How Are Satellite Images Used in Agriculture?

Satellite images are used in agriculture to map fields, monitor crop development, reveal unusual vegetation or moisture patterns, estimate crop water use, classify planted areas, document weather damage, and support regional production forecasts. Their greatest value is not diagnosing a problem from space, but showing where and when conditions differ so people can investigate more efficiently.

Key Takeaways

• Satellite imagery helps locate and track differences across fields and growing seasons.
• Optical, radar, and thermal sensors provide different—not interchangeable—information.
• NDVI and similar products are calculated indicators, not direct crop diagnoses.
• Resolution, acquisition timing, cloud cover, crop stage, and field size determine whether imagery is useful.
• Important agricultural decisions should combine satellite evidence with field observations and other records.

This guide explains what agricultural satellites actually observe, which common products are calculated or modeled, and how those products can support practical decisions. It also includes an original decision framework, a field-pixel calculation, a tool-selection guide, and two documented examples of operational agricultural remote sensing.

Agricultural satellite imagery is best understood as a screening and monitoring tool. It shows where and when surface conditions differ, but usually cannot explain the cause without additional evidence.

---

How Satellite Images Support Agricultural Decisions

Agricultural remote sensing is the collection of information about crops, soil, water, and land without making direct physical contact with them.

Satellite sensors record reflected light, emitted thermal energy, or returned microwave signals. Those observations can then be calibrated, combined, classified, or entered into models to create agricultural information products.

In practice, satellite imagery usually helps answer one of three questions:

1. Where should someone inspect?
2. Which areas may require different management?
3. How have conditions changed over time?

NASA identifies surface reflectance, vegetation greenness, land temperature, soil moisture, rainfall, crop extent, and drought conditions among the Earth-observation variables used in agriculture. Some are sensor-derived observations; others are calculated or modeled from several inputs.

See NASA Earthdata’s agriculture production resources.

Common Agricultural Applications

Application	What satellite data can contribute	Decision supported
Crop development monitoring	Changes in canopy reflectance over time	Where growth differs from the expected pattern
Irrigation analysis	Vegetation, temperature, and estimated water-use patterns	Which zones deserve inspection
Crop classification	Seasonal spectral and structural signatures	What crops are likely planted and where
Field mapping	Boundaries, cultivated area, water, soil, and land-cover change	Inventory, planning, and reporting
Storm or flood assessment	Before-and-after surface changes	Where damage inspections should begin
Drought monitoring	Vegetation, rainfall, temperature, and moisture anomalies	Which regions show worsening conditions
Sampling-zone design	Persistent within-field differences	Where to collect soil or plant samples
Production forecasting	Crop area and seasonal development combined with other data	Regional supply estimates

Yield and production forecasts are estimates, not guarantees. They may combine satellite data with planted-area records, weather, statistical methods, and crop models. Extreme weather, unusual management, incomplete reference data, and changing field conditions can increase uncertainty.

---

What Satellites Observe—and What Software Infers

A map shown in an agricultural platform is not necessarily a direct measurement.

Satellite instruments record electromagnetic energy. Many familiar agricultural products are created later through equations, retrieval algorithms, classifications, or models.

Measured and Derived Agricultural Information

Product	Information type	Practical meaning
Red reflectance	Sensor-derived observation	Processed estimate of reflected energy in the red band
Near-infrared reflectance	Sensor-derived observation	Spectral response influenced by vegetation structure and other surfaces
Radar backscatter	Sensor-derived microwave response	Signal influenced by moisture, roughness, structure, and viewing geometry
Land-surface temperature	Retrieved geophysical product	Temperature derived from thermal observations and processing assumptions
NDVI	Calculated spectral index	Ratio derived from red and near-infrared reflectance
Crop type	Classification result	Algorithmic label based on imagery and reference data
Evapotranspiration	Modeled estimate	Calculation combining satellite and meteorological inputs
Yield forecast	Statistical or crop-model estimate	Prediction with assumptions and uncertainty
Crop health score	Platform-specific interpretation	Provider-defined score that may differ between services
Soil moisture	Product-dependent information	May be sensor-derived, modeled, assimilated, or downscaled

A product can be useful without being a direct measurement. The important questions are:

• How was it produced?
• What resolution does it have?
• How current is it?
• What uncertainty or quality information is available?
• Was it designed for the decision being made?

