The AI field analysis compares your satellite readings against published scientific benchmarks. Full benchmark coverage is available for:

Fields with crops outside this list will still receive analysis but findings will be based on general vegetation benchmarks rather than crop-specific ranges — contact hello@envirometrics.io to discuss adding your crop.

Benchmark ranges are calibrated for six climate zones:

Fields outside these regions will still receive analysis but benchmark accuracy may be reduced.

Disease and pest risk is calculated from weather conditions and growth stage. The following crops and diseases are actively monitored:

This list covers the most economically significant diseases with published epidemiological thresholds. If you grow a crop or face a disease not listed, contact hello@envirometrics.io — we can discuss adding bespoke monitoring for your needs.

Benchmarks and disease thresholds are sourced from AHDB, the American Phytopathological Society (APS), EPPO (European Plant Protection Organisation), IRRI, FAO, and peer-reviewed crop physiology literature.

Today's maximum and minimum temperature at the field's GPS location, sourced from Open-Meteo archive data. Temperatures above 32°C are flagged as heat stress. Temperatures below 5°C are flagged as cold stress risk.

Total precipitation in mm recorded at the field's GPS location for today. Values above 20 mm are flagged as heavy rain.

Maximum relative humidity percentage recorded today at the field location. Values above 85% are flagged as elevated fungal risk — at this level, fungal spores including late blight, botrytis, and rust can germinate and infect susceptible crops within hours, particularly when humidity persists overnight.

Cumulative heat units over the past 7 days, calculated as max(0, (max temp + min temp) ÷ 2 − 10°C) per day. Base 10°C is the threshold below which most temperate crops stop developing. Above 50 GDD over 7 days indicates above-average heat accumulation and faster-than-expected crop development. Below 20 GDD indicates a cooler-than-average week, meaning key growth stages such as tillering, flowering, or grain fill may arrive later than planned.

Total rainfall in mm over the past 7 days at the field's GPS location. Use this alongside NDMI to judge whether recent rainfall is sufficient to relieve any moisture stress visible on the heatmap.

Season rainfall from 1 January plotted as a cumulative line against the 5-year historical average for the same calendar period and GPS location. The blue line is the current season; the dashed grey line is the historical average. A persistent gap below the average indicates a rainfall deficit — the crop is drawing on stored soil moisture. A gap of 30–50 mm in an unirrigated field is typically sufficient to trigger detectable moisture stress on NDMI. A line running above the average warrants checking MNDWI for waterlogging risk in low-lying areas.

Daily breakdown of max temperature, min temperature, precipitation (mm), and relative humidity for the past 7 days. Temperatures below 5°C are shown in orange as a cold stress indicator. Humidity readings above 85% are shown in red as a fungal risk indicator. Use this table to correlate recent weather conditions with any changes visible in the satellite index trend line.

Full 14-day outlook with one row per day showing: maximum and minimum temperature, precipitation (mm), relative humidity, precipitation probability, wind speed (km/h), wind direction, and the spray risk rating. Use this table to plan field operations — irrigation scheduling, spray windows, harvest timing, or machinery access — with enough lead time to act before conditions change.

When you manage multiple fields, opening each one individually to check for problems is too slow. The Portfolio page solves this by fetching the latest satellite data for all your fields at once and ranking them from worst to best. Fields with the biggest decline in the selected metric appear at the top — so the first row is always the field that needs your attention most urgently right now.

Fields are sorted by percentage change between the two most recent satellite images available for each field in the last 30 days. The calculation is: (latest value − previous value) ÷ previous value × 100. A field with NDVI dropping from 0.65 to 0.55 shows −15.4%. Fields with the most negative percentage change rank highest — meaning the worst-performing field is always row 1. Fields with no data appear at the bottom.

Count of fields where the percentage change in the selected metric is below −5%. These fields have shown a meaningful decline since the last satellite pass and warrant a closer look in the field view.

Count of fields where percentage change is between −5% and +5%. Conditions are broadly unchanged since the last image. Continue monitoring on the normal schedule.

Count of fields where percentage change is above +5%. The selected metric has improved meaningfully since the last satellite pass — either natural recovery, a successful intervention, or favourable growing conditions.

The field's position in the ranking, sorted by percentage change with the biggest decline at row 1.

The field name as saved, with the group name shown below it in grey if one was assigned. Click Investigate → on any row to open that field directly in the field view.

The crop name from the most recent season saved for that field. Blank if no season has been recorded.

