NDVI vs NDRE: picking in-season nitrogen rates

Every grower asking for a nitrogen top-dress script eventually asks the same follow-up: NDVI or NDRE? Both show up in the same imagery providers, both get called a "vigor index," and both will hand you a green-to-red map that looks confident regardless of whether it's telling the truth about the crop. The right index depends on what growth stage you're flying and how much leaf area is already stacked between the sensor and the soil.

Where NDVI runs out of road

NDVI (Normalized Difference Vegetation Index) works off the red and near-infrared bands. Healthy, chlorophyll-rich canopy reflects a lot of near-infrared and absorbs red light, so the ratio climbs as biomass builds. That relationship is strong and well-documented, and it's why NDVI has been the default for two decades of vigor maps.

The problem is saturation. Once a corn canopy closes, usually somewhere around V8 to V10 depending on population and row spacing, NDVI stops climbing even though the crop keeps adding biomass and nitrogen demand keeps rising. The index plateaus near its ceiling, and a field that's genuinely running short on N can look nearly identical to a field that's fine. That's a real limitation for a top-dress decision made at V10 to V12, which is exactly when a lot of retailers are flying the pass that feeds the rate script.

Why NDRE holds up through canopy closure

NDRE swaps the red band for red-edge, a narrower slice of the spectrum between red and near-infrared where chlorophyll absorption is less extreme. The red-edge band doesn't saturate as fast, so NDRE keeps separating zones even after the canopy has closed over. It's a better fit for the growth window most top-dress and sidedress decisions happen in, once NDVI has already flattened out and quit telling you anything useful about where the crop is hungry.

NDRE isn't automatically "more accurate." It's a different piece of the spectrum, and the sensor needs to carry a red-edge band in the first place, which not every low-cost multispectral payload does. Early season, before canopy closure, NDVI and NDRE usually tell a similar story, and NDVI's wider dynamic range at low biomass can make it the easier index to read. Canopy closure sets the crossover point. A late-planted or thin-stand field can stay in NDVI's useful range well past when a thick, early-planted field next door has already saturated.

Picking the index for the top-dress pass

A practical rule that holds up in most corn programs: run NDVI for early-season stand and emergence checks, switch to NDRE once the canopy has closed and you're building the zone map that drives the variable-rate nitrogen script. If you're only pulling one index per season and trying to make it do both jobs, NDRE is the safer default because it stays useful later into the window when the actual rate decision gets made. The tradeoff is that NDRE needs a sensor built for it and a bit more interpretation for a grower who's used to reading NDVI maps from a previous provider.

Either index is only a gradient map until it's turned into zones and rates a controller can read. That's the step where a lot of retailers are still exporting a PDF, eyeballing the breaks, and hand-drawing zone boundaries in a separate GIS tool before the file ever reaches the cab. Field Variability builds that step into the weekly imagery pass itself, turning the biomass and vigor gradient straight into a per-zone seed, fertilizer, and chemical rate file the controller runs without a manual rezone every time new imagery comes in.

If you're scripting top-dress nitrogen off red-edge imagery this season and tired of redrawing zones by hand every pass, that's the problem Field Variability is built to take off your desk.