Agricultural Research Project · Ukraine · Slovakia · Portugal

Could this crop fit this location?

Choose a country, region and crop. AgroPredict checks verified annual-rainfall and surface-soil-pH references, then gives you a simple preliminary result with the evidence behind it.

Check a Crop Explore Advanced Tools
Research focus: Ukraine (pilot) · Slovakia · Portugal · Three founder-connected countries
Start Here

Can I grow this crop here?

Select a location and crop. AgroPredict will compare verified annual-rainfall and predicted surface-soil-pH data with published FAO EcoCrop reference ranges.

Preliminary Crop Check

Designed as a simple first screening for Ukraine, Slovakia and Portugal. It does not replace field sampling or advice from a qualified agronomist.

Preliminary result
Annual precipitation
—
Surface-soil pH
—
What the result means: the headline uses only two independently checked factors, with no weighting or percentage score. Temperature is shown in the detailed tools but does not affect this result because the FAO EcoCrop temperature range does not specify a compatible time basis. SoilGrids values are model predictions at approximately 250 m resolution, not laboratory measurements.
Project Scope

A simple answer, with the evidence still visible.

AgroPredict turns two comparable public-data factors into a preliminary crop check, while keeping the limitations and source data easy to inspect.

Choose a location and crop
Select a named region in Ukraine, Slovakia or Portugal and one of five crops with verified FAO EcoCrop records.
Compare verified factors
NASA POWER annual precipitation and SoilGrids 0–5 cm pH are compared independently with published crop ranges. No hidden weights are used.
Read the result and reasons
You receive a plain-language headline plus the value, reference range and status of each factor. Advanced tools remain available for deeper analysis.
Important limitation: this is a preliminary two-factor screening, not a prediction of yield or crop survival. It does not yet account for cultivar, sowing date, irrigation, pests, management practices, full soil profile or laboratory measurements.
User Research

Who this is built for — and what we still need to learn.

The product is being tested with potential users. The research assumptions remain visible, but they are secondary to the working crop check.

View user-research status and open questions

User Research Status

No completed interviews recorded yet

Target Users Being Explored

  • Farm managers and agronomists at medium-to-large Ukrainian grain holdings (size range, decision role, and willingness to pay: unverified)
  • Agricultural cooperative managers who consolidate decisions across multiple farms (segment size: unknown)
  • Agronomic consultants advising farms in Slovakia and Portugal (professional norms, pricing expectations: unknown)
  • Agricultural researchers and students interested in climate-data tools (use-case fit: to be explored)

Research Questions Still Needing Answers

  • How do target users currently access climate and soil data for crop planning?
  • What decisions does climate data actually influence, and at what planning stage?
  • What would make a tool like this trusted rather than ignored?
  • Is the farm manager or the agronomist the decision-maker for software adoption?
  • What is the real willingness to pay — if any?
  • Do farmers in Slovakia and Portugal face similar problems to those in Ukraine?

Interview Questions Planned

  • Walk me through how you decide which crop to plant in a given field each season.
  • What data sources do you currently use for climate or soil information?
  • What would you need to see to trust a new data tool?
  • Describe a time when a poor crop decision cost you money. What information was missing?
  • If a tool gave you monthly climate indicators for your region, would that change any decisions?

What Is Verified vs. Assumed

  • Verified: NASA POWER API returns real climate data for the selected coordinates
  • Verified: Ukraine, Slovakia and Portugal have agricultural sectors where crop planning occurs
  • Assumed (not verified): Target users exist and would find this tool useful
  • Assumed (not verified): Farmers lack access to accessible climate data
  • Assumed (not verified): Any willingness to pay for such a tool

Want to see the details?

The tools below expose the underlying climate, soil, rainfall, growing-season and farm-economics calculations. They support the main crop check but no longer compete with it for attention.

Climate Soil Growing season Rainfall Farm economics
Live Climate Data

Climate Indicator Tool

Retrieves modelled daily climate data from the NASA POWER API (MERRA-2 reanalysis) for a selected region and period. This detailed tool shows temperature, precipitation and humidity; it does not produce the Crop Check headline.