Evapotranspiration is a clear example. NASA explains that evapotranspiration cannot be measured directly by a satellite instrument. It is estimated using variables such as land-surface temperature, air temperature, solar radiation, vegetation information, and meteorological data.

See NASA Earthdata’s evapotranspiration overview.

---

Which Satellite Sensors Are Used in Agriculture?

Agricultural remote sensing commonly uses optical multispectral imagery, synthetic aperture radar, thermal infrared observations, and broader-scale microwave products.

Each data type responds to different physical properties. A sensor that is useful for tracking canopy development may not be the best choice for flood mapping, regional soil-moisture monitoring, or field-scale water-use estimates.

Optical and Multispectral Imagery

Optical satellites measure sunlight reflected from Earth in visible and invisible wavelength bands.

Healthy green vegetation generally absorbs much of the visible red light used in photosynthesis. Internal leaf structure reflects more near-infrared energy, creating a spectral contrast that can be used to study vegetation.

Red-edge and shortwave-infrared bands can add information related to canopy development, leaf water content, soil, crop residue, and surface moisture patterns.

The Sentinel-2 Multispectral Instrument records 13 spectral bands:

• Four bands at 10 m resolution
• Six bands at 20 m resolution
• Three atmospheric bands at 60 m resolution

Not every Sentinel-2 band is available at 10 m. The effective detail of an agricultural product therefore depends on the bands used and whether any bands were resampled during processing.

ESA lists agriculture, land-cover monitoring, leaf-area analysis, chlorophyll-related variables, and leaf-water-content mapping among Sentinel-2 applications.

See ESA’s official Sentinel-2 facts and figures.

Landsat 8 carries the Operational Land Imager and the Thermal Infrared Sensor. Its standard optical bands have 30 m spatial resolution, while the panchromatic band has 15 m resolution. The thermal bands have a coarser native resolution and may be resampled in distributed products.

Each Landsat 8 or Landsat 9 satellite passes over the same location on a 16-day cycle. Their orbits are offset, creating a combined opportunity approximately every eight days.

That schedule does not guarantee a clear, usable agricultural observation every eight days. Clouds, haze, acquisition planning, shadows, and processing delays may reduce actual availability.

See:

• USGS Landsat 8 mission specifications
• USGS Landsat acquisition schedules

Optical imagery is especially useful for:

• Comparing crop development among fields
• Tracking planting, growth, senescence, and harvest
• Calculating vegetation indices
• Mapping soil, water, residue, and vegetation
• Supporting crop classification
• Identifying locations that deserve scouting

Its main limitation is obstruction by clouds, haze, smoke, and shadows.

Synthetic Aperture Radar

Synthetic aperture radar, or SAR, transmits microwave energy toward Earth and measures the returned signal.

Radar does not require sunlight and is generally much less affected by ordinary cloud cover than optical imagery. Its returned signal, called backscatter, can change with:

• Surface roughness
• Soil moisture
• Crop structure
• Vegetation water content
• Polarization
• Incidence angle
• Recent field operations

Agricultural radar applications include:

• Flood and standing-water mapping
• Monitoring during cloudy periods
• Crop-structure analysis
• Crop classification
• Tillage and harvest-event detection
• Tracking larger moisture-related changes

ESA has demonstrated how Sentinel-1 radar time series and interferometric coherence can contribute to crop-type classification.

See ESA’s Sentinel-1 crop-mapping example.

Radar should not be described as unaffected by weather or as a direct measurement of crop moisture. Strong precipitation, soil conditions, surface roughness, crop structure, and acquisition geometry can all influence the signal.

Thermal Infrared Observations

Thermal sensors detect emitted energy rather than reflected sunlight. Processing converts that signal into a land-surface-temperature product.

Plants can cool their leaves through transpiration. When water becomes limited, stomata may close and canopy temperature can rise. However, temperature is also affected by:

• Air temperature
• Wind
• Humidity
• Soil exposure
• Canopy density
• Irrigation timing
• Acquisition time
• Cloud conditions

Thermal information can contribute to:

• Water-stress screening
• Evapotranspiration modeling
• Irrigation-zone comparison
• Drought analysis
• Surface-temperature mapping

A thermal satellite product should not automatically be described as the exact temperature of a crop canopy. More accurate terms include:

• Retrieved land-surface temperature
• Thermal observation
• Processed temperature product
• Satellite-derived surface-temperature estimate

Thermal sensors do not depend on reflected sunlight in the same way as optical sensors, but clouds can still prevent useful surface observations. Daytime and nighttime temperature products also should not be compared without accounting for acquisition conditions.

