2016 Intermountain GIS Conference

Increasing adoption of precision agriculture (PA) technologies by farmers in Montana has led to an abundance of valuable spatial data that can be used to inform management decisions. Seeding, fertilizing, and weed control can be applied at variable rates (VR) across a field using modern farming equipment and data is collected automatically. By using VR technology a producer can increase profits by maximizing yield and reducing input costs. The most commonly varied input is nitrogen, often a farmer’s greatest expense. Nitrogen is most often varied based solely on previous years’ yield maps and the current year’s yield goals. Although useful, yield data is limited, as it often co-varies with other drivers that influence crop yield. Additional information such as soil fertility tests, cropping history, weed density, and weather can and should be used in the decision making process to further optimize input efficiency. This is difficult, however, as significant obstacles exist for farmers and consultants in collecting, organizing, synthesizing, and analyzing these data. For example, agricultural companies force consumers to rely on their non-interoperable file types and proprietary data management software programs that often lack desired features. In response, we have implemented an on-farm precision experiment (OFPE) to begin to address the lack of a free, non-proprietary data organization, analysis, and management recommendation software infrastructure. We have created a participatory research network (PRN) of interested farmers throughout the state to help us design and test our approach. Cooperators in the PRN will provide us with previous years’ yield data, fertility treatments, herbicide treatments, and soil tests, and will manipulate nitrogen levels in two fields on each farm for our experiment. These data, in addition to derived products such as digital elevation models, topographic wetness indices, and climate data will help us to design the database schema, populate the database, parameterize models, and test the eventual graphical user interface through which farmers will be able to access and display their data in the future. The greater OFPE project involves agricultural scientists, database design specialists, computer scientists and agricultural economists. This presentation will focus on the challenges associated with first-level data collection and transformation to non-proprietary formats, yield data QA/QC and cleaning, creation of derived products, the various tools used to accomplish each step, and obstacles to overcome to reach the eventual goal of automating data flow from farm equipment to the OFPE user interface.