Location & Period Live API
Data retrieved from NASA POWER · power.larc.nasa.gov
Location mode
The selected region is represented by one fixed reference coordinate. Results describe the NASA POWER grid cell containing that point, not the full administrative region or an individual field.
API endpoint: Select a region to preview API URL
When enabled, a second request retrieves NASA POWER daily data for 1991–2020 to calculate a numerical difference from that reference period. Requires a separate large API call — may take 15–90 seconds.
Requires internet · Fetches daily data · May take 3–8 seconds

Configure a location and period, then click Fetch Climate Data to retrieve NASA POWER modelled climate data.

No synthetic data will be shown. If the API is unavailable, an error state will be displayed.

⟳

Fetching from NASA POWER API…

Requesting daily data

⟳

Retrieving 1991–2020 historical reference…

Requesting ~30 years of NASA POWER daily data · This may take 15–90 seconds

API request failed
Unknown error

Possible causes: CORS restriction in this browser · network timeout · NASA POWER rate limit · API unavailable.
Recommendation: A lightweight backend proxy (e.g. a single Cloudflare Worker or Vercel serverless function) would resolve CORS issues and allow caching. No credentials are required — NASA POWER is a public API.
—
—
Live API data
Mean Temperature
NASA POWER T2M
—
°C · Temperature at 2 Metres · MERRA-2
—
Total Precipitation
NASA POWER PRECTOTCORR
—
mm · Precipitation Corrected · daily sum aggregated
—
Mean Relative Humidity
NASA POWER RH2M
—
% · Relative Humidity at 2 Metres · MERRA-2
—
Data Coverage
Calculated
—
days with valid data / total days requested
—
Data source: NASA/POWER CERES/MERRA2 Native Resolution Daily Data · power.larc.nasa.gov
Version: —  ·  Coordinates: —  ·  Retrieved: —
Limitation: Native grid resolution 0.5° latitude × 0.625° longitude; physical dimensions vary by latitude. Values represent a grid cell average, not a field-level measurement. Data may differ from on-site measurements.
Important: These indicators describe modelled conditions for the requested period. They do not constitute crop suitability assessments. Converting climate indicators to agronomic recommendations requires verified crop-requirement thresholds from authoritative sources (e.g. FAO crop production guides) — work not yet completed in this project.
Farm Economics Calculator

Farm Economics Calculator

Calculate revenue, production costs, gross margin and break-even values using your own farm assumptions.

Working calculator · User-provided inputs
Farm Scenario Inputs
All financial inputs are provided by you. No defaults are inserted.
AgroPredict does not insert assumed market prices, yields or production costs. Enter values that reflect your own farm, quotation, contract or planning scenario.
Scenario identity
Selecting a crop does not insert any price, yield or cost value
Country selection does not change calculations
No currency conversion is performed. All monetary inputs use the selected currency.
Production assumptions
hectares
tonnes per hectare
UAH per tonne
Variable costs — UAH per hectare
Blank optional cost fields are calculated as 0.
UAH/ha
UAH/ha
UAH/ha
UAH/ha
UAH/ha
UAH/ha
UAH/ha
UAH/ha
UAH/ha
Fixed costs (total for this scenario)
Total amount in UAH — not per hectare. Taxes, financing costs, depreciation and items not entered here are excluded.
Inputs changed — press Calculate to update results.
Financial inputs are calculated locally in this browser and are not transmitted by this prototype.
Results
Complete the required inputs and press Calculate.
Complete the required inputs and press Calculate to see results.
Production
Area
Expected yield
Total production
Revenue and costs
Total revenue
Variable cost per hectare
Total variable costs
Other fixed costs
Total scenario costs
Cost per tonne
Margins — gross margin excludes fixed costs; operating result includes all entered costs
Gross margin
Gross margin per hectare
Operating result
Operating result per hectare
Operating margin
Break-even — yield or price required for revenue to equal entered total costs
Break-even yield
Break-even price
Production:
Total production = Area × Expected yield

Revenue:
Total revenue = Total production × Sale price per tonne

Variable cost/ha:
= Seeds + Fertiliser + Crop protection + Fuel & machinery + Labour + Irrigation + Insurance + Land rent + Other variable

Total variable costs:
= Variable cost/ha × Area

Total scenario costs:
= Total variable costs + Other fixed costs

Gross margin (revenue minus variable costs only):
= Total revenue − Total variable costs

Operating result (revenue minus all entered costs):
= Total revenue − Total scenario costs

Operating margin:
= Operating result / Total revenue × 100 (when revenue > 0)

Break-even yield:
= Total scenario costs / (Area × Sale price)

Break-even price:
= Total scenario costs / Total production

Cost per tonne:
= Total scenario costs / Total production
Sensitivity analysis — operating result across yield and price scenarios

Sensitivity analysis changes only yield and sale price. All user-entered costs remain constant across the table. This is scenario analysis, not a forecast.

