# Capabilities & Datasets

## :star:Get started

<figure><img src="https://213166889-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FpN6g4jM2ghngm26l7cbt%2Fuploads%2FJpHJsCczSKy943ohHZZ5%2FUPDATE%20APRIL%202026.png?alt=media&#x26;token=154a2da1-feaf-4575-ab1a-a3eeac87d0d9" alt=""><figcaption></figcaption></figure>

<figure><img src="https://213166889-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FpN6g4jM2ghngm26l7cbt%2Fuploads%2F2JSNC6Xs0df5VZh2uJ9q%2FUPDATE%20APRIL%202026%20V2.png?alt=media&#x26;token=151ca18f-f0b9-4d40-a3ee-2a34addf12d8" alt=""><figcaption></figcaption></figure>

## :clipboard:Overview

Global Nature Watch offers a set of core capabilities designed to help you explore, analyze and make sense of geospatial data to help monitor nature. &#x20;

&#x20;With Global Nature Watch, you can ask questions in plain language to:&#x20;

* Discover and learn about peer-reviewed data from Land & Carbon Lab and Global Forest Watch&#x20;
* Analyze data for your areas of interest and get insights to help make more sense of the data &#x20;
* Get answers that are tailored to your context as a starting point for reports, policies or project decisions&#x20;

&#x20;When you ask a question or provide a prompt (see [prompt-guidance](https://help.globalnaturewatch.org/resources/prompt-guidance "mention")), Global Nature Watch's AI assistant processes your intent using a system of agents (see [models](https://help.globalnaturewatch.org/get-started/models "mention")) to understand your request and set the context — data layer, area, date — for the response.

As you explore, you can save both individual insights and full conversations on the platform. Your saved areas, layers and conversations stay organized so you can return to earlier work, build a portfolio of places and questions and pick up right where you left off. &#x20;

Because Global Nature Watch is in preview, its capabilities will continue to evolve. It is designed to be open, collaborative and transparent so anyone can build on it. Over time, you can expect new features to be added, existing ones to improve and some to be retired as we learn more about how best to help you work with nature data.&#x20;

## :information\_source:How Global Nature Watch Processes Your Request&#x20;

Global Nature Watch combines a conversational AI interface with structured geospatial data systems. While you interact with Global Nature Watch in natural language, your request is processed through a deterministic[^1], step-by-step analytical workflow to ensure scientific accuracy and reproducibility.&#x20;

At a high level, Global Nature Watch uses:&#x20;

* A large language model (LLM) to interpret your request and determine intent&#x20;
* Metadata-aware retrieval (Retrieval-Augmented Generation, or RAG) to understand available datasets and boundaries&#x20;
* Deterministic data services to fetch and aggregate verified geospatial data&#x20;

Once a prompt is entered, Global Nature Watch follows a sequential workflow.&#x20;

{% hint style="info" %}

#### **How AI and Deterministic Systems Work Together**&#x20;

Global Nature Watch combines two types of systems:&#x20;

**AI (Language Models)**&#x20;

* Understand your question&#x20;
* Select datasets and parameters&#x20;
* Explain trends and suggest follow-up questions&#x20;

**Deterministic Data Services**&#x20;

* Retrieve peer-reviewed data from reliable sources (e.g., Global Forest Watch and Land & Carbon Lab)&#x20;
* Perform geospatial calculations, drawing on the Data API and using [CodeAct ](#user-content-fn-2)[^2]
* Generate charts and visual outputs using CodeAct&#x20;
* Return the same results every time for the same inputs&#x20;

&#x20;AI helps interpret and explain. Deterministic systems perform the scientific analysis.&#x20;
{% endhint %}

#### Step 1: Understanding Your Request&#x20;

Global Nature Watch's language model reads your question and identifies three key components:&#x20;

* Where — the geographic area you're asking about&#x20;
* What — the type of environmental data you need&#x20;
* When — the time period of interest&#x20;