Passive Microwave and Soil-Moisture Products

Passive microwave instruments detect naturally emitted microwave energy associated with soil moisture and other surface properties.

NASA’s Soil Moisture Active Passive mission, or SMAP, provides global observations of surface soil moisture and freeze-thaw conditions. Other products combine remote sensing with land-surface models to estimate surface or root-zone moisture.

See:

• NASA Earthdata’s soil-moisture overview
• NASA’s SMAP mission information

These products can support:

• Regional drought monitoring
• Weather and hydrological modeling
• Watershed analysis
• Soil-moisture anomaly tracking
• Large-scale agricultural condition reporting

Many global soil-moisture products have pixels much larger than a single field. They can provide regional context without being suitable for deciding how much water one small field or irrigation zone needs.

Sensor Comparison

Sensor type	Strongest agricultural role	Main limitation
Optical multispectral	Vegetation, crop development, and land cover	Clouds and shadows
Synthetic aperture radar	Floods, structure, and cloudy-season monitoring	Complex, context-dependent interpretation
Thermal infrared	Surface temperature and modeled crop water use	Temperature differences have several possible causes
Passive microwave	Regional soil moisture and drought context	Often too coarse for field-level decisions

---

Using Vegetation Indices Responsibly

Vegetation indices combine spectral bands mathematically to emphasize selected surface characteristics.

They can make vegetation patterns easier to compare, but no index removes every influence from soil background, atmosphere, illumination, canopy structure, or sensor geometry.

What NDVI Shows

The Normalized Difference Vegetation Index compares near-infrared and red reflectance:

NDVI = (NIR − Red) ÷ (NIR + Red)

In words, NDVI subtracts red reflectance from near-infrared reflectance and divides the result by their sum.

See NASA Earthdata’s NDVI explanation.

NDVI can help compare relative vegetation greenness and canopy development. It is not a universal crop-health score.

A value that is normal for one crop, field, soil type, date, or growth stage may be unusual in another setting.

NDVI Calculation Example

Assume one crop pixel has:

• Near-infrared reflectance: 0.50
• Red reflectance: 0.10

The calculation is:

NDVI = (0.50 − 0.10) ÷ (0.50 + 0.10)
NDVI = 0.40 ÷ 0.60
NDVI = 0.67

An NDVI of 0.67 may correspond to a dense green canopy under many conditions. It does not prove that the crop is disease-free, adequately irrigated, or likely to produce a specific yield.

The result becomes more useful when compared with:

• Other zones in the same field
• Previous dates from the same season
• Similar crops at similar growth stages
• Historical imagery from comparable seasons
• Irrigation and weather records
• Field observations

The spatial pattern and trend usually matter more than one isolated value.

Choosing an Indicator

Indicator	Most useful for	Main caution
NDVI	General vegetation and canopy development	Can become less sensitive in dense vegetation
NDRE	Chlorophyll-related differences later in canopy development	Requires red-edge data
SAVI	Sparse vegetation where soil background is influential	Includes an adjustment factor
NDMI	Relative canopy moisture patterns	Does not directly measure root-zone soil water
Surface temperature	Thermal comparison and water-stress screening	Strongly affected by weather and exposed soil
Radar backscatter	Structural and moisture-related surface changes	Sensitive to geometry, roughness, and crop structure

Choose an indicator because it has a defensible relationship with the management question—not because it creates the most dramatic map.

---

Agricultural Satellite Monitoring Workflow

A useful workflow begins with a decision, not with downloading an image.

1. Define the Decision

State what action the information could change.

Useful questions include:

• Which irrigation zones should be inspected?
• Where should soil or plant samples be collected?
• Did flooding affect the whole field or only lower areas?
• Is one field developing differently from comparable fields?
• Which areas appear to have been harvested?

“Monitor crop health” is too broad unless the response to an unusual pattern is also defined.

2. Check the Field Boundary

The mapped field should exclude roads, drainage channels, buildings, tree lines, neighboring crops, and unrelated bare soil where possible.

Boundary errors can create mixed pixels and misleading edge patterns.