Percentage (e.g. 10 means ±10%)
Percentage (e.g. 10 means ±10%)
Enter yield step and price step to display the sensitivity table.
Operating result is a planning indicator, not audited profit. Taxes, financing costs, depreciation, subsidies, payment timing and costs not entered by the user are excluded.
Climate results and financial scenarios are currently calculated independently. AgroPredict does not yet have a verified model connecting these climate indicators to crop yield or financial performance. The user may manually enter a different yield assumption after reviewing climate information, but the site does not claim that the climate tool generated that yield.
Crop Screening

Crop Climate Suitability Screening

Compare the site's 1991–2020 annual precipitation mean against FAO EcoCrop published ecological ranges. Temperature figures from EcoCrop are shown as reference only — no comparison is calculated because the temporal definition of published temperature ranges is not sufficiently specific for comparison with the selected NASA POWER indicator. Each factor is reported independently. No overall score is calculated.

Run the NASA POWER climate analysis for a location before starting crop screening.
Soil Profile

Soil Property Profile

pH, soil organic carbon, sand, silt, clay, cation exchange capacity and total nitrogen for the 0–5 cm depth layer. Source: ISRIC SoilGrids 2.0 via WCS at approximately 250 m spatial resolution. Values are model predictions, not field measurements.

Run the climate/location analysis first to select a location for soil data.
Tools

Growing Season Calculator

Calculate accumulated growing degree days, frost-exposure days and heat-threshold days from NASA POWER modelled daily grid-cell climate data. Available after analyzing a location with the Climate Tool.

Analyze a location using the Climate Tool above to enable the Growing Season Calculator.
Tools

Rainfall Pattern & Dry Spell Analyzer

Calculate selected-period precipitation totals and ETCCDI/Climdex-style wet-day, heavy-rainfall and consecutive dry/wet spell indicators from NASA POWER daily PRECTOTCORR data.

Analyze a location using the Climate Tool above to enable the Rainfall Pattern Analyzer.
Research Markets

Three countries. Three reasons.

AgroPredict focuses on Ukraine, Slovakia and Portugal for reasons directly connected to the founder's personal experience — not because of modelled market scores. No market-opportunity numbers are shown here; none have been validated.

All descriptive content is qualitative context — not validated market analysis. No numerical market-opportunity scores are assigned to any country.

Research Rationale

Why these three countries.

Each country is selected for a specific personal reason. This is not a market-attractiveness model — it is an honest statement of how the research scope was chosen.

🇺🇦

Ukraine — Pilot

Ukraine is the founder's home country and the primary research context for AgroPredict.

The public beta now includes the preliminary Crop Check plus climate and SoilGrids tools for Ukrainian regions. It has no validated commercial customers or partnerships, and yield datasets remain unintegrated.

🇸🇰

Slovakia — Validation target

The founder studied at university in Slovakia for three years. This personal connection provides a basis for understanding Central European agricultural context and accessing local networks for future user research. Slovakia operates under EU agricultural policy and shares data infrastructure with the broader European data environment.

AgroPredict has not launched in Slovakia. No partnerships, customers, or pilots exist in this market.

🇵🇹

Portugal — Validation target

The founder studied in Portugal for part of a semester through an exchange programme and participated in research and volunteer experiences connected with agricultural entrepreneurship. Portugal's Mediterranean climate — characterised by seasonal drought and heat stress — offers a meaningfully different climate context from Ukraine and Slovakia.

AgroPredict has not launched in Portugal. No partnerships, customers, or pilots exist in this market.

Business Model

Business model hypothesis — not a validated plan.

The following describes a potential business model structure. None of these elements have been tested with real customers. Pricing, subscription tiers and revenue projections have not been validated.

