Forest Os Global
Global map0.25° overview aggregated from H3 resolution 7 · loading
Global data layerLOADING
Checking the sourceno substitute data
Map signalscoverexcess moisturedrought / firesloss / degradationresearch forecast
2026
no dataH3 cells
319months of data
14public layers
Open methodologysources and limitations ›
PolygonClick on the map to add points.
Access to analytics

From a public map to an evidence-based report

Detailed indicators, scenarios and exports are available after payment or under contract.

One-off report

Basic area analysis, PDF

from 49 000 ₽
Subscription

Ongoing portfolio monitoring

from 20 879 ₽/year
For business

Data exchange, team roles and service terms

on request
AI interpretation grounded in actual data

Area-specific findings are generated only after calculation using available data layers and include the data date, limitations and evidence level. Synthetic answers and non-production events are not published.

Verifiable resultsA basic report is prepared within 3–5 business days after payment and receipt of sufficient data. Forest asset reviews and scenario assessments follow an agreed schedule.
Public datageneralised overview without internal thresholds
Scientific methodologyverifiable and transparent
Independent sourcesGFW, ESA, FIRMS and ERA5 are used for comparison but do not replace independent validation
Programmatic data access and reportsavailable after payment or under contract

The public map shows a 0.25° overview surface aggregated from H3 resolution 7. It is not exact H3 detail; parcel-level conclusions are available only in a report with sources and limitations.

Reports Forest OS

A report for a specific decision

Choose the depth of analysis. Before payment, we confirm the area, deliverables, available sources and delivery time. Every report states its data date, limitations and clear next steps.

Initial assessment

Basic report

49 000 ₽

For a rapid assessment of one area using available data.

  • area map and profile;
  • history of available indicators;
  • observed risk signals;
  • sources, limitations and recommended next steps.
Check an area
Planning

Scenario assessment

from 450 000 ₽

For comparing area-management options under an agreed brief.

  • baseline and business-as-usual scenarios;
  • intervention and monitoring options;
  • ranges of assumptions and risks;
  • review points and conditions for revisiting the decision.
Request scenarios
Limits of use. Model-based estimates do not guarantee future results and do not replace mandatory field surveys, an official opinion or independent validation where these are required. Methodology · Validation
Order a report

Check an area and receive a report

Выберите участок на карте, введите координаты или нарисуйте полигон. Бесплатная сводка показывает агрегированные показатели; полный анализ, сценарии и рекомендации включаются в PDF после оплаты.

1. Area

If no area is selected, clicking the map fills in the coordinates. The free summary is limited to 10 000 ha. Area reports use detailed L1/L2 analysis, not just H3 resolution 7.

Mean vegetation index (NDVI)after calculation
Burn historyafter calculation
Degradation riskafter calculation
Area / biomeafter selection
Report detail leveldetailed area analysis

2. Report type

Delivery terms

A basic report is emailed within 3–5 business days after payment and receipt of sufficient data. Delivery dates for other reports are agreed before payment.

Workspace

Sign in to Forest OS

Individual customers receive a shared workspace after selecting an area and completing a short registration. Institutional access is provided under contract only.

Already have an account

Use your email and password. The system opens the individual or institutional workspace according to your verified access rights.

Sign in

Forest Os Global / Solutions for forest territories

From understanding forests —
to a plan of action.

Analytics, restoration and forest asset management. Choose a task and we will define the scope, data requirements and deliverables for your team.

Forest OS analyticsFieldwork with specialist providersScope and price agreed before work begins

A specific task.
A clear deliverable.

From a consultation on an existing report to a territory management strategy. Every package specifies the deliverables, analytical scope and delivery terms.

Loading current packages and prices…

What the base price covers. Project packages cover one area of up to 1 000 ha, standard available data and one round of revisions. Consultation and training limits are specified on their cards. Field visits, commercial imagery, laboratory work, mandatory project documentation and additional calculations are charged separately. The final price and tax terms are confirmed in the proposal before work begins.

From analysis
to implementation.

Field services and supplies are quoted individually. Availability, region and specialist provider are confirmed before contracting; there is no single rate for every territory.

Loading field services…

Special programmes

For tasks requiring a dedicated methodology, training programme or external specialists.

Your task first. Then a proposal.

01 / DESCRIBE YOUR TASK

Area and objective

Choose a service and specify the region, approximate area and the outcome your team needs.

02 / AGREE THE TERMS

Data, scope and price

We check data sufficiency and agree the providers, schedule and deliverables. Where necessary, we propose a preliminary survey.

03 / DELIVER TO THE BRIEF

Results with explanations

We deliver the agreed materials with sources, limitations and next steps. Fieldwork follows a separate agreed estimate.

Discuss your project?

Leave your contact details and a short description. We will clarify the input data and prepare a proposal.

Submitting an enquiry does not create an obligation to pay. Work begins after the terms are agreed.

Contact Forest OS ↗

Choose a service; its name will be included in your enquiry.

How does an analytics package differ from official project documentation?

Analytics packages help assess data and inform management decisions. They do not replace mandatory surveys, official project documentation, independent validation or certification. Model estimates and scenarios include limitations; the readiness of a particular method is checked when agreeing the brief.

What if my territory exceeds 1 000 ha?

Include the area and task in your enquiry. We will prepare a separate quote for large territories, portfolios, additional data sources or integrations. Prices do not increase automatically at a single per-hectare rate.

Are one-off reports and subscriptions included in service prices?

No, unless the proposal explicitly includes them. Services on this page cover project work and support. Terms for standalone reports and platform access are available under “Pricing”.

Forest Os Global / Pricing and access

Your scale.
Your workspace.

One-off area analysis, regular monitoring or forest estate management. Choose your format and compare capabilities, area limits and pricing.

All prices in Russian rublesAnnual subscriptionsClear data and implementation terms
01 / Ongoing access

From a single area
to a forest portfolio.

Your plan sets the access limits: area, sites, number of reports and tools. Module availability and applicability are confirmed before contracting.

For small landowners

Personal

20 879 ₽
per year · access under contract
  • Up to 500 ha / up to 3 areas
  • Monthly monitoring and alerts
  • 2 basic reports per year
For professional land users

Land Pro

68 727 ₽
per year · access under contract
  • Up to 5 000 ha / up to 10 areas
  • More frequent alerts under the agreed schedule
  • Browser app, documents and 6 reports per year
For forest businesses and teams

Team

400 000–800 000 ₽
per year · access under contract
  • Up to 50 000 ha / 10 users
  • Basic forest inventory estimates, reporting and data export
  • Notifications and 12 reports per year
For large area portfolios

Portfolio

Custom quote
Annual contract · terms on request
  • Up to 250 000 ha / 30 reports per year
  • Scenarios and limited data exchange
  • Organisation-wide single sign-on quoted separately
For large portfolios

Enterprise

Agreed service levels, single sign-on, data exchange and dedicated support. Area, usage rights and modules are specified in the contract.

Custom quote

Annual contract. Additional large-portfolio rate: 1.74–4.35 ₽/ha

Not sure which plan to choose?

Initial recommendation based on area and number of sites.