3. Match the Sensor to the Question

Use:

• Optical imagery for canopy development and spectral differences
• Radar for flooding, structure, and cloudy periods
• Thermal and evapotranspiration products for water-use analysis
• Regional soil-moisture products for drought context
• Drones or direct inspection for very small features

4. Inspect Image Quality

Before interpreting a map, check:

• Acquisition date
• Cloud and shadow masks
• Missing data
• Processing level
• Field coverage
• Recent rain or irrigation
• Planting, spraying, tillage, or harvest records
• Whether the crop stage makes the comparison meaningful

5. Compare Several Dates

One image shows a moment. A time series shows whether a pattern is persistent, developing, or temporary.

Persistent zones may relate to soil, drainage, elevation, salinity, or infrastructure. A new anomaly may reflect weather, equipment failure, crop damage, field operations, or a processing error.

6. Inspect Normal and Unusual Areas

Visit at least one unusual location and one apparently normal comparison area.

Record crop stage, visible symptoms, soil condition, irrigation status, photographs, weeds, and recent operations. This reduces confirmation bias.

7. Take a Proportionate Action

A satellite anomaly may justify scouting, sampling, checking irrigation pressure, inspecting drainage, or reviewing application records.

It does not automatically justify applying more water, fertilizer, or crop-protection products.

8. Review the Outcome

After an inspection or intervention, compare later imagery and field observations.

A monitoring system is valuable only when it improves the timing, location, or confidence of a real decision.

CosmoBasics editorial observation: The operational value of an agricultural map depends less on how sophisticated the algorithm appears and more on whether the user has a defined response when an alert occurs.

---

The CosmoBasics SCALE Framework

The CosmoBasics SCALE Framework for Agricultural Satellite Decisions is an editorial decision tool created for this guide.

It is not a statistically validated agronomic model, diagnostic system, or substitute for a qualified agricultural adviser. It is intended to help readers judge whether satellite imagery is a reasonable first tool for a specific problem.

Use SCALE before purchasing a platform, downloading imagery, or acting on an agricultural alert.

S — Specify the Decision

Identify the action the information may change.

“Find irrigation blocks requiring inspection” is specific. “Understand the farm better” is not.

C — Check the Spatial Fit

Ask whether the feature is large enough to influence several pixels.

An individual plant, blocked nozzle, narrow leak, or small patch may be invisible in moderate-resolution imagery.

A — Assess Availability and Timing

Determine whether usable data can arrive before the decision deadline.

Cloud cover, acquisition schedules, processing delays, and data gaps can make a theoretically suitable sensor operationally unsuitable.

L — Link the Signal to the Condition

Explain why the suspected condition should affect reflectance, temperature, radar backscatter, or another observable variable.

Without a plausible physical relationship, the map may be visually impressive but operationally weak.

E — Evaluate With Ground Evidence

Decide how the interpretation will be checked.

Possible evidence includes:

• Field scouting
• Soil testing
• Plant tissue testing
• Irrigation logs
• Weather records
• Photographs
• Yield maps
• Machinery records

SCALE Practical Fit Checklist

SCALE question	Yes	No or unsure
Is the management decision clearly defined?	1	0
Does the target feature cover several pixels?	1	0
Can usable data arrive before the deadline?	1	0
Should the condition affect an observable signal?	1	0
Is ground verification possible?	1	0

Interpreting the Checklist

• 4–5 “Yes” answers: Satellite imagery is usually a strong candidate for screening or monitoring.
• 2–3 “Yes” answers: Satellite imagery may contribute, but it should not independently determine the action.
• 0–1 “Yes” answers: Direct inspection, machinery data, field sensors, aircraft, or drones may be a better starting point.

The result indicates practical fit, not scientific certainty. It should not be used as an agricultural prescription, insurance rule, or proof of a field condition.

---

How Spatial Resolution Changes What You Can See

Spatial resolution describes the ground dimensions represented by one pixel.

Smaller pixels can reveal finer patterns, but smaller does not automatically mean more accurate. Spectral bands, image quality, atmospheric correction, acquisition timing, crop stage, and interpretation remain important.

Field-Pixel Calculation

Suppose a field covers 5 hectares.

One hectare equals 10,000 m²:

5 hectares × 10,000 m² = 50,000 m²

A 10 m × 10 m pixel covers:

10 m × 10 m = 100 m²

The field therefore contains approximately:

50,000 m² ÷ 100 m² = 500 nominal 10 m pixels

A 30 m × 30 m pixel covers:

30 m × 30 m = 900 m²

At 30 m resolution, the field contains approximately:

50,000 m² ÷ 900 m² = 56 nominal 30 m pixels

These are idealized counts. Boundary pixels may contain a mixture of crop, road, water, trees, soil, or neighboring land.

