PRECISION AGRONOMY: WHAT YOU WANT VS WHAT YOU NEED

Some would suggest you want an EM survey.

Some would suggest you want an NDVI map.

But what you want is a map that reflects the problem you are trying to solve. Semantics? Maybe, but I’ll try and elaborate.

An EM survey, or the process of collecting soil electrical conductivity (EC) data, gives you information related to soil salts, texture, and moisture variability. Sometimes it really nails what drives crop variability and response to soil inputs, but sometimes it doesn’t. And therein lies the problem. What you really want as a farmer or precision agronomist is a map that represents the response probability to soil applied inputs such as fertilizers, soil amendments, and seed. That could include the EC data. But it could also require topography, elevation, soil color, vegetative indices, and/or gamma layers to be incorporated with the EC data.

An NDVI map reflects the relative “greenness” of a crop as seen from satellite or drone. It’s affected by both biomass and the color of that biomass. If your goal is a map that represents the relative rate of a desiccant needed across a field, that is probably what you want since it doesn’t matter why it’s green or not green, or even what is green. But if you want to understand crop emergence and early season vigor, it often falls apart due to noise from residue color, soil color, and weeds. If you want to understand where to topdress additional nitrogen in season, NDVI is a great supporting tool. But it doesn’t distinguish too wet from too dry, or sandy areas from sodic areas. And these are important factors that determine whether the crop will respond to additional nitrogen or not. You want a nitrogen response map, not an NDVI map.

Let’s take last year’s yield map and make a variable rate phosphate map! You could, but you’re presumably making the assumption that the soils’ ability to supply phosphate is the same everywhere (it’s not). What seemed logical actually may have just led you to put high rates in the least responsive soils, and the lowest rates in the most responsive soils. Oops.

Too often, I see farmers and agronomists ask for data that doesn’t give them the information they need. It might be a good supporting tool, or part of the final solution, but as a stand-alone data layer can lead to disappointment or worse – mistakes. That high NDVI area you wanted to apply the highest rate of fungicide too? It’s actually weeds. That high EC soil that you assumed was saline? It’s actually just a heavy clay with a full moisture profile and high productivity, despite having the same EC as the salt affected loam on the other side of the field.

It’s taken me a few years to train myself to first ask, “what am I trying to solve?”. And then work backwards to ensure I have a process to collect data that will get me the solution. Because I cover a diverse area with many different soil types and landscapes, I need a repeatable but adaptable process. One that doesn’t assume anything. I can’t spend 2 days mapping a farm to just collect EC data and find out later that’s not going to give me what I need. That’s not a solution, that’s an exercise of frustration! And unfortunately, is something many agronomists and farms have experienced over the years.

If you’ve experiences similar frustration, or are just embarking on your precision ag journey, feel free to contact us at Croptimistic Technology. We’ve been there, seen a lot of things, and made our own mistakes. We can help you avoid the same mistakes too.