02 / Comparison

What your subscription includes

Limits not specified in the base package are agreed during onboarding. On small screens, scroll the table horizontally.

Compare subscription capabilities and limits
CapabilitiesPersonalLand ProTeamPortfolioEnterprise
Portfolio area, haUp to 500Up to 5 000Up to 50 000Up to 250 000Under contract
Number of areas310Under contractUnder contractUnder contract
TeamAs agreed during onboardingAs agreed during onboarding10 usersUnder contractUnder contract
Reports per year261230Under contract
Monitoring frequencyMonthlyHigher frequency under an agreed scheduleUnder the project scheduleUnder the project scheduleCustom service schedule
Oracle optionClient access — subject to plan termsClient access — subject to plan termsClient access — subject to plan termsClient access — subject to plan termsAccess level and capabilities — as agreed
ScenariosNot includedNot includedSeparate packageAs specifiedAs agreed in the brief
ExportIncluded in reportsReports and documentsIncluded as specifiedIncluded as specifiedUnder contract
Data exchange with other systemsNot includedNot includedSeparate packageAgreed datasets for external systemsExtended data exchange — separate module
Employee single sign-onNot includedNot includedUnder a separate briefOptionalUnder contract
Service and supportAs agreed during onboardingAs agreed during onboardingAs agreed during onboardingAs agreed during onboardingCustom service schedule / dedicated specialist

The site is a separate area of analysis. The personal rate includes 2 basic reports per year. The composition of reports and the rules for using the annual limit are fixed when you connect. Subscription does not mean unlimited calculations, uploads or analyst work.

03 / One-off tasks

A report for a specific decision

The range takes into account the composition of raw data and analytical work. First, agree on the boundaries, applicability of methods, materials and final cost.

Basic evidence-based report

Basic area analysis

49 000–79 000 ₽
one-off · up to 1 000 ha
  • Observation history and trends in key indices
  • Condition, anomalies, fire history and moisture stress
  • Area map, sources and limitations
Format, delivery time and larger areas

Print-ready document (PDF), area map (GeoJSON); tables (CSV/XLSX) if required

3–5 business days after sufficient data are received

Above the base area: a landmark of 4.35–8.70 ₽ /ha on an agreed scale. The final price depends on the data and amount of work of the analyst.

Comprehensive forest asset due diligence

Forest asset review

180 000–350 000 ₽
one-off · up to 5 000 ha
  • Everything in the basic report plus analyst input
  • Model-based forest inventory estimates where accuracy is sufficient
  • Review of risks and supporting evidence; comparison of available documents
Format, delivery time and larger areas

Analytical report and agreed exports

Schedule agreed after data review and approval of the brief

Above the base area: a landmark of 6.96–13.05 ₽ /ha on an agreed scale. The final price depends on the data and amount of work of the analyst.

Scenario and investment assessment

Scenario assessment

450 000–900 000 ₽
one-off · up to 5 000 ha
  • Four scenarios: baseline, stress, restoration and forest use
  • Ranges for project net present value, revenue and costs
  • Action plan and limitations of model-based findings
Format, delivery time and larger areas

Assessment, scenario comparison and intervention-zone map

Schedule and scenario applicability defined in the brief

Above the base area: a landmark of 13.05–26.10 ₽ /ha on an agreed scale. The final price depends on the data and amount of work of the analyst.

Model estimates include sources and limitations. Forest inventory characteristics require sufficient accuracy; comparison with FGIS LK is possible only with authorised access. Scenario assessments follow an agreed brief and do not guarantee investment returns.
04 / Corporate and public institutions

Operations centre
for your territory.

We will calculate the license, implementation and support for the tasks of the region. Before the contract, we fix the composition of modules, boundaries of responsibility, readiness of integrations and measurable terms of service.

Regional operations workspace

Region: Core

Custom quote
Annual contract · terms on request
Implementation: custom quote one-off
  • Operational map, priority zones and event register
  • Roles, checks, reporting and exports
  • Preparation for federal information system integration, subject to authorised access
  • Business-hours support under the contract
Regional operations workspace

Region: Extended

Custom quote
Annual contract · terms on request
Implementation: custom quote one-off
  • Everything in the core regional package plus extended assessments
  • Forest inventory, restoration and data exchange as specified
  • Oracle recommendations and probabilistic hypotheses reviewed by a specialist
  • Sensor integration and seasonal 24/7 support under contract
Start with a verifiable result

Paid pilot

The territory, deadline and criteria for the result are agreed in advance. Check the work on your task, then decide on a one-year contract.

20–30%

of the annual price

50–100% of the pilot fee credited towards the annual contract, subject to agreed terms.

Complex hardware integrations, cameras, exits and equipment - on a separate task. The composition of the expanded regional package and the calculations available for the territory are confirmed before connection.

05 / An option within your plan

Oracle: from explanation
to informed decisions.

Oracle helps interpret Forest Os Global data, compare evidence and explore courses of action. Its access level determines which data and tools it can use.

Oracle is an option within your main plan. Capabilities and pricing are specified in the main contract; no separate subscription is required. If the option is already included, you do not pay for it twice.
Understand the bigger picture

Public access

  • Explanations of forest condition indicators
  • Overview of published map data
  • Sources and limitations

Open explanations on the public map

Open the public map ↗
Work with your areas

Client access

  • Answers about your areas and observation history
  • Explanations of risks, reports and available scenarios
  • Work with client portfolios and materials

An option for individual or corporate workspaces

Coordinate your team

Operational access

  • Review events and the evidence behind decisions
  • Compare available documents and observations
  • Decision log and action approval controls

Project-based onboarding for organisations

Compare long-term options

Strategic access

  • Development scenarios for territories and forest assets
  • Compare economic assumptions and risks
  • Alternatives, limitations and testable hypotheses

Custom configuration based on your brief

Oracle capabilities by access level
CapabilitiesPublic accessClient accessOperational accessStrategic access
DataPublic indicators and the general mapOnly authorised client areas and documentsAgreed organisational data and documentsIsolated customer data and assumptions
AnalysisExplanations of indicators and their limitationsObservation history, changes and report explanationsData quality checks and event comparisonCompare risks, objectives and economic assumptions
ScenariosNot includedAvailable operational scenarios relevant to the taskCompare courses of action under the agreed briefLong-term portfolio development scenarios
WorkflowBrowse and ask about published dataWork with areas, reports and field observationsDecision log, specialist review and approvalCompare alternatives and supporting evidence
AccessOpen explanations on the public mapAn option for individual or corporate workspacesProject-based onboarding for organisationsCustom configuration based on your brief

What determines the option’s terms

  • Selected access level and tasks
  • Area and number of sites
  • Number of users and calculation volume
  • Connected sources, systems and support level

What we agree before onboarding

Sites, sources, available functions, number of users, settlement limits and maintenance procedure. Extended loops connect in a consistent configuration for the organization and do not allow access to other customers’ data or internal management of the platform.

Observations and scenarios differ. Conclusions are accompanied by sources and limitations; significant actions are confirmed by a specialist. Available settlements for your territory are checked before the contract.

