Methodology & Data Sources
Agronomic decision support requires verifiable transparency. Every score, envelope boundary, and farm-gate revenue estimate is grounded in deterministic physics and published benchmarks.
How to use the evidence
Use this page to trace a report's inputs and assumptions, then challenge the limiting factors before making a field, conversion, or investment decision. A mapped or modelled result is a screening layer; local soil tests, water checks, buyer terms, professional agronomy, and legal or financial due diligence remain separate evidence.
Source trace
Reports and tools state the dataset or benchmark used for each decision layer.
Resolution
Coverage and precision vary by coordinate, dataset, time window, and user-supplied evidence.
Verification
Local manual inputs automatically override gridded HWSD2 and ERA5 baselines.
| Requirement | Observation | Fit / Shortfall Rule | Primary Source / Next Check |
|---|---|---|---|
| Climate and Soil | Coordinate-based mapped & modelled layers with stated fallback paths. | Resolution varies; local measurement replaces mapped inputs. | Open-Meteo, FAO HWSD2, NASA POWER |
| Economics | Dated benchmarks and editable assumptions feed the commercial screen. | Reference economics do not establish a buyer, forecast, or guaranteed margin. | FAOSTAT, World Bank, Local Records |
How We Price Crops for Your Location
A single global commodity price does not reflect what a farmer in Kenya receives versus one in Germany. Our multi-layer pricing pipeline bridges the gap between global benchmarks and local farm-gate realities.
Farm-Gate Price Calculation Architecture
4-Stage PipelineFarmGatePrice = (Tradability × GlobalSpot) + ((1 − Tradability) × BasePrice × PLI × FFPI_Ratio)5-Tier Price Sourcing Chain
We source prices through a cascading fallback chain: (1) Live futures from financial APIs, (2) Alpha Vantage commodity feeds, (3) Cached recent prices, (4) FAOSTAT Producer Prices by country, (5) Static FAO benchmarks. The highest-confidence available source is always used.
Dual-Market PLI Blending
Each crop has a tradability score (0–1). Highly tradable commodities like wheat use global prices; locally consumed crops like cassava are adjusted via the World Bank Price Level Index (PLI).
FFPI Inflation Normalization
Historical FAOSTAT prices are dynamically adjusted to current-year equivalents using the FAO Food Price Index (FFPI) sub-indices, tracking real macroeconomic shocks rather than flat linear inflation.
Dynamic Corridor Validation
All prices (including user overrides) are validated against a dynamic corridor. Globally traded commodities use a ±50% band; local-only crops use ±80%, preventing anomalous spikes from distorting projected gross margins.
Biological Suitability Model
Can this crop physically survive and produce yields at your coordinates? Each environmental factor is scored as a fuzzy-membership suitability (0–1) built from the crop's own threshold envelope, then combined with an AHP-weighted geometric mean and blended with Liebig's Law of the Minimum.
Parametric Fuzzy-Membership Envelope
Each factor becomes a suitability membership (0–1) from the crop's threshold envelope: thermal regime, heat-unit accumulation (GDD), frost & heat-extreme safety, chill/vernalization, cold survival, soil pH, drainage, salinity, aluminium toxicity, photoperiod, moisture adequacy, and topography (slope).
Hard Biological Gates
Crops that cannot survive are actively blocked: tropical crops with any frost exposure are capped at 15%, extreme heat beyond crop tolerance + 2°C triggers a death gate, and severe photoperiod violations cap the biological score at 20%. Below 10%, all revenue projections are zeroed.
AHP-TOPSIS Ranking Engine
After individual scoring, crops are ranked using a hybrid Analytic Hierarchy Process (AHP) and TOPSIS multi-criteria decision analysis—the same methodology used in peer-reviewed agricultural research.
Veto penalties: where pure TOPSIS falls short
TOPSIS is fully compensatory: a high score on one criterion can offset a fatal weakness on another. Taken alone it will happily rank a crop first when its biological fit is excellent but the coming season carries a severe climate hazard. Outranking methods such as ELECTRE III solve this with an explicit veto threshold, where poor performance on a single criterion cannot be bought back with strength elsewhere.
We do not run ELECTRE III as a second ranking pass — the ordering is TOPSIS throughout. We apply its veto principle as a post-ranking multiplier on four conditions: fundamentally unprofitable economics, an experimental-risk business case, severe groundwater depletion, and a high-probability seasonal climate hazard.
seasonal penalty = 1 − (expected loss × forecast reliability × ensemble probability)
The seasonal multiplier is derived, not tuned. Expected loss is set at 0.35, the mid-range of documented frost damage in stress-sensitive crops (20–80% depending on temperature, duration, cultivar and growth stage). Forecast reliability is set at 0.55 because seasonal ensembles have real but modest skill at these lead times and are known to be over-confident. A unanimous ensemble therefore costs a crop roughly 19% of its rating — the same weight we already give severe groundwater depletion.
The seasonal veto applies only to a current-season audit. In historical mode you are asking what a typical year looks like, so this season’s anomaly must not move the ranking. Biological fit is always computed from the typical-year climate, so the penalty is not counting the same hazard twice.
AHP Pairwise Matrix
Six criteria are weighted via a Saaty pairwise comparison matrix (CR < 0.01): Biological Fit, Management Ease, Market Score, Profitability, Volatility Cost, and Carbon Bonus. An optional 7th criterion—Groundwater Stress Index—activates for water-scarce regions.
TOPSIS Closeness Vector
Viable crops are ranked by their Euclidean distance to the ideal and anti-ideal solutions across all weighted criteria. A profitability veto (40% penalty for negative ROI) and groundwater depletion penalty prevent linear compensation from masking critical risks.
Data Sources & Climate Telemetry
Climate & Weather (3-Level Fallback)
Historical mode: (1) Open-Meteo ERA5 reanalysis archive, (2) Open-Meteo CDN mirror, (3) NASA POWER Agroclimatology API. Forecast mode uses CMIP6 MPI-ESM1-2-XR under SSP2-4.5 scenarios. All climate windows are derived from the FAO-defined crop calendar for the target latitude.
FAO HWSD v2.0 Soil Mapping
Soil properties are sampled from the raster grid cell containing your selected coordinates. Top-layer physical properties—pH, nitrogen, SOC, CEC, texture, AWC, salinity (ECe), ESP, and drainage—are queried from the locally hosted HWSD2 database.
Research report model card
Responsible organization: AgriAuditor · Last reviewed: 9 August 2026
Import Gap
Annual U.S. net imports (imports minus exports), shown in rounded kilotonnes from 2023 FAOSTAT Trade bulk snapshots as a screening signal.
Search Interest
Google Trends is sampled, normalized, and relative on a 0–100 scale. It is not a purchase, planting-intent, or market-demand measure.
Climate Screen
Deterministic climate-envelope output is a separate evidence layer. It does not establish soil or infrastructure feasibility for a specific field.
Primary source terms and report-specific settings are recorded on each report. See Import Gap and Search Interest. Send corrections or custom requests through the contact page.
Authoritative Data Sources
FAO Harmonized World Soil Database v2.0 · FAO FAOSTAT Producer Prices · FAO Food Price Index (FFPI) · Open-Meteo ERA5 Reanalysis · NASA POWER Agroclimatology · CMIP6 SSP2-4.5 Climate Projections · World Bank Price Level Index (PLI) · World Bank Agricultural Wage Data · Alpha Vantage Commodity Feeds · MERIT DEM Topography
This platform produces agro-economic intelligence reports intended to support—not replace—professional due diligence by qualified agronomists, financial analysts, and legal advisors. It does not constitute financial advice or investment recommendation.