Satellite and drone imagery can show growers precisely where unhealthy plants stand in a field, but it cannot explain why those plants are stressed — and it cannot generate fertilizer rates. That is the core message of a new Michigan State University Extension guide on interpreting remotely sensed imagery in precision agriculture.

MSU Extension notes that interpreting imagery is rarely straightforward because weather, soils and management practices all shape crop growth. Used correctly, however, imagery reveals spatial patterns in crop health and productivity, showing where and how field conditions change across a season.

How does remote sensing measure crop health?

Current sensors measure sunlight reflected by plants across visible wavelengths — red, green and blue (RGB) — and near-infrared (NIR) wavelengths the eye cannot see. Chlorophyll in leaves strongly absorbs blue and red light, while healthy leaf structure strongly reflects NIR. Comparing reflectance across these bands produces vegetation indices (VIs) that highlight differences in growth, vigor and stress across a field.

Different bands carry different agronomic signals:

• Blue (~450 nm): tracks chlorophyll absorption.
• Green (~560 nm): indicates leaf vigor, canopy color and senescence.
• Red (~650 nm): reflects photosynthetic activity and vegetation cover.
• Red-edge (~730 nm): signals chlorophyll absorption and nitrogen status.
• Near-infrared (~840 nm): captures canopy density, plant health and vigor.

Because plants prefer blue and red light to build chlorophyll, they reflect more green — which is why vegetation looks green to begin with. The contrast sharpens in the NIR range: healthy plants reflect far more NIR than stressed ones, which is why indices built on NIR, such as the Normalized Difference Vegetation Index (NDVI), dominate precision agriculture.

What do NDVI numbers actually mean?

The MSU guide translates index values into on-the-ground conditions. NDVI values at 0 or below represent dead material or artificial structures such as roads and buildings. Unhealthy plants — those showing yellow, brown, orange or red hues — fall between 0 and 0.33. Moderately healthy, light-green plants score between 0.33 and 0.66. Very healthy plants reach values between 0.66 and 1.

Growers do not necessarily need paid imagery to start. Climate's FieldView platform offers Field Health imagery of enrolled fields for free, using satellites to generate vegetation maps based on the Climate Crop Index (CCI), a proprietary index similar to NDVI that estimates relative biomass accumulation.

Beyond NDVI, the guide lists other common indices. NDRE measures relative chlorophyll and indicates nitrogen status. GNDVI works better in late growth stages on developed canopies for biomass accumulation. EVI handles dense canopies without saturating.

Can imagery replace yield data?

MSU Extension frames remote sensing as one tool among several, working alongside soil maps and historical yields. Combine yield monitors reveal persistent spatial patterns that growers can use to delineate management zones based on long-term productivity. But where high-quality yield data is missing, recent research published in Precision Agriculture evaluated whether in-season satellite imagery could substitute.

The researchers compared satellite imagery collected throughout the growing season with yield patterns measured at harvest. A key finding: multiple images across the season provided more useful information than a single image, because crop conditions shift as plants respond to weather and management differences.

The guide is explicit about imagery's three hard limits:

• Imagery shows where good and bad plants are, quantifying spatial differences and tracking the same field areas repeatedly through the season. Plants act as sensors of their current growing conditions.
• Imagery cannot say why plants are unhealthy. Nutrient limitations, water stress, pests and disease all produce similar signals, so growers need additional data to isolate causes.
• Imagery does not provide fertilizer or amendment rates. Vegetation indices assess condition, but rate decisions require soil sampling or crop modeling based on expected yield response.

What this means for growers

MSU Extension positions remote sensing as a strong tool for identifying and monitoring within-field spatial differences during the growing season, with the caveat that imagery alone cannot diagnose stress causes. Combining it with soil and yield data supports better-informed management decisions, and MSU Extension says it can help farmers organize, analyze and interpret precision agriculture data to turn imagery observations into actionable decisions.