06 / For integrators and developers

API
with other systems

Annual agreement, agreed limits and separate payment of excess by area, calculations or forecast units. Derivative results are licensed.

Data exchange with external systems

Developer

434 982 ₽
annual contract · overage charged separately
  • Sandbox
  • Derived indices and low-risk features
  • Quotas agreed during onboarding
Data exchange with external systems

For services and applications

Custom quote
Annual contract · terms on request
  • Higher throughput
  • Automatic event notifications and available forecasting methods
  • Limited commercial use
Data exchange with external systems

Enterprise data exchange

Custom quote
Annual contract · terms on request
  • Agreed service levels, batch calculations and dedicated quotas
  • Private endpoints for an approved client
  • Redistribution rights defined in the contract

In the tariffs "Personal" and "Land User" program data exchange is not included. The Portfolio tariff provides limited results for external systems; extended access is agreed separately. Raw unloading of the whole cube and internal models are not included in these tariffs.

07 / Extend your capabilities

Modules for your task

They are connected separately or included in a harmonized corporate specification. The same module is not charged twice if it is already included in the contract.

Separate module

Fire safety

Priorities of territories, readiness and work with events. Composition of operational methods - after verification of applicability.

Custom quote

Additional rate: 1.74–6.96 ₽/ha per year.

Separate module

Forest inventory and verification

Inventory estimates and monitoring. Detailed analysis is charged separately; field validation is not included.

4,35–17,40 ₽/ha per year

Detailed analysis: 8.70–34.80 ₽/ha.

Separate module

Forest restoration

Project and subsequent monitoring. Connection after checking the readiness of the module and agreeing on the field circuit.

2,61–8,70 ₽/ha per year

Onboarding: 8.70–34.80 ₽/ha, one-off.

Separate module

Forest risk analytics for insurance

Analysis of Portfolio Risk, Events and Implications for Insurance Teams It is not a substitute for an insurance decision.

Custom quote

Insurance pilot: on request.

Next step

Let us build your
workspace.

Specify the selected package and territory. We will prepare a proposal with the final cost, composition of the result and connection conditions.

An enquiry does not charge you or create a paid subscription.

Before onboarding

Important terms

Aligning cost, expectations and results.

What is charged separately?

Commercial satellite imagery, ground surveys, laboratory work, equipment, complex integrations, travel and exceeding agreed quotas. These positions are highlighted in the proposal before work begins.

Can I pay monthly?

Tariffs “Personal”, “Land User”, “Team”, “Portfolio”, “Corporate” and data exchange packages are designed for a year. Schedule of payments, transfer between packages and revision of limits are agreed in the contract. There is no automatic monthly write-off on this page.

How do I change plans?

You report a new area, number of territories or module requirements. We will agree on new limits, date of transition and cost calculation. Access rights change after the terms and conditions are issued, not just when choosing a card.

Does Oracle make decisions on its own?

No. Oracle helps analyse data and possible decisions. Findings include sources and limitations, and significant actions require specialist approval. Event-cause classification remains a probabilistic hypothesis.

What determines the price and custom quote?

It depends on the area, refresh rate, data, volume of analytical work, level of service, connections and rights of use. Prior to the contract, these parameters and the total amount are fixed. The tax conditions are set out in the proposal.

Conversion of USD-denominated items: Central Bank of Russia rate on 03 September 2026 — 86.9963 ₽ per 1 US dollar. Prices are rounded to the nearest ruble and per-hectare rates to the nearest kopeck; existing ruble ranges are preserved. This is a fixed pricing edition, not an automatic daily revaluation. Existing contracts and amounts of previously created orders remain unchanged.
For business

Institutional access

Для крупных лесных портфелей, банков, страховых, государственных и научных команд: согласованный доступ к данным, API, роли, уровень обслуживания и настройка отчётов под задачи.

Enquiry

Your enquiry will be sent to the service provider. Scope, price and a delivery period of up to 30 calendar days are agreed in the specification or contract before payment.

Forest Os Global / Public information

Forest territory data.
A clear API.

Documentation for developers and integrators: available endpoints, parameters, responses, limits and pricing. From your first request to an agreed data integration.

API documentation 2.1.1 · 03 September 2026 · Validation statuses have their own dates

01 / First request

Get a time series by coordinates

The application programming interface (API) delivers Forest Os Global data to your system. The six available public read endpoints below require neither registration nor an API key. Private client endpoints, internal website requests and payments are outside this specification.

Base URLhttps://forestosglobal.com
TransportHTTPS · GET · JSON UTF-8
Space / timeH3 resolution 7 · calendar month
cURL · live public endpoint · two months of NDVI
curl --get "https://forestosglobal.com/api/public/h3-at-point" \
  --header "Accept: application/json" \
  --data-urlencode "lat=55.75" \
  --data-urlencode "lon=37.62" \
  --data-urlencode "var=ndvi" \
  --data-urlencode "start=2026-06" \
  --data-urlencode "end=2026-07"

The line continuations above are for Bash. For PowerShell, use the single-line command below. Coordinates are latitude lat and longitude lon in WGS 84 degrees. Dates are explicit and both ends of the period are included. Start with one indicator and a short period.

Windows PowerShell · single-line command
PowerShell
curl.exe --header "Accept: application/json" "https://forestosglobal.com/api/public/h3-at-point?lat=55.75&lon=37.62&var=ndvi&start=2026-06&end=2026-07"
02 / Endpoint reference

What to request and what to send

Parameters are passed in the query string after ?. A GET request body is not required. Values of var are case-sensitive; use the names below without spaces. These endpoints do not support pagination or multi-cell selection.

GET /api/public/h3-at-point

Time series by coordinates

Resolves the coordinates to an H3 resolution 7 cell, then returns its monthly values. The result applies to the whole cell, not an individual tree or pixel.

Parameters for /api/public/h3-at-point
ParameterType / requiredRuleExample
latnumber / yesWGS 84 latitude in decimal degrees, using a decimal point.55.75
lonnumber / yesWGS 84 longitude in decimal degrees. Do not swap lat and lon.37.62
vararray / noLowercase names, comma-separated, without spaces. If omitted, all currently allowed indicators are returned; explicit selection is recommended. Maximum 20 names; duplicates are unnecessary. Check availability at summary.spectral_indices.indices.<name>.public_ready.ndvi,evi
startstring / noFirst month, inclusive. Defaults to 11 months before end. The start must not follow the end; the maximum period is 12 calendar months, inclusive. Explicit dates are recommended.2026-06
endstring / noLast month, inclusive. Defaults to the server’s current UTC month, not the latest month with data. Explicit dates are recommended.2026-07
Example URL
https://forestosglobal.com/api/public/h3-at-point?lat=55.75&lon=37.62&var=ndvi&start=2026-06&end=2026-07
GET /api/public/h3-cell

Time series by cell identifier

The same value format, without query_point. Use this if you already have the H3 identifier. One cell and up to 12 months per request; this endpoint does not support multiple cells or full-cube exports.

