This is Part 2 of a talk given by our COO Paul Miller at the American Chemical Society conference in San Francisco. It goes into more detail about Morning Farm Report.
Morning Farm Report is the delivery platform for our data products. We offer many different products that growers can use to make data-driven decisions. The first is Rainfall, which is field specific, real time, and can be used to make simple planning decisions based on rainfall. Did it rain too much for in general for planting?
We also offer Field Intel, which is a data science-based product. We use crop models to aggregate weather data into growth stage related bins. Forecasts through historical data are available at an ag-field scale.
Trafficability is a sophisticated soil model that allows planning for advanced field operations for a few to thousands of fields.
Finally, we have a yield-modeling product that integrates sophisticated stress modeling with nutrient sub models to understand fertilizer applications – to make them more efficient. With this approach, model scenarios are used to predict influencing factors on corn and soybean yields.
All of these different tools can be used by growers to run their operations at all levels of sophistication more efficiently.
We deliver Rainfall data for thousands of fields on a daily basis to both growers and developers for their software applications. This map shows the spatial distribution of these fields across the United States. Our Rainfall product is sent by email at 5am for all of a Grower’s fields. This simple approach allows data-driven decision making right when it is needed which can save on fuel costs and equipment setup.
More sophisticated longer-term views of data are part of our Field Intel product. While Field Intel includes all of the date-based analyses a grower might expect to see, an agronomic-based data science approach is the main focus. Field Intel is used to help growers truly understand the reaction of their crops (past and present) to meteorological factors, which influence over 50% of variability in yields. These science-based approaches really help growers fine tune seed selection, nitrogen applications, and other agronomic decisions. The graphic above showcases the drought in Champaign-Urbana in 2012 compared to this year. June and especially July were incredibly low in Rainfall totals compared to 2014, whereas rainfall totals for 2014 were only slightly above normal in April and May and significantly greater for 2014 in June and July. This will most likely not affect grain yield this year, though it is too soon to know.
Morning Farm Report’s Field Fit Trafficability product represents over 5 years of development and scientific thinking. Our approach allows growers and ag businesses to schedule fields for work for the current day and up to 14 days in the future at 3 different weight classes of vehicles (more to come). Combined with additional routing algorithms and based on soil physics and hydraulics, Trafficability represents the probability that one can get into a field to perform timely operations. During a wet year where scheduling is packed, getting seeds in the ground without destroying the crop is critical. In addition, our approach allows efficiency to be increased for an operation–spending time loading up vehicles to travel only to get to field that cannot be worked is wasted time and money. Data driven decision-making saves on both fuel and time, and is critical to improving a grower’s business and food security at the same time. The only way to grow more is to produce as efficiently as possible. This is critical in times of lower grain prices.
Shown in the map above is a spatial view where fields fit to work are shown in green, marginal shown in orange, and fields where operations cannot be performed are shown in red. These simple map views allow growers to scan spatially to see which fields are fit to work. A preliminary graph for Trafficability (shown below) shows the classes of field operations that can be performed for a location in Champaign versus time. With the inclusion of our Forecast product, these graphics are generated up to 14 days in the future with the best results available in 2 and 5 days.
Finally, we offer a yield calculation product that allows a grower to integrate the weather, planting decisions, seed decisions, soils, and nutrients into a data-driven analysis approach. This package allows a grower to understand all of the operation decisions made during the season, puts the past into perspective, integrates the future with Forecast, and aids in development of different operation approaches (including fertilizer applications) to increase yields on all fields. An excerpt is shown in the table where yields are estimated using our crop modeling approach. This product includes our finest field-level Rainfall and meteorological data.
Finally, Yield produces an estimate of where a grower’s fields fit across the growing area for the selected crop. We model everywhere to give a grower a market advantage (see map below for 2007 growing season). Data-driven decision making can be utilized not only on the farm but also across the market.
Getting these kinds of tools into the hands of growers is critical to bridging the data science divide that may exist when looking at a grower’s own data. Agronomic efficiency and sustainability requires access to all time scales of data, from slow (historical) to fast (real-time). Agriculture is one of the most complicated human endeavors; these data driven tools allow growers and agricultural businesses to be as efficient and knowledgeable as possible to try to meet long term food security goals.