The most recent satellite-derived value for the selected index, taken from the latest available cloud-free image in the past 30 days. The small bar beneath the number shows the value as a proportion of the 0–1 scale — red for low, amber for moderate, green for healthy. The label updates to match whichever metric is selected in the Index dropdown.

The percentage change between the two most recent satellite images for this field. A negative number (shown in red) means the metric declined. A positive number (shown in green) means it improved. The sign convention is consistent across all metrics — red always means worse, green always means better, regardless of whether higher or lower values are desirable for a given index.

A small sparkline bar chart showing the last 6 data points for this field. The height of each bar represents the relative value at that point in time — a declining staircase pattern confirms a deteriorating trend, while rising bars indicate recovery. Bars are coloured red, amber, or green based on the absolute value at each point.

A one-word health status badge based on the current metric value. High risk (red): the value is in the lower third of the healthy range for this index. Monitor (amber): the value is in the middle range. Healthy (green): the value is in the upper range. The thresholds are adjusted per metric — for indices where lower values indicate stress (such as BSI or NDWI), the logic is inverted accordingly.

The date of the most recent cloud-free Sentinel-2 image used to calculate the current value for this field. If this date is more than 10–12 days ago, cloud cover may have prevented a more recent acquisition — use the field view to check the available dates calendar.

Opens the field view for that specific field, pre-loaded with the selected metric. Use this to drill into the heatmap and see exactly where on the field the problem is occurring — which zones are declining, and whether the issue is localised or spread across the whole polygon.

Wheat, barley, corn, oilseed rape, sorghum, cotton, soybean, potato, sugar beet, and leafy vegetables are classified as biomass-coupled crops. For these species, green canopy growth directly equals yield — more biomass means more output. All nine spectral indices are used: NDVI, EVI, SAVI, NDRE, CIre, NDWI, NDMI, MNDWI, and BSI. The analysis interprets each index relative to the expected development stage for the days since sowing and accumulated heat units.

Olive, vine, apple, citrus, avocado, almond, walnut, pistachio, stone fruits, mango, pear, cherry, peach, nectarine, pomegranate, and fig are classified as perennial crops. For these species, canopy health does not predict fruit yield — a tree with moderate NDVI may be producing excellent fruit while a tree with high NDVI may be under-producing. Vegetation indices (NDVI, EVI, SAVI, NDRE) are excluded entirely to avoid misleading conclusions. Only NDWI, NDMI, MNDWI, CIre, and BSI are used, focused on water status and soil health rather than biomass accumulation.

Before identifying problems the analysis establishes what normal looks like for this specific field right now. For biomass crops, the baseline comes from published agronomic research — expected index ranges for this crop type, in this climate region, at this growth stage. Growth stage is derived from the sowing date and accumulated growing degree days. For example, wheat at flag leaf stage in a UK temperate climate should have NDVI between 0.70 and 0.85, NDRE above 0.45, and CIre above 2.8. A reading below that range is a deviation from baseline — and the size of that deviation determines the severity of the finding.

For perennial crops the baseline comes from the field's own five-year Sentinel-2 satellite history, pulled automatically at the time of analysis. For each of the five metrics used, the system retrieves imagery from the same ±7 day window in each of the five preceding years — the same trees, the same field, the same time of year. This produces a field-specific mean, minimum, and maximum for each metric. Deviation from that field's own history is the primary signal. A metric reading two standard deviations below the five-year mean is a meaningful anomaly for that specific orchard, regardless of how it compares to published benchmarks for the crop type.

For each relevant metric the analysis assesses the last 60 days of available satellite imagery — direction (rising, flat, or declining), rate of change (slow, moderate, or rapid), and whether that trend is expected or unexpected for this crop, location, and growth stage. Context is critical: a declining NDVI on a pre-emergence field is expected and normal. The same rate of decline on a wheat field at flag leaf stage — when NDVI should be stable or rising — is a serious problem requiring immediate attention.

If fewer than three satellite observations exist for a metric within the analysis window, findings based on that metric are flagged with an explicit caution note. The analysis still runs and includes those findings — they may still be valid — but the user is warned that the trend data is limited. Cloud cover is the most common cause of sparse data, particularly in temperate climates from October to March.

The first level of analysis compares the whole-field average for each metric against the established baseline — the agronomic benchmark for biomass crops, or the five-year field history for perennial crops. This answers the question: is this field as a whole performing normally for its crop type, growth stage, and location right now?