Parameters for /api/public/h3-cell
ParameterType / requiredRuleExample
h3string / yesOne H3 resolution 7 cell present in the land cube. Use a lowercase hexadecimal identifier. The cell parameter remains a legacy alias; use h3.8711aa7aaffffff
vararray / noLowercase names, comma-separated, without spaces. If omitted, all currently allowed indicators are returned; explicit selection is recommended. Maximum 20 names; duplicates are unnecessary. Check availability at summary.spectral_indices.indices.<name>.public_ready.ndvi,evi
startstring / noFirst month, inclusive. Defaults to 11 months before end. The start must not follow the end; the maximum period is 12 calendar months, inclusive. Explicit dates are recommended.2026-06
endstring / noLast month, inclusive. Defaults to the server’s current UTC month, not the latest month with data. Explicit dates are recommended.2026-07
Example URL
https://forestosglobal.com/api/public/h3-cell?h3=8711aa7aaffffff&var=ndvi&start=2026-06&end=2026-07
GET /api/public/summary

Overview and indicator registry

Returns the platform overview and index availability. generated_at is the summary generation time, not the latest observation date. Select var using public_ready, not the total number of registered indices.

Parameters for /api/public/summary
ParameterType / requiredRuleExample
If-None-Matchstring / noETag from a previous response at the same URL. If it matches, the server returns 304 without a body. Supported by summary and validation.W/"etag-from-previous-response"
Example URL
https://forestosglobal.com/api/public/summary
GET /api/public/validation

Dated validation snapshot

Aggregated validation results and limitations. HTTP 200 may contain ok=false when no snapshot is available. stale means the information is outdated; date_unconfirmed means the date is unverified. The number of checks is not a model accuracy percentage.

Parameters for /api/public/validation
ParameterType / requiredRuleExample
If-None-Matchstring / noETag from a previous response at the same URL. If it matches, the server returns 304 without a body. Supported by summary and validation.W/"etag-from-previous-response"
Example URL
https://forestosglobal.com/api/public/validation
GET /api/public/tariff-catalog

Plan and option catalogue

Public packages in rubles, including api and oracle_option. price_mode=quote indicates a custom quote: no numeric price is supplied. This catalogue does not create orders, charge payments or grant access rights.

No parameters required.

Example URL
https://forestosglobal.com/api/public/tariff-catalog
GET /api/public/pricing-settings

Price labels for interfaces

A compatibility dictionary of price labels. Use tariff-catalog to select a package. If a key is in quote_required_keys, settings[key]=null: display labels[key], not a zero price.

No parameters required.

Example URL
https://forestosglobal.com/api/public/pricing-settings
03 / Data contract

Month first. Cell second.

Axis order matters: variables[name][month index][cell index]. The outer array corresponds to months; the inner array corresponds to h3. A public response contains one cell, so its index is always 0. The example below comes from a public endpoint; it illustrates the response format, not a forest condition benchmark.

HTTP 200 · Example response for NDVI, June–July 2026
{
  "ok": true,
  "h3": [
    "8711aa7aaffffff"
  ],
  "months": [
    "2026-06",
    "2026-07"
  ],
  "variables": {
    "ndvi": [
      [
        0.2637999951839447
      ],
      [
        0.19259999692440033
      ]
    ]
  },
  "encodings": {
    "ndvi": {
      "dtype": "int16",
      "fill_value": -32768,
      "scale": 0.0001,
      "offset": 0.0,
      "units": "1"
    }
  },
  "encoded": false,
  "resolution": "H3 r7",
  "public_policy": "one_cell_max_12_months_no_raw_imagery",
  "query_point": {
    "lat": 55.75,
    "lon": 37.62
  }
}
Time-series fields
FieldMeaning and handling
okApplication-level success. Check this after the HTTP status.
h3[]Cell IDs; there is exactly one in this contract.
months[]Actual selected months in ascending order. Do not reconstruct months from your request. A period outside the available time axis may return 503 rather than a successful empty response.
variables.<name>[][]A number or null. July NDVI in the example is variables.ndvi[1][0]. JSON field order does not determine time order.
encodings.<name>dtype, fill_value, scale and offset describe the source encoding; units describes the returned units.
encodedfalse: values have already been decoded into physical units. Do not multiply by scale again.
resolution / public_policySpatial resolution and public-response limitations.
query_pointThe original lat/lon, present only in h3-at-point. These are not the cell-centre coordinates.
04 / Indicator glossary

Nine indices and four additional layers

Indicators in the public time series
Name for the var parameterIndicatorUnitsHow to interpret
ndviVegetation index (NDVI)DimensionlessGreenness and vegetation activity.
eviEnhanced Vegetation Index (EVI)DimensionlessAn additional measure of vegetation condition.
nbrNormalised Burn Ratio (NBR)DimensionlessSpectral response sensitive to vegetation cover condition and burning.
dnbrChange in Normalised Burn Ratio (dNBR)DimensionlessIn the global series: previous-month NBR minus current-month NBR. This does not prove that a fire occurred.
ndwiNormalised Difference Water Index (NDWI)DimensionlessSpectral indicator of open water and moist surfaces.
ndmiNormalised Difference Moisture Index (NDMI)DimensionlessA spectral measure of vegetation moisture.
laiLeaf Area Index (LAI)m²/m²Leaf area per unit of ground area.
faparFraction of absorbed light (fAPAR)DimensionlessFraction of absorbed photosynthetically active radiation.
sipiStructure Insensitive Pigment Index (SIPI)DimensionlessPigment spectral index; not a diagnosis of tree condition.
lst_daySurface temperatureKelvins (K)Daytime land surface temperature in kelvins, not air temperature. °C = K − 273.15.
soil_moistureSoil moisturem³/m³Volumetric soil moisture as defined by the source product; not a percentage unless converted.
disturbanceLand-cover disturbanceBit flagsA bitmask, not a probability. Do not interpret bits without the agreed layer dictionary.
burn_dateBurn historyDay of yearDay of year from the historical product. Not a real-time alert; null does not mean no fire.

units=1 means dimensionless. Do not transfer thresholds between indicators or territories without validation. Ranges and descriptions are available in spectral_indices.indices within the summary response. A layer’s presence in the cube does not guarantee an observation for every cell and month.

CCI and BAIS2 are used for detailed area analysis and are not included in this global public series. They may appear in the registry with public_ready=false. History and availability depend on the source; 319 months on the shared axis do not mean 319 measurements for every indicator.