A large rectangular grain field may be represented adequately by 10 m or 30 m imagery. A narrow vineyard, irregular vegetable plot, orchard edge, or terraced field may contain a much larger proportion of mixed pixels.

The most useful resolution is the one that matches the management unit—not necessarily the smallest pixel advertised by a provider.

---

Which Tool Should Be Used First?

The best first tool is not always the one with the highest resolution. It is the tool that can answer the question at the required scale, time, and cost.

Agricultural problem	Recommended first tool	Why
Regional drought pattern	Satellite drought and vegetation products	Broad, repeatable coverage
Flooding during cloud cover	Radar satellite imagery	Microwave observations are less obstructed by clouds
Seasonal crop-development comparison	Optical satellite time series	Consistent multi-date reflectance
Exact disease diagnosis	Field scouting and laboratory testing	Diagnosis requires direct evidence
Small broken sprinkler	Direct inspection or drone imagery	The feature may be smaller than satellite pixels
Field-scale crop water-use estimate	ET product, weather data, and farm records	Water use is modeled from several inputs
Individual plant symptoms	Direct scouting	Individual plants are usually below satellite resolution
Large storm-damage footprint	Pre-event and post-event satellite imagery	Efficient area-wide comparison
Fine within-row variability	Drone, machinery, or close-range sensing	Greater spatial detail
National crop-area estimate	Satellite classification plus reference data	Scalable mapping across large regions

A productive evidence chain is often:

Satellite screening → targeted inspection → sampling or diagnosis → documented action

---

Agricultural Satellite Programs in Practice

Operational agricultural products usually combine several images, reference data, quality controls, and models. They are rarely created from one picture.

USDA Cropland Data Layer

The U.S. Department of Agriculture’s National Agricultural Statistics Service produces the annual Cropland Data Layer, a crop-specific land-cover classification.

The 2025 Cropland Data Layer:

• Has 10 m spatial resolution
• Uses harmonized Sentinel-2, Landsat 8, and Landsat 9 surface-reflectance inputs
• Uses cloud-masked 10-day median composites
• Incorporates NDVI and several spectral bands
• Uses agricultural reference data and additional land-cover inputs

See:

• USDA 2025 Cropland Data Layer metadata
• USDA Cropland Data Layer releases

The product illustrates several important principles:

1. Crop maps depend on multiple dates rather than one image.
2. Reference data are needed to train and evaluate classifications.
3. A crop label remains an algorithmic classification.
4. Resolution alone does not describe product quality.
5. Metadata and accuracy information are essential for interpretation.

USDA states that farmer-reported information cannot be derived from the public Cropland Data Layer.

The 2025 product is used here as a documented processing example. A newer annual layer may become available after this article’s review date.

OpenET and Agricultural Water Use

OpenET combines Landsat observations with an ensemble of models to estimate evapotranspiration.

Inputs include information related to:

• Land-surface temperature
• Vegetation greenness
• Weather conditions
• Solar energy
• Model assumptions

See NASA’s explanation of how OpenET uses Landsat data.

OpenET is a useful example of why processing levels matter. A temperature product, vegetation index, and evapotranspiration estimate may appear in the same platform, but they represent different types of information and uncertainty.

An evapotranspiration layer should be described as:

• Estimated evapotranspiration
• Modeled evapotranspiration
• A satellite-informed ET estimate

It should not be described as an exact, direct measurement of irrigation demand.

Illustrative Irrigation Scenario

The following scenario is hypothetical. It is not a documented farm trial and does not claim measured water or yield savings.

A manager sees a new low-NDVI strip across part of an irrigated maize field.

Instead of assuming that the crop needs more water, the manager:

1. Checks the image date and cloud mask.
2. Confirms that the strip covers several pixels.
3. Compares the pattern with earlier images.
4. Reviews rainfall and irrigation records.
5. Checks a thermal or estimated-ET product.
6. Visits the strip and a normal comparison zone.
7. Inspects crop stage, soil moisture, irrigation pressure, weeds, and drainage.
8. Records the verified cause before changing irrigation.

The map creates value by narrowing the inspection area. It does not determine the cause.

---

Can Satellites Detect Disease, Pests, or Nutrient Problems?