Pricing has not been validated. Subscription plans will be designed only after user interviews and operating-cost analysis are complete. The numbers previously shown on this page ($200/month Pro, $6,000–15,000/year Enterprise) were illustrative placeholders with no basis in customer research or cost modelling — they have been removed.
Hypothesis 01 Not validated
SaaS subscription

Farms and agribusinesses pay a recurring subscription for access to climate indicators, soil data and crop planning tools. Price point, billing cycle and feature packaging: unknown — requires willingness-to-pay interviews.

Hypothesis 02 Not validated
Enterprise or API licensing

Agricultural cooperatives, insurance companies or agricultural banks pay for bulk data access or white-label integration. Whether these organisations would pay for this type of data: unknown.

Hypothesis 03 Not validated
Consulting or advisory

Climate data is delivered as part of an advisory service to agribusinesses. Whether a data-as-service model fits the agricultural consulting market in these geographies: not yet researched.

Market sizing: No TAM, SAM or SOM figures are shown because none have been calculated from verified primary sources. The ~18,000 farms figure and ~$43M SAM previously cited were illustrative calculations not traceable to a verified dataset — they have been removed. Market size research using Eurostat agricultural census data (dataset: ef_kvaareg, ef_m_farmleg) and Ukrainian State Statistics Service data is a future task.
Positioning

Where AgroPredict fits — as a research project.

A qualitative comparison of capability categories. AgroPredict's active capabilities include: climate retrieval, farm economics, crop screening and a SoilGrids profile for pH and SOC at 0–5 cm. This is not a validated competitive analysis.

Capability Traditional Consulting Farm Management Software Ag Data Platforms AgroPredict (current)
Real-time climate data retrieval ✗ No Partial ✓ Yes ✓ Working (NASA POWER)
Preliminary crop check Manual ✗ No Partial ✓ Working (rainfall + surface pH)
Soil data integration Partial ✗ No ✓ Yes (some) ✓ Working (SoilGrids)
Business impact projections ✓ Yes ✗ No ✗ No ✗ Removed — not validated
International market comparison ✗ No ✗ No Limited → Planned (3 countries)
Validated agronomic recommendations ✓ Yes (expert-led) ✗ No Partial ✗ Not yet built

Competitor capabilities based on general product category knowledge — not a primary research audit. Actual competitive landscape requires desk research and product testing.

Methodology

How the climate tool works.

Every step in the data pipeline is described explicitly, including what is verified and what is not.

01
Region selection
The user selects a country and region. Each region is mapped to a fixed representative regional reference point. Coordinates are not field-level GPS points. Results describe the NASA POWER grid cell at that location, not the full administrative region.
VerifiedApproximate
02
NASA POWER API call
A fetch() request is sent to power.larc.nasa.gov with parameters T2M, PRECTOTCORR, RH2M, community=AG, and the selected date range. No API key required.
Live API · NASA POWER v2
03
Data validation
The response is checked for HTTP errors, malformed JSON, and missing parameter keys. Days where NASA POWER returns -999 (fill value for missing data) are excluded from calculations and counted in data coverage.
Error handlingFill value: −999
04
Aggregation
Mean temperature = arithmetic mean of valid daily T2M values. Total precipitation = sum of valid daily PRECTOTCORR values. Mean humidity = arithmetic mean of valid daily RH2M values.
CalculatedDocumented formula
05
What is NOT done
The Climate Indicator Tool itself computes no crop score. The separate preliminary Crop Check applies only verified FAO annual-rainfall and soil-pH ranges, with no weighting or probability. Soil data is fetched only after a user action. No yield or business impact is predicted.
BoundaryFuture work
Previous scoring model (removed): This project previously displayed a composite 0–100 suitability score based on manually assigned base scores, fixed soil bonuses (e.g. Chernozem +8), drought penalties (−8/−18), and yield-index adjustments. These weights were not derived from verified agronomic research. The scoring engine has been removed and will not be reintroduced until each component has a named source, documented threshold, and citation.
Development Roadmap

From climate data to decision support.

A staged plan. Future milestones are directional goals — not confirmed commitments or timelines.