05 / Reference responses

Reading summaries and prices

Reference-response dictionary
Endpoint / fieldContent
summary · cubeSize and state of the overall data environment. READY and readiness percentages do not establish the accuracy of every value.
summary · spectral_indicesVersion, groups, index specifications and availability flags. Querying an index requires public_ready=true.
summary · policy / map / l1_detailPublic-read limits, map overview and detail-level information.
summary · benchmark / external / heavy_layersAdditional technical context. Do not use it as stand-alone evidence of accuracy or an event.
summary · pricing_labels / pricing / paymentInterface reference information. Use tariff-catalog.api to select a data-exchange package.
validation · updated_at / snapshot_stateSource snapshot date and freshness: published, stale or date_unconfirmed.
validation · summary / components[]Component counters and cards: id, name, status, decision, scope, limitation, checks_total, checks_passed, warnings, failures, blocked.
validation · controls / limitations / evidence_levelsWhat is checked, the applicable limitations and the different evidence levels.
validation · external_reference_* / disclosure_policySources and disclosure boundaries; not a list of organisations that certified the platform.
tariff-catalog · version / currency / api[]Catalogue version, RUB and API packages. Each package contains id, name, features and price_mode; published prices include min and sometimes max.
tariff-catalog · oracle_optionOracle is an option within the main contract. Capabilities and pricing depend on the access level and agreed configuration.
tariff-catalog · portal / reports / institutional / modulesOther product groups do not automatically include API access. fx documents the conversion basis, not the current exchange rate.
pricing-settings · labels / quote_required_keysReady-to-display price labels and custom-quote keys. settings=null for a price does not mean free access.

Component status codes retain their machine-readable spelling. validated: checks passed within the stated scope; limited_operational_use: limited use; research_only: research use; model_estimate_only: model estimates; draft_scenarios_only: preliminary scenarios; not_approved_for_production: not approved for operational use; validation_failed: validation failed; not_started: no confirmed status. A missing counter must not be treated as zero.

Additional fields may be introduced. Clients should ignore unknown fields, verify required value types and not treat operational summaries as an immutable measurement contract. Exact schemas for the main responses are provided in the OpenAPI file.

06 / Reliable integration

Errors, limits and retries

HTTP and error handling
HTTP / codeMeaningAction
200 / ok=trueRead operation completedCheck actual months, null values, dates and fitness for use.
200 / ok=falseNo validation snapshot availableShow that the snapshot is unavailable; do not claim that checks passed.
304Snapshot unchanged under If-None-MatchUse the cached response body; 304 has no body.
400 / invalid_coordinatesCoordinates missing, non-numeric or out of rangeCheck lat in −90 to 90 and lon in −180 to 180.
400 / invalid_h3_r7Invalid identifier syntaxSupply one valid H3 resolution 7 identifier.
400 / unsupported_public_variablesUnknown, private or too many indicatorsCheck var and public_ready. Do not request CCI/BAIS2 through this endpoint.
400 / invalid_monthThe date is not in YYYY-MM format or falls outside the supported rangeCorrect the calendar year and month.
400 / invalid_month_rangeThe start of the period is later than the endSwap the period boundaries or choose a new range.
400 / public_range_exceeds_12_monthsPeriod exceeds 12 months inclusiveShorten the period; arrange commercial access for larger volumes.
404 / not_availableIndicator excluded from public accessChange the selected indicators; retrying unchanged will not help.
404 / not_foundUnknown routeCheck the base URL and exact path spelling.
503 / cube_read_unavailableSource could not be read; a missing cell or period in the cube may produce the same resultCheck parameters and coverage. Use bounded retries for temporary failures.
503 / h3_lookup_unavailableCell lookup temporarily unavailableDo not replace the response with zero; retry later or contact support.
Other 4xx / 5xxApplication, security or gateway errorCheck HTTP and Content-Type first; the body is not necessarily JSON.
Example parameter error
{"ok":false,"error":"invalid_coordinates"}
Volume

1 cell · up to 12 months

Up to 20 names in var; the currently available list is smaller. There is no public bulk export of the entire cube. Request-size limits are not requests-per-second quotas.

Load

Agree limits before integration

Public access does not promise unlimited request rates, response times or service levels. For commercial projects, we agree frequency, concurrency, volume and overage charges.

Caching

Follow the HTTP headers

summary and validation support ETag / If-None-Match and 304. At this edition: Cache-Control: public, max-age=15, stale-while-revalidate=30. Series and pricing responses use no-store.

Client recommendation: use a bounded timeout and up to three retries for temporary network failures or 502/503/504 responses, with increasing randomised delays. This is guidance, not a guaranteed server response time. Never retry indefinitely. If 429 includes Retry-After, respect it; no universal numeric quota or header scheme is currently claimed.

Responses use application/json; charset=utf-8. Large summaries support gzip when Accept-Encoding: gzip is supplied. Do not depend on key order. For calls from another website, use your own server-side intermediary: actual GET responses currently lack an allowing CORS header. An OPTIONS response alone does not permit cross-origin reading.

This public specification includes no API key, Bearer token, OAuth login, callbacks to client URLs or guaranteed set of private endpoints. Commercial authentication, address restrictions, logging, retention and service levels are agreed separately. Never place secrets in request URLs or browser code.

07 / Working examples

Connect from Python and JavaScript

Minimal server-side read examples preserve missing values and the correct axis order. Add rate control, logging and bounded retries for a production application.

Python · standard library
Python 3
import json
from urllib.parse import urlencode
from urllib.request import Request, urlopen
from urllib.error import HTTPError, URLError

params = {"lat": 55.75, "lon": 37.62, "var": "ndvi",
          "start": "2026-06", "end": "2026-07"}
url = "https://forestosglobal.com/api/public/h3-at-point?" + urlencode(params)
try:
    with urlopen(Request(url, headers={"Accept": "application/json"}), timeout=30) as r:
        if r.headers.get_content_type() != "application/json":
            raise RuntimeError("Expected JSON")
        data = json.load(r)
except HTTPError as error:
    raise RuntimeError(f"HTTP error: {error.code}") from error
except URLError as error:
    raise RuntimeError("Connection error") from error
if not data.get("ok"):
    raise RuntimeError(data.get("error", "Data unavailable"))
for index, month in enumerate(data["months"]):
    value = data["variables"]["ndvi"][index][0]
    print(month, "no data" if value is None else value)
# If months is empty, there are no observations in the requested interval.
# Values are already decoded; do not apply scale again.
JavaScript · server-side Node.js
Node.js 18+
// Node.js 18+; server-side example; this does not bypass CORS restrictions.
const url = new URL("https://forestosglobal.com/api/public/h3-at-point");
url.search = new URLSearchParams({
  lat: "55.75", lon: "37.62", var: "ndvi",
  start: "2026-06", end: "2026-07"
}).toString();
const response = await fetch(url, {
  headers: { Accept: "application/json" },
  signal: AbortSignal.timeout(30000)
});
if (!response.ok) throw new Error(`HTTP error: ${response.status}`);
if (!(response.headers.get("content-type") || "").includes("application/json")) {
  throw new Error("Expected JSON");
}
const data = await response.json();
if (!data.ok) throw new Error(data.error || "Data unavailable");
data.months.forEach((month, index) => {
  const value = data.variables.ndvi[index][0];
  console.log(month, value === null ? "no data" : value);
});
08 / Pricing in rubles

From exploration to your scale

Public read access

Explore the format

The six public GET endpoints above are available without payment or an API key. This is not a commercial licence for bulk export or data resale.

Free of charge
For an initial integration

Developer

from 434 982 ₽
per year · final price agreed in advance

Derived indices and an agreed feature set for integration. Test access, quotas and usage terms are specified before onboarding.

For a live service

For services and applications

Custom quote
Volume, schedule and terms tailored to the task

Higher volumes and agreed commercial use. Event coverage, notifications and available forecast outputs are checked for each project.