Satellite imagery can identify areas showing patterns consistent with plant stress. It usually cannot determine the exact cause without additional evidence.

Disease, insects, nutrient deficiency, drought, excess water, compaction, salinity, weeds, frost, lodging, poor emergence, and natural senescence can produce overlapping spectral or thermal changes.

Satellite observations may help answer:

• Where did an unusual pattern appear?
• How large is the affected zone?
• When did the change become visible?
• Is the pattern expanding?
• Does it align with irrigation, soil, drainage, or management boundaries?

They usually cannot answer independently:

• Which pathogen is present?
• Which insect caused the damage?
• Which nutrient is deficient?
• Which treatment should be applied?
• Whether a particular input is legally or agronomically appropriate?

Conditions That Affect Stress Detection

More favorable	Less favorable
The affected area covers several pixels	Only individual plants are affected
The canopy changes consistently	Symptoms remain below the canopy
Clear observations are available at the right time	Clouds block the critical period
Comparable historical imagery exists	No seasonal baseline exists
The anomaly persists across several dates	It appears once in a poor-quality image
Ground verification is possible	The field cannot be inspected

Satellite imagery can prioritize scouting. It should not replace diagnosis.

---

Common Interpretation Errors

Treating Map Color as Severity

Agricultural platforms often stretch colors to make small numerical differences visible.

A red area does not automatically indicate severe crop damage. Check the legend, numeric range, date, quality information, and comparison method.

Applying One NDVI Threshold Everywhere

NDVI changes with crop type, growth stage, soil background, canopy density, illumination, and processing.

A threshold copied from another crop or region can create false alarms. Compare similar crops at similar stages whenever possible.

Ignoring Clouds and Processing Artifacts

Thin clouds, haze, shadows, missing data, and tile boundaries can create convincing false patterns.

Check the quality layer and original scene before treating an unusual shape as a field condition.

Assuming Higher Resolution Means Higher Reliability

Fine pixels can show smaller features, but poor timing or inconsistent processing can make them less useful than a reliable moderate-resolution time series.

Resolution is one part of data quality—not a substitute for it.

Acting Before Verifying the Cause

A low-index zone may result from insufficient water, excess water, late planting, weeds, soil variation, harvest, shadow, or a cloud-mask error.

Applying water, fertilizer, or crop-protection products based only on a color map can waste resources or worsen the real problem.

Comparing Different Growth Stages

Later planting, replanting, uneven maturity, and different crop varieties can create legitimate differences between fields.

A valid comparison requires crop-stage and management context.

Treating a Health Score as a Diagnosis

A crop health score is usually defined by the platform producing it.

Before relying on a score, determine:

• Which sensor was used
• When the image was acquired
• Which bands or indices were included
• Whether the score is relative or absolute
• How clouds and missing data were handled
• Whether the method is documented
• Whether uncertainty is shown

Scores from different platforms should not be assumed to be equivalent.

---

Troubleshooting Agricultural Satellite Maps

What appears on the map	Possible explanation	What to check
Sharp rectangular anomaly	Tile edge, missing data, or processing artifact	Original scene and quality mask
Weak values around all field edges	Mixed pixels containing non-crop surfaces	Boundary accuracy and inward buffering
Sudden whole-field change	Growth, harvest, atmosphere, weather, or processing	Field records and nearby comparison fields
Known small problem is invisible	Feature is below the pixel size	Drone imagery or direct inspection
Radar values change unexpectedly	Moisture, roughness, crop structure, or geometry changed	Longer radar time series
Hot zone appears after irrigation	Bare soil, sparse canopy, timing, or weather dominates	Canopy cover and acquisition time
Two platforms disagree	Different sensors, dates, corrections, or formulas	Product documentation and metadata
Imagery looks healthy but yield was poor	Stress was temporary, hidden, late, or unrelated to greenness	Yield, weather, soil, and management records

A single unexplained map should be treated as a question, not an answer.

---

Agricultural Satellite Platform Evaluation

A useful platform should explain where its data came from, when the image was collected, and how the displayed product was created.

Data Transparency Checklist

Before subscribing or relying on alerts, confirm that the platform provides:

• Satellite or data-source name
• Image-acquisition date
• Spatial resolution
• Cloud and shadow information
• Processing level
• Index formula or product definition
• Historical image access
• Correctable field boundaries
• Downloadable values or legends
• Documentation for model-derived products
• Processing-delay information
• Quality flags or uncertainty information
• Export options compatible with the farm workflow
• A review step before an alert becomes an action

Privacy and Data Ownership

Farm boundaries, yields, crop records, and management histories can have commercial value.