&#x20;If any of these are missing or unclear, Global Nature Watch will ask you to clarify before proceeding.&#x20;

#### Step 2: Identify the Location (Where)&#x20;

Global Nature Watch first determines the specific geographic area you are asking about.&#x20;

* It matches your place name to a predefined Area of Interest (AOI) in its spatial database&#x20;
* If a place name is ambiguous (for example, multiple places share the same name), Global Nature Watch will ask you to clarify&#x20;
* Global Nature Watch analyzes one defined area at a time&#x20;

No analysis begins until a specific location is confirmed.&#x20;

*AI interprets your place name. A rule-based system matches it to a predefined boundary.*&#x20;

#### Step 3: Dataset & Time Selection (What and When)&#x20;

Next, Global Nature Watch determines which dataset best answers your question.&#x20;

* It interprets keywords such as “fire,” “deforestation” or “land cover change”&#x20;
* It selects the appropriate data layer&#x20;
* It confirms or requests a time range (for example, “2018–2023” or “last three months”)&#x20;

If the time period is missing or unclear, Global Nature Watch will ask for clarification before proceeding.&#x20;

*AI selects the relevant dataset and parameters based on your request.*&#x20;

#### Step 4: Data Retrieval (Deterministic Query)&#x20;

Once the location, dataset and time range are confirmed, Global Nature Watch retrieves the relevant data from its underlying data sources.&#x20;

The data returned reflects:&#x20;

* The selected geographic boundary&#x20;
* The specified time range&#x20;
* The chosen dataset&#x20;

If any required information is missing, this step does not proceed.&#x20;

*This step is fully deterministic: data is retrieved directly from structured geospatial services.*&#x20;

#### Step 5: Analysis & Insight Generation&#x20;

Global Nature Watch then analyzes the retrieved data to produce user-friendly outputs. This includes:&#x20;

* Creating a chart or simple visualization&#x20;
* Summarizing key trends in plain language&#x20;
* Suggesting follow-up questions for deeper exploration&#x20;
* Providing access to the underlying code used to produce visualizations &#x20;

The goal is to make complex environmental data understandable and actionable without requiring technical expertise.&#x20;

*Deterministic systems provide underlying value for charts. AI: generates executable code when creating charts (rather than manipulating data directly) using CodeAct and explains and summarizes the results.* &#x20;

#### Status Updates & Transparency&#x20;

Throughout the process, Global Nature Watch may provide short updates (for example, confirming the selected location or dataset). These updates are intended to keep you informed and ensure accuracy before analysis continues.

## :gear:Capabilities & Limitations&#x20;

This section summarizes what Global Nature Watch does and does not support across locations, datasets, analysis types and outputs.&#x20;

### :round\_pushpin:Areas of Interest (AOI)&#x20;

Global Nature Watch can analyze data only for predefined geographic boundaries. The table below outlines which types of Areas of Interest (AOIs) are supported and which are not.&#x20;

<table data-header-hidden="false" data-header-sticky><thead><tr><th width="297">Capability</th><th width="123">Supported</th><th>Notes</th></tr></thead><tbody><tr><td>Countries (admin 0) </td><td>✅ </td><td>Predefined boundaries only </td></tr><tr><td>States / Provinces (admin 1) </td><td>✅ </td><td>Varies by country </td></tr><tr><td>Districts / Municipalities (admin 2) </td><td>✅ </td><td>Where available </td></tr><tr><td>Protected Areas (WDPA) </td><td>✅ </td><td>Limited to groups of &#x3C;25 WDPA areas </td></tr><tr><td>Key Biodiversity Areas (KBA) </td><td>✅ </td><td>Limited to groups &#x3C;25 KBA areas </td></tr><tr><td>Indigenous &#x26; Community Lands </td><td>✅ </td><td>Limited to groups of &#x3C;25 areas  </td></tr><tr><td>Custom Polygons </td><td>✅ </td><td>Can be uploaded (GeoJSON, up to 1 MB) or drawn on the map </td></tr><tr><td>Global / Continental Areas </td><td>❌ </td><td>e.g., Africa, Europe </td></tr><tr><td>Ecosystems or Regions </td><td>❌ </td><td>e.g., Amazon Basin, Latin America </td></tr><tr><td>Buffers </td><td>❌ </td><td>e.g., 10km around a defined area </td></tr></tbody></table>