2025–26
Current — Working public beta
✓NASA POWER API integration (live)
✓Temperature, precipitation, humidity retrieval
✓Error handling and data-coverage reporting
✓Ukraine, Slovakia, Portugal regions
✓Preliminary crop check using annual rainfall and surface-soil pH
✓SoilGrids integration for seven surface-soil properties
✓Transparent methodology documentation
Next step
User research
→Complete first 5 user interviews
→Validate or reject target-user hypotheses
→Document actual information needs
→Assess willingness to pay (or not)
Not started · No timeline confirmed
Future
Extend crop evidence
→Add more independently comparable crop factors
→Verify crop-specific growing-season definitions
→Validate the screening method with agronomists
→SNISolos (Portugal) · VÚPOP (Slovakia)
Planned · Depends on source compatibility and validation
Long term
Decision support
→Validated suitability scoring model
→GPS coordinate-level analysis
→ML model (if yield data available)
→First paying customers (if validated)
Aspirational · Depends on prior stages
AgroPredict
Independent research project
Agricultural Technology Climate Data Open APIs Ukraine Research Prototype
Project Story

An independently developed research project at the intersection of agricultural data and international business.

AgroPredict began with a concrete observation: Ukraine — one of the world's most agriculturally significant countries — was rebuilding its agricultural sector after significant disruption, yet accessible tools for data-driven crop planning remained limited. The project explores whether open climate data, combined with soil and yield information, can provide meaningful support for crop decisions.

AgroPredict currently provides a preliminary Crop Check based on annual rainfall and surface-soil pH, climate retrieval and historical comparison (NASA POWER), user-input farm economics, detailed crop reference screening (FAO EcoCrop), climate extremes, growing-season and rainfall indicators, and seven SoilGrids surface-soil properties at 0–5 cm. Yield datasets are not integrated. The Crop Check is descriptive screening, not a validated agronomic recommendation. Farm economics uses user-provided values only.

The three countries are included because of the founder's direct personal experience: Ukraine as home country, Slovakia as the country of university study, and Portugal through an exchange semester and agricultural entrepreneurship research.

Verified data only

Every data-derived result is traceable to a named public source; farm-economics figures come only from user-entered assumptions. No synthetic climate, soil or crop-reference values are generated. Missing data is shown as missing, not filled.

Honest scope

The project is a public beta. It has no validated product-market fit, and the roadmap is directional rather than a commitment.

Documented methodology

Every calculation is documented: formula, inputs, source, and limitations. The source ledger below lists every data source used or planned.

Built for learning

This project explores whether an independent researcher can build a credible agricultural data tool using only open APIs and public datasets — without fabricating data to fill gaps.

Data & Methodology

Source ledger.

Every data source used or planned in AgroPredict is listed here. Sources marked Live API are actively queried. Sources marked Planned have not yet been integrated.