Before uploading data, review:

• Who owns the uploaded field information
• Whether the provider may share it
• Whether it may be used to train models
• Whether users can delete historical records
• Whether information may be supplied to insurers, suppliers, or other third parties
• What happens to stored data after account closure
• Whether exported records remain usable after cancellation
• Which law and jurisdiction govern the agreement

This is a practical contract-review checklist, not legal advice. Material agreements should be reviewed under the laws and contractual rules applicable to the farm and service provider.

Warning Signs

Be cautious when a service promises:

• Guaranteed yield increases
• Perfect irrigation recommendations
• Automatic disease diagnosis
• Real-time imagery without explaining delays
• Unlimited or near-perfect accuracy
• Input recommendations without field context
• A health score with no methodology
• Crop diagnosis without ground verification
• Commercial superiority without a documented comparison

No satellite platform should be treated as an automatic agronomic authority.

---

Benefits and Limitations of Agricultural Satellite Imagery

Core Benefits

• Repeated coverage over large areas
• Consistent comparison among fields
• Historical archives for seasonal analysis
• Spectral information beyond normal vision
• Faster prioritization of scouting and sampling
• Support for crop classification and damage assessment
• Public access to major Landsat and Sentinel datasets

Landsat archive products have been available for download without data charges since 2008.

See USGS Landsat data access.

Copernicus Sentinel data are available through the Copernicus Data Space Ecosystem, which provides free access to Sentinel data and related discovery, viewing, download, and processing services.

Core Limitations

• Clouds and shadows obstruct optical imagery
• Pixels may contain several surface types
• Different problems can create similar signals
• Small or rapidly developing problems may be missed
• Model outputs contain assumptions and uncertainty
• Image availability may not match the decision deadline
• Ground evidence remains necessary

Satellite information is most valuable when it reduces the area that must be inspected or improves the timing of an investigation.

---

Recommendations for Different Users

Small Farms and First-Time Users

Begin with one field and one management question. Review several dates before paying for complex alerts or automation.

Do not judge a field from one image or assume that a software score explains the cause of an anomaly.

Large Farms and Agricultural Enterprises

Combine satellite time series with field boundaries, crop records, weather, machinery data, and a documented inspection process.

Automated alerts are useful only when each alert leads to a defined verification step.

Irrigation Managers

Prioritize estimated evapotranspiration, thermal information, weather, water-delivery records, and soil measurements.

Do not convert NDVI directly into an irrigation volume without a locally validated method.

Agronomists and Crop Consultants

Use imagery to choose representative sampling locations and determine when a pattern first appeared.

Do not replace field judgment with a universal vegetation-index threshold.

Insurers and Damage Assessors

Use pre-event and post-event imagery with weather records, photographs, field reports, and applicable policy requirements.

Satellite imagery can help document the extent of an event, but it should not independently determine claim eligibility or payment.

Researchers and Public Agencies

Use reproducible products with documented processing, quality flags, reference data, and uncertainty analysis.

Long-term Landsat and Sentinel archives are valuable because their documentation allows methods to be examined and repeated.

---

How This Article Was Researched and Reviewed

This article was prepared from publicly available first-party documentation rather than commercial marketing materials.

The review included:

• Checking Sentinel-2 band and resolution information against ESA documentation
• Checking Landsat 8 and Landsat 9 resolution and acquisition schedules against USGS documentation
• Checking agricultural applications against NASA Earthdata and NASA Science resources
• Checking the 2025 Cropland Data Layer against USDA NASS metadata
• Distinguishing sensor-derived observations from indices, classifications, and model estimates
• Treating evapotranspiration as a modeled estimate rather than a direct satellite measurement
• Labeling the irrigation scenario as hypothetical
• Avoiding claims that satellite imagery can independently diagnose crop disease or guarantee agricultural outcomes

No commercial agricultural platform was independently tested or ranked for this article. No independent agronomist, agricultural engineer, or remote-sensing scientist is presented as having reviewed it.

Mission status, data portals, and product methods can change. The technical statements and official links in this article were checked on August 2, 2026.

---

A Practical Starting Point

Begin with one field and one management question.