### :map:Datasets&#x20;

Global Nature Watch draws on a trusted set of geospatial datasets from world-leading researchers to provide timely and actionable insights on land use, forest change and nature-based solutions. These datasets include deforestation alerts, historical tree cover loss, land cover maps and more.&#x20;

We are committed to transparency in how data is sourced and used. Most of the data powering Global Nature Watch originates from or is also available on [Land & Carbon Lab](https://landcarbonlab.org/data/) and [Global Forest Watch](https://globalforestwatch.org/), where you can explore the full range of methodologies, data providers and technical documentation. All data made accessible through Global Nature Watch has been peer reviewed.&#x20;

The tables below summarizes the data themes currently supported and their general characteristics.

#### Ecosystem Monitoring

| Dataset                            | Description                                                        | Notes                                                                   |
| ---------------------------------- | ------------------------------------------------------------------ | ----------------------------------------------------------------------- |
| Tree Cover                         | 2000 baseline, 30m resolution, global \| Source: UMD               | data can be filtered by % canopy cover, Forest Carbon Stocks            |
| Tree Cover Gain                    | 2000-2020, 5-year intervals, 30m resolution, global \| Source: UMD | data can be filtered by % canopy cover, Forest Carbon Removals          |
| Tree Cover Loss                    | 2001-2024, annual, 30m resolution, global \| Source: UMD           | data can be filtered by % canopy cover, Forest Greenhouse Gas Emissions |
| Tree Cover Loss by Dominant Driver | 2001-2024, \~1km resolution \| Source: WRI/Google DeepMind         | data can be filtered by % canopy cover, Forest Greenhouse Gas Emissions |
| SBTN Natural Lands Map             | 2020, 30m resolution \| Source: SBTN/LCL/WWF/Systemiq              |                                                                         |

#### Land Cover Change Monitoring

| Dataset                                              | Description                                                         | Notes                                                                                                                                                                                         |
| ---------------------------------------------------- | ------------------------------------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| Global All Ecosystem Disturbance Alerts (DIST-ALERT) | 2023-present, updated weekly, 30m resolution, global \| Source: UMD | alerts can be filtered by any one of the following datasets at a time: drivers (LDACS), SNTB Natural Lands Map, Global Natural/Semi-Natural Grassland Extent (2022), Global Land Cover (2024) |
| Global Land Cover                                    | 2015-2024, annual, 30m resolution \| Source: UMD                    | analysis of land cover change is only available for 2015 to 2024 (not annually), and land cover analysis is only available for 2024                                                           |
| Land Disturbance Alert Classification System (LDACS) | 2024-present, updated quarterly, 30m resolution \| Source: WRI/UMD  |                                                                                                                                                                                               |

#### GHG Monitoring

| Dataset                         | Description                                                   | Notes                                                                           |
| ------------------------------- | ------------------------------------------------------------- | ------------------------------------------------------------------------------- |
| Forest Carbon Removals          | average over 2001-2024, 30m resolution \| Source: WRI         | can be analyzed but isn't available to view on the map as a separate data layer |
| Forest Greenhouse Gas Emissions | 2001-2024, annual, 30m resolution \| Source: WRI              | can be analyzed but isn't available to view on the map as a separate data layer |
| Forest Greenhouse Gas Net Flux  | average over 2001-2024, annual, 30m resolution \| Source: WRI |                                                                                 |