For enterprise systems

Enterprise data exchange

Custom quote
Volume, schedule and terms tailored to the task

Batch calculations, dedicated quotas and contractual service levels. Redistribution rights are agreed separately.

Prices come from the platform’s shared catalogue. The annual contract defines volumes, territories, outputs and usage rights. Implementation, overage and additional work are quoted separately unless included. The final amount and tax terms are confirmed in the proposal.

Public and commercial access compared
TermPublic read accessCommercial access
AccessNo key required; overview data contractAgreed specification and authentication method
VolumeOne cell, up to 12 months per requestAreas, periods, frequency and concurrency under contract
OutputsDocumented series and public reference dataAgreed set of derived outputs
ServiceNo numerical service-level commitmentAgreed targets, support and incident handling
OraclePublic exploration through the platform interfaceAn option within the main contract; capabilities depend on access level
Source cube and modelsFull export is not providedNot included automatically; output rights are agreed separately

Oracle is not sold here as a separate plan. Public, client, operational and strategic access differ in their data, tools and permissions. API access to the option is included only if explicitly agreed in the main contract; buying an API package does not automatically unlock every access level.

09 / Onboarding and support

Agree the deliverables before work starts

01

Describe your task

Areas, periods, required indicators, your system’s format, expected request frequency and output rights.

02

Agree the contract

Endpoints, authentication, limits, acceptance criteria, service levels, pricing and change management.

03

Test the integration

Reference requests and responses, missing-value and error handling, followed by acceptance of the agreed configuration.

For support, include the endpoint, request time in UTC, HTTP status, error code and a minimal example without secrets. Send sensitive coordinates and client data only through an agreed secure channel. A universal request_id is not currently guaranteed in responses.

10 / Glossary

Technical terms explained

API terminology
TermMeaning
APIApplication programming interface: a way for another program to retrieve platform data.
HTTP / HTTPSNetwork communication protocols; HTTPS uses an encrypted connection.
GETA read method: a program requests data at a specified URL. The method name is not translated in requests.
JSON / UTF-8A structured data format and a text encoding supporting multiple languages, including Russian.
OpenAPI / JSON SchemaStandards for describing endpoints, parameters and response structures for automated validation.
URL / URIA resource address or identifier. Copy it unchanged from the examples.
UTCCoordinated Universal Time, not Moscow time or the browser’s local time.
H3 resolution 7 / WGS 84A spatial cell system and a geographic coordinate reference system. Grid resolution is not the accuracy of the original measurement.
null / true / falseIn the data: a missing value, Boolean true and Boolean false. Their code spelling remains unchanged.
CORSBrowser rules for reading data from another origin. Server-side access does not imply browser access is allowed.
ETag / If-None-MatchA response version tag and a conditional request using that tag, allowing clients to detect changes.
Cache-Control / Retry-After / gzipInstructions for caching a response, waiting before retries, and a data compression format.
Bearer / OAuthAn access-token scheme and an authorisation protocol. Neither is required for the documented public endpoints.
WebhookAn automatic notification sent to an agreed URL in another system.
MRV / SLA / AOIMonitoring, reporting and verification; service-level agreement; area of interest. These concepts are explained in the interface.
P10—P90The range between the 10th and 90th percentiles of model estimates. It does not by itself guarantee that a future observation will fall within the range.
Validation / data provenanceChecks of result quality and applicability; information about where data came from and how they were obtained.

The Forest Os Global name, URLs, commands, field names, scientific indices and format names retain their original spelling for compatibility. Their purpose is explained in this glossary.

11 / Version and disclosure

A verifiable specification without private internals

API documentation edition 2.1.1, dated 03 September 2026: labels and the technical glossary were expanded without changing URLs, parameters or data structures. The document edition does not change a model version or its operational approval. Public endpoint URLs currently contain no version number.

The description uses OpenAPI 3.1.1 and JSON Schema 2020-12. HTTP methods and statuses are explained with reference to HTTP Semantics (RFC 9110). This is a machine-readable specification, not a claim of certification, software-register inclusion or compliance with every possible standard.

We publish purpose, parameters, formats, limits and access terms. We do not disclose keys, internal addresses, infrastructure diagrams, client data or private model parameters. Interface changes and migration for commercial projects are governed by their specifications. New capabilities are not included until the agreed delivery scope confirms them.

Forest Os Global / Public information

What has been validated.
Where limitations remain.

Dated validation results for Forest Os Global. Data integrity, model quality and permitted use are reported separately, without a single score that hides differences between components.

Edition 2.0 · 03 September 2026 · Validation statuses have their own dates

Published snapshot date (UTC): 2026-09-02T09:38:00Z. This is not the date of the latest forest event. Refresh the snapshot to confirm that the status is current.

01 / Data

Scope of engineering validation

25datasets checked in the snapshot
17verified data manifests
36external references and specifications considered

These are validation counters, not a count of independent auditors or a forecast accuracy percentage. An external specification establishes the meaning of a source product; it does not by itself validate a Forest Os Global model.

Status of checked datasets
Dataset validation resultCount
No issues6
With warnings19
With errors0
02 / Applicability

Status of each component

Data and cube

Validated within the stated scope

The stated checks passed; the conclusion applies only to the specified validation scope.

Passed
13
Total in summary
13
Warnings
0
Failures
0
Unconfirmed
0
What was checked and what limitations apply

Integrity, spatial keys, dates and absence of duplicates.

Engineering data checks do not establish the accuracy of every analytical model.

F-12 · Active fire detection

Not approved for operational use

Quality checks have not passed. No validated operational forecast is claimed for this model.

Passed
20
Total in summary
35
Warnings
0
Failures
15
Unconfirmed
3
What was checked and what limitations apply

Detection timeliness, missed events and false alarms.

Operational release requires successful validation on a new independent period. Earlier detection is not guaranteed.

F-09 · Fire spread

Research use only

The result is not used as a validated operational forecast.

Passed
15
Total in summary
18
Warnings
0
Failures
0
Unconfirmed
3
What was checked and what limitations apply

Comparison of forecasts saved in advance with actual fire perimeters.

Applicability to a territory requires independent time-stamped perimeters and reliable initial conditions.

Forest inventory and planning

Model estimates only

Only explicitly labelled model estimates are permitted; they do not replace field surveys.

Passed
18
Total in summary
25
Warnings
3
Failures
0
Unconfirmed
4
What was checked and what limitations apply

Consistency of model estimates with available reference materials.

Regional aggregates do not replace a field inventory of a specific area.

Forest restoration

Preliminary scenarios only

Scenarios require specialist review and agreement on their initial assumptions.

Passed
18
Total in summary
26
Warnings
2
Failures
0
Unconfirmed
6
What was checked and what limitations apply

Scenario justification, input data and specialist review.

Survival probabilities require observations of actual project outcomes.

Oracle · Four access levels

Limited use

Use is permitted subject to the stated limitations and specialist review.

Passed
39
Total in summary
49
Warnings
1
Failures
0
Unconfirmed
9
What was checked and what limitations apply

Approval handling, data separation and human review of significant decisions.