SRC-001 Live API
Temperature at 2 Metres (T2M), Precipitation Corrected (PRECTOTCORR), Relative Humidity at 2 Metres (RH2M)
Source: NASA/POWER CERES/MERRA2 Native Resolution Daily Data
Institution: NASA Langley Research Center — Prediction of Worldwide Energy Resources (POWER) Project
API endpoint: https://power.larc.nasa.gov/api/temporal/daily/point
Documentation: https://power.larc.nasa.gov/docs/services/api/temporal/daily/
Parameters used: T2M (°C), PRECTOTCORR (mm/day), RH2M (%) · community=AG
Spatial resolution: 0.5° latitude × 0.625° longitude (native grid; physical dimensions vary by latitude — this is regional-scale data and is not field-level)
Temporal resolution: Daily · Data underlying source: MERRA-2 reanalysis
Transformation applied: Mean of daily T2M values (valid days only); sum of daily PRECTOTCORR values (valid days only); mean of daily RH2M values (valid days only). Fill value −999 excluded.
Limitation: Grid-cell averages do not reflect field-level conditions. Reanalysis data may differ from on-site measurements. Coverage gaps are possible.
Licence: Public domain / open access — no key required for point queries
SRC-002 Planned — not integrated
Soil moisture — European coverage
Source candidates under evaluation:
(a) Copernicus Land Monitoring Service — Daily Surface Soil Moisture v1.0: land.copernicus.eu
(b) Copernicus Soil Water Index — Europe 1km v2: land.copernicus.eu
Status: Access requirements, CORS availability for browser use, and licensing terms not yet confirmed. Direct browser queries may require a backend proxy.
Action required before integration: Verify API access method, spatial resolution for target regions, temporal coverage, and whether registration is required.
SRC-003 Planned — not integrated
Soil type and agricultural land classification — Ukraine
Source candidates under evaluation:
(a) State Statistics Service of Ukraine (ukrstat.gov.ua) — agricultural census data
(b) data.gov.ua — open government data portal
(c) Ministry of Agrarian Policy and Food of Ukraine
(d) European Soil Data Centre (ESDAC): esdac.jrc.ec.europa.eu
Status: No dataset has been downloaded or verified. Soil type labels used in earlier versions (Chernozem, Chestnut Soil, etc.) were manually assigned without a verified primary source — they have been removed.
SRC-004 Planned — not integrated
Soil data — Slovakia
Source candidates under evaluation:
(a) VÚPOP / NPPC — Slovak soil information system
(b) Slovak National Geoportal
Status: API availability, spatial resolution and licensing terms not yet confirmed.
SRC-005 Planned — not integrated
Soil data — Portugal
Source: SNISolos — Direção-Geral de Agricultura e Desenvolvimento Rural: snisolos.dgadr.gov.pt
Status: Portal identified. Data download format, API availability, and licensing terms not yet confirmed.
SRC-006 Planned — not integrated
Agricultural census statistics — EU (farm counts, land area by country)
Source: Eurostat Agriculture Database: ec.europa.eu/eurostat/web/agriculture/database/
Relevant datasets: ef_kvaareg (farm structure by region), ef_m_farmleg (farms by legal status and size)
Status: Datasets identified. Specific figures (farm counts for Slovakia, Portugal, Ukraine) not yet extracted and verified. No numbers from these datasets appear in the current version of this site.
SRC-007 Planned — not integrated
Crop production statistics
Source: FAOSTAT: fao.org/faostat
Status: Country and regional crop-production statistics have not yet been extracted or integrated. Crop prices and yields used in an earlier prototype (for example, wheat at 4 t/ha and $220/t) were manually assigned without verified sources and remain removed. Crop ecological reference ranges used by the working Crop Check are documented separately in SRC-008 and SRC-015.
SRC-008 Active — used in comparison
Common wheat climate reference ranges — Triticum aestivum
Source title: EcoCrop Data Sheet — Triticum aestivum (id: 2114)
Institution: Food and Agriculture Organization of the United Nations (FAO) — EcoCrop database
Direct URL: https://ecocrop.apps.fao.org/ecocrop/srv/en/dataSheet?id=2114
Access date: 2026-09-01
Crop scope: General Triticum aestivum (common wheat). The source does not distinguish winter-wheat-specific requirements. No cultivar-specific values are provided.
Values used:
  Rainfall (annual) — Optimal: 750–900 mm/year; Absolute: 300–1600 mm/year; unit: mm/year; temporal basis: annual (explicitly labeled as such on the source page).
  Temperature requirements — Optimal: 15–23 °C; Absolute: 5–27 °C; unit: °C; temporal basis: not specified on the source page (labeled only as "Temperat. requir." with no annual/seasonal/growth-stage breakdown). These values are displayed as a reference but no comparison is calculated.
Transformation performed: FAO's published annual ecological rainfall reference is compared with NASA POWER cumulative precipitation for one rolling 12-month period, subject to the ≥90% data-coverage display rule (an AgroPredict product rule, not an FAO threshold). Temperature reference displayed alongside NASA POWER reanalysis data without a comparison calculation.
