At R3, canopy closure is the key signal
Satellite crop monitoring commonly uses NDVI to indicate biomass or chlorophyll, with useful results through vegetative growth. Soybean changes once it enters reproduction. At R3 (beginning pod), the canopy is either closed or not, and that distinction explains more of harvest outcome than any individual vegetative-stage reading we track.
In a well-managed soybean field, rows generally close between R1 and R3, shaped by seeding rate, row spacing, and seasonal water availability. At R3, 140,000 plants per acre in 30-inch rows should produce nearly complete cover between rows when viewed from above. Harva models compare reflectance at the row center with reflectance from the plant crown. A near-zero gap in visible and near-infrared bands marks closure. If the gap remains through the R2-R3 window, we flag the field for review.
What R3 looks like to the satellite
From orbit, a fully closed R3 soybean field appears as a continuous dark-green pattern of trifoliate leaves, with little soil reflectance between rows. Near-infrared reflectance is high and fairly even, while red-band absorption is strong. Under normal mid-summer conditions in the Midwest and Central Valley, a healthy closed canopy typically produces NDVI values from 0.75 to 0.88. Region, planting date, and variety can change the result.
Late or partial closure produces another signature. Soil between rows enters the reflectance mix, lowering NDVI and raising the bare-soil fraction. The result can resemble water stress with rolled leaves, even where leaf area has not actually declined. We therefore pair a bare-soil fraction index with NDVI to separate a thin canopy from a rolled, stressed canopy. Those conditions call for very different management responses.
How late closure relates to pod abortion risk
Soybean pod abortion is driven by stress and is concentrated in the R3-R5 window. Plants set more pods than they can fill, removing the surplus when resources become insufficient. Those resources include photosynthate from existing leaves and soil water that supports its production.
Incomplete closure at R3 means a field starts with structurally lower leaf area index, reducing light capture and daily photosynthate production. That shortfall may not produce visible pod abortion by itself, but it leaves the crop more exposed to added stress. If heat or dryness arrives at R3-R4, pod abort rate can rise faster in a field already behind on canopy than in a fully closed neighboring field.
Across early-access pilot fields in the San Joaquin Valley and parts of the Midwest corn belt, fields scoring as delayed at R3 later displayed more variable harvest zones in combine yield maps layered over our satellite data. We do not present those comparisons as a direct causal count, since season-to-season variation, hybrid differences, and agronomist decisions all affect closure timing. The correlation is nevertheless strong enough for us to flag delayed fields for a closer irrigation review during R3-R5.
Late closure: what it signals
A soybean field can miss full closure by R3 for several reasons, and their urgency differs.
On sandy or variable-texture ground, too few plants is the most common explanation. A stand below 90,000 plants per acre in wide rows leaves interrow gaps through reproductive growth. In that case, a satellite closure flag confirms the planter monitor rather than revealing something new. Few management choices remain at that stage.
Vegetative-stage water stress during V3-V6 can limit lateral branching and internode elongation. Plants then stay shorter and narrower, closing late or never. With adequate vegetative-stage rainfall, delayed closure points more toward population, soil, or variety. A June VWC drawdown in the satellite record instead makes water the likelier cause. Before flagging, we compare closure scores with soil moisture history in each zone.
Disease, especially Phytophthora root rot or sudden death syndrome in susceptible environments, can produce zone-level gaps that resemble low population from above. A field walk distinguishes them. Satellite data identifies where to walk, not what the walk will find.
Limits of satellite closure readings
The limits matter. Closure scores work best on clear-sky passes taken within three to five days of R3. Clouds and haze force interpolation from neighboring passes, shifting the analysis and potentially making a field appear less developed than it is.
Row spacing also changes the reading. Narrow-row soybean (15-inch) closes weeks before wide-row soybean (30-inch), so closure date matters less in isolation than against the date expected for that field configuration. Harva accounts for row spacing, but a recent spacing change that is missing from the grower's record can set the flag threshold incorrectly.
Satellite closure is an average across a field or zone. If half the area is closed and half remains open, the result is partial closure: the closed half may be fine while the open half is not. We map the signal at 10-meter resolution instead of reducing it to one field average. Even so, that scale misses individual stands and wet micro-topographic spots that can differ sharply inside one pixel.
Harva's R3-R5 watch window
For weekly prescriptions, R3-R5 is soybean's most water-sensitive period. Our flag threshold is tighter then than during vegetative growth. A VWC drawdown considered "watchable" at V5 becomes a "prescribe irrigation" call at R4 on that same field.
The latest satellite-pass closure score feeds into that threshold. Delayed closure produces a tighter VWC threshold, so irrigation is recommended at a higher soil water content than for a fully closed neighbor. A lower-LAI field has less canopy buffering through a hot afternoon, raising effective water demand per unit of yield potential.
Your agronomist should decide whether that recommendation merits action. We provide the combined evidence: satellite closure score, depth-profile VWC trend, and the 5-day temperature forecast, which together inform irrigation timing. Rate, duration, and whether to prioritize one field over another remain decisions for you and your agronomist.
Pairing canopy data with field notes
Canopy closure data is most useful when checked against planting records. As-planted population, seeding depth, row spacing, and the variety's usual canopy architecture establish the expected pattern. If satellite closure trails that expectation, the possible causes narrow quickly.
In a normal season, a soybean field planted at 130,000 seeds per acre in 30-inch rows with an aggressively branching variety should close by R2. Incomplete closure at R3 points to a specific problem: poor soil fit at planting that weakened early roots, a mid-June moisture event that limited lateral expansion, or an overstated stand count. Each possibility leads to a different management path.
Harva flags the closure anomaly; the field walk finds its cause. Together, these signals support data-directed scouting instead of guesswork or satellite reliance, helping mid-size farms find yield-limiting conditions before harvest maps confirm them.