Testing software safeguards does not replace blind expert assessment of recommendations and observation of their consequences.

Core and data updates

Validated within the stated scope

The stated checks passed; the conclusion applies only to the specified validation scope.

Passed
12
Total in summary
12
Warnings
0
Failures
0
Unconfirmed
0
What was checked and what limitations apply

Reproducible updates, recovery and preservation of results.

Engineering resilience does not guarantee availability under any workload.

F-07 · Fire fuel index

Limited use

Use is permitted subject to the stated limitations and specialist review.

Passed
16
Total in summary
19
Warnings
1
Failures
0
Unconfirmed
2
What was checked and what limitations apply

Consistency of the analytical indicator and quality of input data.

The index is not a standalone fire probability; its relationship to field measurements requires separate validation.

F-14 · Event reports

Limited use

Use is permitted subject to the stated limitations and specialist review.

Passed
15
Total in summary
18
Warnings
1
Failures
0
Unconfirmed
2
What was checked and what limitations apply

Quality of permitted sources, duplicate reports and operator confirmation.

A report remains evidence to investigate, not an established fact.

F-13 · Fire type

Not approved for operational use

Quality checks have not passed. No validated operational forecast is claimed for this model.

Passed
15
Total in summary
19
Warnings
2
Failures
1
Unconfirmed
1
What was checked and what limitations apply

Classification quality and robustness across classes.

Fire type and ignition cause are different questions. Establishing the cause requires separate evidence.

F-15 · Fire risk forecast

Not approved for operational use

Quality checks have not passed. No validated operational forecast is claimed for this model.

Passed
17
Total in summary
27
Warnings
0
Failures
7
Unconfirmed
3
What was checked and what limitations apply

Accuracy, false alarms, calibration and transferability across territories.

A favourable result on an internal sample does not replace successful validation on new regions and periods.

Counters follow the source snapshot’s categories. Do not add them up to estimate the accuracy or readiness of the entire platform. A release decision requires new validation; redesigning the page does not change model approval.

03 / Evidence levels

External data and independent assessment

Engineering checks

Does the process work correctly?

Data integrity, reproducibility, access controls and missing-value handling.

Comparison

Is the result supported by observations?

Comparison with historical, satellite, event and field references under specified conditions.

Independent organisation

Can the assessment be reproduced?

A separate protocol and conclusion from external experts. This cannot be claimed solely because NASA data or government aggregates were used.

04 / Limitations

Considerations for use

  • The status applies to a dated snapshot, not every future version.
  • The number of checks is not a forecast accuracy percentage. Source-summary categories should not be added without their methodological definitions.
  • Validation against external data does not mean an independent organisation has certified the platform.
  • A global grid and time axis do not guarantee complete coverage of every layer.
  • Field accuracy and model transferability require separate field and cross-region validation.
  • Recommendations do not replace the decisions of an authorised specialist.
05 / Joint validation

Open to independent validation

We can agree areas, authorised field data, periods, metrics and a reproducible protocol with a research or specialist organisation. Access terms and any research quota are agreed separately; a positive result is not a condition of participation.

06 / Disclosure policy

Share evidence. Protect confidential information.

Public

Results and limitations

Sources, indicator definitions, dates, aggregated validation results and scope of use.

By agreement

Project validation package

Specifications, methodology, anonymised examples and additional evidence within an agreed scope.

Private

Intellectual property and data

Source code, trained weights, internal features and thresholds, infrastructure diagrams, keys and other customers’ materials.

Forest Os Global / Public information

Forest territories.
One digital environment.

Forest Os Global brings together satellite observations, analytics and tools for managing forest territories. For organisations that need to see what changed and understand the evidence behind each finding.

Edition 2.0 · 03 September 2026 · Validation statuses have their own dates

01 / Scale and meaning

Global history. Local validation.

25 063 102cells in the global land grid
319 monthstime axis: January 2000 – July 2026
9 + 2public indices and two for detailed area analysis

These figures describe the analytical infrastructure at the edition date, not a count of confirmed observations. Completeness depends on the layer, territory, season and satellite operating period.

02 / Workflow

From observations to informed decisions

01

Observe

Compare vegetation history, thermal anomalies, moisture and land-cover changes.

02

Validate

Distinguish observations, derived estimates and scenarios; see the source and limitations of each result.

03

Act

Prepare reports, prioritise survey areas and agree decisions with specialists.

03 / Capabilities

One platform. Different tasks.

Monitoring

Territory condition

Maps and change histories for areas, portfolios and regions. Detail depends on available data.

Fires

Events and risk

Current satellite signals are separated from historical burned-area records. Forecast-model approval is shown on the validation page.

Forest inventory and planning

Assessment and planning

Model estimates, forest-use scenarios and portfolio analysis. These do not replace field forest inventories.

Restoration

Restoration scenarios

Prepare and compare restoration options, with mandatory specialist review.

Oracle

Assistance with data analysis

An option within the main plan: public, client, operational and strategic access, with different permissions and tools.

Science

Comparison with observations

Work with field time series and satellite history under an agreed project. Independent validation is arranged separately.

04 / Who it is for

Different roles. Shared data.

Organisational tasks and deliverables
UserPractical taskDeliverable
Forest operatorMonitor your territories and plan surveysReport, change map and survey priorities
Public institutionCompare available information about a territoryOverview and analytical materials for domain specialists
IntegratorEmbed analytics in your information systemAgreed data exchange and service terms
Research institutionTest hypotheses using field and satellite dataReproducible comparison protocol and assessment of limitations
05 / Verifiability

Trust starts with clear boundaries of applicability

Software reliability, scientific model quality and independent assessment are different evidence levels. We report them separately and do not present the number of passed checks as forecast accuracy.

Forest Os Global / Public information

Everything for an informed
decision.

Documents for customers, integrators, research institutions and domain specialists. Service terms, methodology and verifiable information about the platform.

Edition 2.0 · 03 September 2026 · Validation statuses have their own dates

01 / Library

Documents for your task

Methodology · edition 2.0

How we reach our findings

Sources, spatial scale, indicators, validation and limitations. The browser version corresponds to the downloadable Russian document.

Validation · see snapshot date

Validation status

Data and component checks, unresolved issues and permitted-use boundaries.

For integrators · interface documentation

Public API

Query parameters, limits, formats and errors for the basic read endpoints. Private interfaces are not included in this file.

For customers

Order and terms of service

The composition, price, terms and procedure for obtaining the result are determined by the selected service and agreed conditions.

02 / Working terms

Commercial and legal documents

Conditions of service
DocumentWhat to check
OfferProcedure for concluding the contract, subject and conditions of service
Payment and provision of servicesHow the order is made and the result transmitted
Refunds and cancellationHow to request cancellation and a refund
Personal dataContact data processing and applicant’s rights
03 / Government and corporate projects

Integration under an agreed brief

Preparing materials for an external information system does not mean access to it has been authorised. FGIS LK requires coordination with its operator, granted access and verification of applicable official formats. Until then, Forest Os Global does not replace state forest register data.

