly Pomona for the research and the current climate conditions are being modeled. Then, through analyzing the data, the team is trying to see where the same climate conditions will happen in the future where the crops can grow healthily, according to Granco.
The modeling will also help farmers on figuring out decisions such as the best irrigation practices, how they should maintain the soil and how the plants are responding to the given environment at a given time. This is the future Granco and the team envisions.
“It’s like a marriage of the three — you have the plants and then you have the technology, machine learning, and then you have environment that’s changing. And that’s what we’re trying to learn,” Saxena said. “So, having any data will be of use for us. This has been working; this model is perfect; OK, we need to tweak something in this model to make it more applicable. So, this is also like learn by doing, at the same time we know we’ll get something good from the project.”
Feature image courtesy of Gabriel Granco.