Select imagery whose resolution and timing fit that question. Review several dates, inspect the quality information, and compare both unusual and normal areas before taking action.

For regional monitoring, crop mapping, flood assessment, and seasonal comparisons, satellite data may be an efficient starting point. For individual plants, small equipment failures, urgent diagnosis, or within-row detail, direct inspection or finer-scale sensing is usually more appropriate.

Use satellite images to locate and track differences, then use field evidence to explain those differences before taking action.

---

Frequently Asked Questions

Can satellites see individual crop plants?

Most freely available agricultural satellite products cannot resolve individual plants.

A 10 m pixel covers 100 m², while a 30 m pixel covers 900 m². Commercial satellites, aircraft, and drones may provide finer detail, but visibility still depends on plant spacing, image quality, canopy structure, and the surrounding surface.

How often are agricultural satellite images updated?

Update frequency depends on the mission, location, sensor, processing system, and cloud conditions.

A revisit schedule describes when a satellite may have an opportunity to observe an area. It does not guarantee a usable image. Optical observations may be unavailable for extended periods in cloudy regions, while radar can provide additional coverage.

Can satellite images identify a specific crop disease?

Satellite imagery can reveal areas that may require disease scouting, but it usually cannot identify a specific pathogen.

Water stress, nutrient problems, insects, weeds, soil variation, frost, and disease can produce similar spectral or thermal patterns. Diagnosis requires field evidence and may require laboratory testing.

Are agricultural satellite images free?

Many major public datasets are available without imagery charges.

Landsat products can be downloaded through USGS services, and Sentinel data can be accessed through the Copernicus Data Space Ecosystem. Commercial platforms may charge for processing, alerts, storage, integrations, support, or proprietary imagery.

Can radar satellites monitor fields through clouds?

Radar is generally much less affected by ordinary cloud cover than optical imagery and does not require sunlight.

Radar backscatter is still influenced by soil moisture, surface roughness, crop structure, acquisition geometry, polarization, and sometimes precipitation. Radar maps therefore require contextual interpretation.

Can satellite images replace crop scouting or soil tests?

No.

Satellite imagery provides broad, repeated evidence of surface and canopy patterns. Scouting, soil tests, tissue tests, equipment inspections, and farm records provide the direct evidence needed to identify causes and select appropriate actions.

---

Sources

Accessed August 2, 2026.

1. NASA Earthdata — Agriculture Production
   Agricultural Earth-observation variables, applications, direct observations, and modeled products.

2. NASA Earthdata — Evapotranspiration
   Definition of evapotranspiration and explanation that ET is modeled rather than directly measured by satellite instruments.

3. NASA Earthdata — Normalized Difference Vegetation Index
   NDVI definition and its relationship to red and near-infrared reflectance.

4. NASA Earthdata — Soil Moisture and Water Content
   Soil-moisture products, agricultural uses, and distinctions among satellite and modeled datasets.

5. NASA Earthdata — Soil Moisture Active Passive
   SMAP surface soil-moisture observations, coverage, instruments, and products.

6. NASA Science — Agriculture and Food Security
   Landsat applications in crop monitoring, irrigation, acreage analysis, forecasting, and disaster assessment.

7. NASA Science — OpenET Study Helps Water Managers and Farmers Put Landsat to Work
   OpenET’s Landsat inputs, model ensemble, validation, and field-scale evapotranspiration estimates.

8. ESA — Sentinel-2 Facts and Figures
   Sentinel-2 spectral bands, spatial resolutions, coverage, and applications.

9. ESA — Satellite Radar Interferometry Effective for Mapping Crops
   Sentinel-1 radar time series and interferometric coherence used for crop classification.

10. USGS — Landsat 8
    Landsat 8 instrument, spatial-resolution, orbit, and acquisition specifications.

11. USGS — Landsat Acquisition Schedules
    Landsat 8 and Landsat 9 repeat cycles, orbital offset, and acquisition scheduling.

12. USGS — Landsat Data Access
    Official information about no-cost Landsat archive access and download services.

13. USDA NASS — 2025 Cropland Data Layer Metadata
    Resolution, satellite inputs, compositing, training data, and processing information for the 2025 Cropland Data Layer.

14. USDA NASS — Cropland Data Layer Releases
    Official annual product downloads, confidence layers, and release information.

15. Copernicus Data Space Ecosystem
    Official free-access portal for Copernicus Sentinel data, visualization, downloads, and processing services.