Technical specifications and integration assessment materials are provided to identified recipients after the disclosure scope is agreed. Internal connection diagrams, addresses and operational instructions are not published on this page.

04 / Supplier assessment

Technical assessment materials

Within the framework of the project, it is possible to agree on a demonstration, delivery description, results rights, service indicators and acceptance protocol. Official statuses and opinions are published only if there is appropriate confirmation. The presence of this documentation in itself does not mean inclusion in the register of Russian software or certification of the platform.

05 / Transparency

Open information and protected materials

Public

Results and limitations

Sources, indicator definitions, dates, aggregated validation results and scope of use.

By agreement

Project validation package

Specifications, methodology, anonymised examples and additional evidence within an agreed scope.

Private

Intellectual property and data

Source code, trained weights, internal features and thresholds, infrastructure diagrams, keys and other customers’ materials.

Edition 2.0 aligns the time axis, nine public indices, two detailed-area indicators and four Oracle access levels. Validation is linked to a published snapshot; internal test results are not converted into marketing promises.

Forest Os Global / Public information

How data becomes
insight.

Forest Os Global public methodology: sources, scale, indicator definitions and rules for evidence-based validation. Applicability matters more than universal promises.

Edition 2.0 · 03 September 2026 · Validation statuses have their own dates

01 / Scope

What this methodology explains

Forest Os Global links observations, derived indicators and analytical findings to a place and time. This methodology describes how results are produced and checked; it does not establish the accuracy of every model.

This edition describes the public and contractual analytical environment. Source code, trained parameters and infrastructure internals are not disclosed. Standard scientific definitions and source specifications are available through the links.

02 / Space and time

A global grid does not replace local detail

At the edition date, the global land grid contains 25 063 102 H3 resolution 7 cells. Its time axis contains 319 months, January 2000 – July 2026. This is a structure for storing and comparing data, not a guarantee of an observation for every indicator, cell and month.

The average area of an H3 resolution 7 hexagonal cell is about 5.16 km²; actual area varies across the Earth. Grid resolution does not improve the resolution of the satellite product. Small areas require detailed imagery and, where necessary, field measurements.

Some source time series start later. Clouds, snow and a lack of suitable scenes create missing values. Page date, validation snapshot date and observation date are different quantities.

03 / Sources and quality

Preserve data provenance

Sources include NASA MODIS, Copernicus Sentinel-2 and USGS Landsat satellite products, NASA FIRMS thermal anomalies, ERA5-Land climate reanalysis and available authorised field materials. Products differ in spatial and temporal resolution.

Source quality flags determine which observations are usable. Depending on the product, observations form a monthly series or retain an event date. Aggregation across sources and rescaling must be defined for each result.

Reproducing a result requires the area boundary, period, source product and version, units, aggregation method, completeness and limitations. The scope of supplied provenance documentation is agreed in the brief.

04 / Indicators

Nine public indices and detailed-area analysis

The public registry includes NDVI, EVI, NBR, dNBR, NDWI, NDMI, LAI, fAPAR and SIPI. CCI and BAIS2 support detailed analysis of a selected area where suitable Sentinel-2 data are available; they are not claimed as continuous global historical series.

NDVI and EVI describe vegetation spectral response; NDMI is moisture-sensitive; the selected NDWI definition measures water-related spectral contrast. LAI describes leaf area, and fAPAR describes absorption of photosynthetically active radiation. These indicators are not direct forest inventory measurements, disease diagnoses or carbon stock measurements.

NBR and dNBR support spectral change analysis. The difference between adjacent monthly values is not equivalent to the severity of a specific fire: that requires suitable before-and-after windows, seasonality checks and independent evidence.

Different sensors and approximated derived indicators are not assumed interchangeable without validation. The shortwave infrared band used for NBR is particularly important. An index name alone does not establish comparability between two series.

05 / Interpretation

Observations, estimates and scenarios are different outputs

A missing observation remains missing, not zero. Zero is valid only as defined by the source product. Failure to detect a signal does not prove that a small or missed event did not occur.

A FIRMS hotspot is a satellite detection, not an established ignition cause. Burned-area archives, hotspot counts, fire area and the number of separate incidents are not interchangeable.

Fire type, degree of land-cover change and ignition cause are separate questions. Causal hypotheses require event, field and other independent evidence; the system does not make legal determinations.

Climate scenario context uses available NASA NEX-GDDP-CMIP6 regional data, including SSP2-4.5 and SSP5-8.5. These are pathways under specified assumptions, not site-level fire probabilities or a proprietary dynamic climate model from Forest Os Global.

06 / Validation

Three levels of evidence

Engineering checks establish integrity, consistency and reproducibility of data and software processes. Backtesting compares a result with an observed outcome. External validation uses a reference not used for tuning; it is not equivalent to an independent organisation’s formal assessment.

A valid test specifies the task, forecast issue date, territory, horizon, observed-outcome source, metrics and acceptance criteria in advance. Samples are separated by time and, where appropriate, geography. Future information must not leak into input features.

Event evaluations measure detections, missed events and false alarms. Probability evaluations assess calibration and ranking; quantitative estimates assess error magnitude. Metrics are published with their validation scope, sample and limitations, not as a universal platform accuracy percentage.

If the actual outcome is not yet observed, a reference is missing or the sample is insufficient, the result remains unconfirmed. Technical readiness does not turn a failed test into a pass. Current component decisions appear in a separate dated validation summary.

07 / Model improvement

Feedback without automatic release approval

Events from authorised sources and user reports can expand the reference dataset for forecast comparison. User reports first require verification, deduplication and confirmation of event time and location.

A new model version must be compared with its predecessor using independent data. Training or accumulating observations does not by itself authorise publication of a result. This section describes how improvements are verified, not a promise of automatic gains.

Oracle operates within its access level and permitted data. It helps explain results and compare options, but significant actions require specialist confirmation.

08 / Limits of use

What must be agreed with the customer

Forest inventory characteristics remain model estimates until supported by appropriate field evidence. Restoration scenarios do not guarantee survival. Carbon and economic estimates require their own methodology, assumptions and validation.

Access to FGIS LK and other restricted sources requires lawful authorisation. The system does not replace official documents, state information or decisions by authorised persons.

Sources, definitions, validation scope, aggregated results and limitations are disclosed publicly. Additional materials for reproducing a study are supplied within an agreed scope, respecting data and intellectual property rights.

09 / Primary sources

Official specifications

NASA MODIS

Vegetation history, land surface temperature and burned-area archives. Each product has its own dates, resolution and quality flags.

Open source documentation

NASA FIRMS

Current satellite thermal anomalies. A hotspot is not equivalent to a separate fire, its area or an established cause.

Open source documentation

For comparisons of fire-related indices, use the USGS definition of NBR, including the correct band mapping for the sensor.

USGS: definition of NBR
Payment for Forest OS

Checking payment

Returning from the payment form does not confirm payment. Forest OS is waiting for a verified server notification from Robokassa.

Order status

Retrieving order status…

Order
Service
Amount
Receipt
Report payment

Payment incomplete

Money should not be considered received without server confirmation. The order can be reopened after the status check.

The payment has been cancelled or the payment form has returned an error.