Limitations: EcoCrop ecological ranges are broad climatic envelopes for species occurrence, not agronomic planting thresholds. Being within a range does not guarantee crop success; being outside does not mean crop failure. No growth-stage or soil context is included. This source is used for educational reference only.
SRC-009 Referenced — used in feature design
WMO Climatological Standard Normal period 1991–2020 — definition of reference period
Source: World Meteorological Organization — Knowledge Hub, Climate Services
URL: community.wmo.int/…/wmo-climatological-normals
Access date: 2026-09-01
Relevant definition (quoted verbatim from source): "Climatological standard normals: Averages of climatological data computed for the following consecutive periods of 30 years: 1 January 1981 to 31 December 2010, 1 January 1991 to 31 December 2020, etc. (Technical Regulations)."
Use in AgroPredict: The 1991–2020 period is used as the fixed historical reference window for the Climate Anomaly feature. AgroPredict uses NASA POWER MERRA-2 reanalysis data (SRC-001) for this period — not official weather-station observations collected by WMO members. The comparison is therefore described as "AgroPredict historical reference based on NASA POWER modelled grid-cell data for 1991–2020" and not as an official WMO local climate normal.
SRC-010 Referenced — coordinates verified
Administrative capital coordinates for Ukraine (25 regions), Slovakia (5 regions), Portugal (5 regions) — GeoNames geographic database
Institution: GeoNames — open geographic database
Dataset: GeoNames.org feature records for administrative cities
URL: https://www.geonames.org/
Access date: 2026-09-03
What is sourced: Latitude/longitude coordinates of regional administrative capital city centres, used as the representative reference point for each region in COUNTRY_REGION_REFERENCE_DATA (agropredict-core-v19.js). Each record in the data file stores the GeoNames numeric feature ID and the direct feature page URL for full auditability.
Methodology: Each coordinate represents the city centre of the formally designated regional administrative capital. Coordinates are not centroids of the administrative region and are not field-level coordinates. For Ukraine, the 25 regions follow the internationally recognized administrative territory of Ukraine per Articles 133 and 134 of the Constitution of Ukraine (adopted 28 June 1996; see SRC-011). Location labels do not describe current territorial control.
Status: Ukraine: 24 oblasts + Autonomous Republic of Crimea = 25 entries. Not included: Kyiv City (special-status city), Sevastopol (special-status city) — these are excluded from the agricultural region selector per SRC-011.
SRC-011 Referenced — administrative scope
Constitution of Ukraine, Articles 133 and 134 — administrative-region scope for the agricultural region selector
Institution: Verkhovna Rada of Ukraine
URL: https://zakonst.rada.gov.ua/docs/en/constitution.pdf
Adoption date: 28 June 1996
Access date: 2026-09-03
Articles used:
  Article 133 — lists the component units of Ukraine's territorial structure: the Autonomous Republic of Crimea, 24 oblasts, and the cities of Kyiv and Sevastopol.
  Article 134 — states that the Autonomous Republic of Crimea is an integral constituent part of Ukraine.
Use in AgroPredict: The agricultural region selector includes 24 oblasts and the Autonomous Republic of Crimea (25 entries total). Kyiv and Sevastopol are excluded from the selector because they are special-status cities with distinct constitutional standing, not agricultural oblast-level entities.
SRC-012 Derived calculation — active
Climate Extremes & Risk Flags — empirical-percentile methodology for descriptive comparison of temperature and precipitation
Source data: NASA POWER daily reanalysis (MERRA-2), already described in SRC-001. No new external data source.
Historical window: 1991–2020 (same reference period as the Climate Anomaly feature).
Comparison: Same calendar window methodology — each historical year contributes one window covering the same calendar start/end month-day as the selected current period.
Empirical midrank percentile formula:
  percentile = 100 × (count of historical values below current + 0.5 × count equal to current) / n
  where n = number of valid qualifying historical windows after coverage filtering.
Qualifying-window requirements (same as Climate Anomaly):
  Temperature (T2M): historical window requires ≥90% of expected daily values; current period requires ≥90%.
  Precipitation (PRECTOTCORR): historical window requires 100% daily coverage; current period requires 100%.
  Minimum qualifying windows: 24 of 30 (AgroPredict display rule, not a NASA or WMO threshold).
Classification boundaries (AgroPredict display rules — not NASA, WMO, FAO, or agronomic thresholds):
  0 ≤ p ≤ 10: Unusually low relative to history
  10 < p < 25: Below central range
  25 ≤ p ≤ 75: Within the central historical range
  75 < p < 90: Above central range
  90 ≤ p ≤ 100: Unusually high relative to history
What this source does not provide: No new crop data. No financial data. No forecast. No drought diagnosis (these flags do not implement a recognized meteorological drought index). No agronomic recommendation. No probability of yield reduction or crop loss.