The second level identifies which specific zones within the field are significantly below the field average — expressed as both a percentage and hectares affected. This is where the analysis becomes actionable. A field-average NDVI reading of 0.68 might look acceptable. But if 32% of the field (32.6 ha) is showing values below 0.55 while the remaining area is at 0.75, the field-average is masking a serious localised problem. Hotspot findings tell the farmer exactly how much of the field needs attention and make targeted, zone-specific treatment possible rather than blanket application.

A single anomalous metric has many possible explanations. The analysis tests for combinations of metrics that all point to the same root cause. One metric anomalous = LOW confidence — flag for monitoring, do not act yet. Two metrics aligned = MEDIUM confidence — investigate soon. Three or more metrics aligned = HIGH confidence — act now.

Seven multi-metric patterns are tested. Nitrogen deficiency: declining NDRE or CIre with stable or rising NDVI. Moisture stress: declining NDWI or NDMI with stable NDVI. Waterlogging: rising MNDWI or NDWI alongside declining NDVI. Poor establishment: high BSI with low NDVI or EVI early in the season. Suspected disease or pest pressure: localised NDVI decline that cannot be explained by moisture or nutrition patterns. Over-fertilisation: very high NDRE or CIre with flat or declining NDVI. Under-fertilisation: low NDRE or CIre alongside uniformly low NDVI across the whole field.

Four patterns are tested. Water stress: declining NDWI, NDMI, and MNDWI together — all three moisture metrics declining in the same direction is a reliable signal of water deficit. Soil health decline: rising BSI alongside declining NDMI — bare soil increasing while plant moisture falls indicates canopy loss or soil structural problems. Anomalous canopy change: an unusual MNDWI, NDWI, or NDMI decline that falls well outside the five-year baseline range for this time of year — something that has not happened in five years warrants investigation. CIre as a contributing signal only: low or declining CIre can support water stress or soil health findings but is never used alone to prescribe fertiliser on perennial crops.

Six weather inputs are used to validate or explain satellite signals: rainfall over the last 7 days and season total compared to the 5-year average for the field's GPS location; consecutive dry days accumulated to the analysis date (drought accumulation, not just a snapshot); growing degree days for the season; frost events in the last 30 days; maximum temperature over the analysis period; and a 14-day forecast summary for context.

If a satellite signal is explained by recent weather it is noted as weather-driven. For example, a decline in NDMI following 21 consecutive dry days in summer is expected — the satellite is correctly detecting moisture stress caused by drought. The finding is still reported, but the cause is clear. If the same NDMI decline persists despite adequate rainfall and normal temperatures, it cannot be explained by weather — that makes it a confirmed finding with no obvious external cause, which raises the urgency and narrows the likely causes to irrigation failure, root disease, or soil compaction.

Each confirmed finding includes: a plain-language problem description; the specific metrics that confirm it and their current values; the area of the field affected expressed in both hectares and percentage; the most likely agronomic cause; a confidence level (LOW, MEDIUM, or HIGH); an urgency rating (MONITOR, SOON, or IMMEDIATE); the recommended input type — fertiliser type or irrigation review — and whether to apply it to the whole field or only the affected zone; and an estimated cost saving from targeted versus blanket treatment.

Four things are explicitly outside the scope of every analysis. Specific fertiliser rates in kg/ha or litres/ha are never prescribed — without a soil test, any rate recommendation would be guesswork. Disease is never confirmed definitively from satellite data — it is flagged for ground inspection only. Fruit yield is never predicted for perennial crops — canopy health and fruit production are not directly coupled. Pesticide products are never recommended by name — that decision belongs to a qualified agronomist with ground knowledge.

Disease risk is calculated by combining hourly weather data (temperature, humidity, rainfall) with the crop's current growth stage. When environmental conditions match published infection thresholds for a specific disease, a risk level is assigned for that day.

No satellite imagery is used for this feature — it is purely weather and growth stage based. Growth stage is estimated from the sowing date and accumulated growing degree days.

Thresholds are sourced from AHDB, the American Phytopathological Society (APS), and EPPO (European Plant Protection Organisation).

Conditions are currently optimal for infection. Scout or apply fungicide as appropriate for the disease and growth stage.

Conditions are partially favourable. Monitor closely over the next 48 hours.

Some conditions met but below the full infection threshold.

No risk detected for this date — conditions do not meet the threshold for this disease.

The Disease Risk column in the 14-day forecast table shows the highest-risk disease per day. Hover a pill to see the full list of diseases and their levels for that day.

The Disease & Pest Risk panel shows a horizontally scrollable 21-day grid — 7 days back to 14 days forward — with one row per disease and coloured pills for each day where risk is detected. Only diseases with at least one risky day are shown; diseases with no detected risk are hidden.