Independence: Temperature and precipitation are classified independently. No combined score or average percentile is calculated. No economic calculation uses these results.
SRC-013 WCS integration active — seven properties, 0–5 cm
ISRIC SoilGrids 2.0 — global model-based soil-property predictions
Provider: ISRIC — World Soil Information
Dataset: SoilGrids 2.0
Citation: Poggio, L., de Sousa, L.M., Batjes, N.H., Heuvelink, G.B.M., Kempen, B., Ribeiro, E., Rossiter, D. (2021). SoilGrids 2.0: producing soil information for the globe with quantified spatial uncertainty. SOIL, 7, 217–240. https://doi.org/10.5194/soil-7-217-2021
License: CC BY 4.0
Approximate spatial resolution: 250 m
Depth intervals: Six standard intervals — 0–5 cm, 5–15 cm, 15–30 cm, 30–60 cm, 60–100 cm, 100–200 cm.
Properties integrated at 0–5 cm: Soil pH (phh2o), soil organic carbon (soc), clay, sand, silt, cation exchange capacity (cec), and total nitrogen. Bulk density remains unintegrated.
Units and conversions: sand/silt/clay g/kg ÷ 10 = %; CEC mmol(c)/kg ÷ 10 = cmol(c)/kg; total nitrogen cg/kg ÷ 100 = g/kg, following SoilGrids documentation.
Access method: ISRIC Web Coverage Service (WCS), fetched by the AgroPredict backend only after the user presses the soil button.
What this source does not provide: No field measurement, laboratory result, soil suitability score, combined soil-and-climate score, or agronomic recommendation.
SRC-014 Definitions integrated
WCRP ETCCDI / Climdex precipitation indices
Definitions used: CDD is the maximum run of days with precipitation < 1 mm; CWD is the maximum run with precipitation ≥ 1 mm; R10mm and R20mm count days meeting 10 mm and 20 mm thresholds; Rx1day and Rx5day are maximum one-day and consecutive five-day precipitation totals.
Input data: NASA POWER PRECTOTCORR daily modelled grid-cell precipitation, mm/day, for the user-selected period.
AgroPredict display rule: indicators are displayed only when at least 90% of requested calendar days contain valid data. This is a product data-quality rule, not an ETCCDI threshold.
Limitations: Selected-period descriptive indicators only. No drought classification, irrigation schedule, weather forecast, crop-loss probability, or agronomic recommendation is produced.
Climdex index definitions · WCRP ETCCDI
SRC-015 Active — primary Crop Check
FAO EcoCrop annual rainfall and soil-pH reference ranges for five crops
Institution: Food and Agriculture Organization of the United Nations (FAO) — EcoCrop database
Records used: common wheat (id 2114), maize (id 2175), sunflower (id 1191), barley (id 1232), and soybean (id 1150).
Comparisons used: NASA POWER mean annual precipitation for qualifying calendar-year windows in 1991–2020 is compared with each crop's published annual rainfall ranges. ISRIC SoilGrids predicted mean pH at 0–5 cm is compared independently with each crop's published soil-pH ranges.
Soil reference-point rule: the service first queries the region's administrative reference point. If SoilGrids returns an all-zero pH tile, AgroPredict checks a fixed, deterministic sequence of eight nearby points offset by 0.05° north, south, east, west and diagonally, and uses the first non-zero pH land cell. When this fallback is used, the exact sampled coordinates are displayed in the result. This nearby cell is not a regional average or a field measurement.
Headline rule: no weights or numerical suitability score are used. Two factors within their optimal ranges produce “The checked conditions broadly match”; at least one factor within the wider absolute range produces “Possible, with conditions to review”; any factor outside the published absolute range produces “One or more checked conditions need attention.” Missing data produces a partial- or insufficient-data result.
Limitations: EcoCrop ranges are broad ecological references, not field-specific planting rules. EcoCrop does not specify a measurement depth for its pH ranges, while the SoilGrids value represents predicted surface soil at 0–5 cm. The result is therefore a preliminary descriptive screen, not a forecast, probability, yield estimate or agronomic recommendation.
Claims removed from previous version: All composite suitability scores (0–100); manually assigned market opportunity/suitability/climate/expansion scores; "Oleksandr M." representative persona and all associated figures ($800K+ input costs, $60–80K avoided loss, $2,400/yr willingness to pay); ROI slider (22% risk exposure, 35% risk reduction, $220/t wheat price, $1,200–$9,600/yr software tiers); TAM/SAM figures (~18,000 farms, ~$43M SAM, ~$23B global agritech); illustrative subscription prices ($200/mo Pro, $6,000–15,000/yr Enterprise); all synthetic regional base scores (Poltava: 72, Kyiv: 65, etc.); all manually assigned crop bonuses and drought penalties. None of these numbers were traceable to a verified primary source.

Start with one simple crop check.

Choose a crop and location, review the two checked factors, then open the advanced tools only if you need more detail.

Check a Crop Data & Methodology