#### Agricultural Monitoring

| Dataset                                                    | Description                                                                  | Notes                                                                                                                                  |
| ---------------------------------------------------------- | ---------------------------------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------- |
| Deforestation (sLUC) Emission Factors by Agricultural Crop | <p>2020–2024, tabular data at national/subnational levels<br>Source: LCL</p> | <p>Covers 42 agricultural crops. Only available for administrative boundaries (not KBAs, protected areas, etc.)<br>\*Analysis only</p> |
| Global Natural/Semi-Natural Grassland Extent               | 2000-2024, annual, 30m resolution \| Source: LCL/OpenGeoHub/LAPIG/IIASA/iDiv | data only available up to 2022                                                                                                         |

### :chart\_with\_upwards\_trend:Analyses&#x20;

GNW performs structured, descriptive analyses based on the selected location, dataset and time range. The table below summarizes the types of analyses supported.&#x20;

<table data-header-hidden="false" data-header-sticky><thead><tr><th>Analysis Type</th><th width="187">Description</th><th width="211">Example</th><th width="93">Status</th><th>Notes</th></tr></thead><tbody><tr><td>Area-based summaries </td><td>Straightforward query targeting one area or dataset </td><td>“How much area was affected by conversion alerts in Peru?” </td><td>✅ </td><td> </td></tr><tr><td>Temporal analysis </td><td>Refers to specific years or date ranges </td><td>“How much forest was lost in Brazil between 2001 and 2020?” </td><td>✅ </td><td> </td></tr><tr><td>Trend analysis over time </td><td>Analyzes trends based on data analysis over time </td><td>“Is deforestation going up or down in Brazil?” </td><td>✅ </td><td>Within selected AOI </td></tr><tr><td>Before / after comparison </td><td>Analyzes data before and after a specified time </td><td>“Compare deforestation in Kalimantan before and after the start of Nusantara in 2022.” </td><td>✅ </td><td>Requires defined time range </td></tr><tr><td>Cross-location comparison </td><td>Compares two or more named areas </td><td>“Which had more cropland in 2015, Nigeria or Ghana?” </td><td>✅ </td><td> </td></tr><tr><td>Nested-area comparison </td><td>Compares sub-areas within a larger one </td><td>“Which state in India lost the most forest cover?” </td><td>✅ </td><td> </td></tr><tr><td>Dataset class comparison </td><td>Compares classes of data in a given location or analyzes a specific class in the data </td><td>“What percentage of disturbances in Occitanie were due to crop management in 2023?” </td><td><p>✅ </p><p>  </p></td><td> </td></tr><tr><td>Ranking </td><td>Ranks subareas within a larger area </td><td>“What states in Brazil had the most tree cover loss in 2024?” </td><td><p>✅ </p><p> </p></td><td>Slow, and limited to 1,000 areas </td></tr><tr><td>Cross-country analysis </td><td>Ranks countries globally </td><td>“Where is agricultural land found in the greatest area globally?” </td><td>✅ </td><td>Up to 1,000 areas  </td></tr><tr><td>Predictive modeling </td><td> </td><td> </td><td>❌ </td><td>Not supported </td></tr><tr><td>Forecasting </td><td> </td><td> </td><td>❌ </td><td>Not supported </td></tr></tbody></table>

### :bar\_chart:Outputs&#x20;

GNW generates standardized visual and narrative outputs designed to make data accessible and interpretable. The table below outlines the supported output types.&#x20;

<table data-header-hidden="false" data-header-sticky><thead><tr><th width="220">Output Type</th><th width="127">Supported</th><th>Notes</th></tr></thead><tbody><tr><td>Time series charts </td><td>✅ </td><td>Primary visualization </td></tr><tr><td>Bar / comparison charts </td><td>✅ </td><td>Simple comparisons </td></tr><tr><td>Written summaries </td><td>✅ </td><td>Plain-language insights </td></tr><tr><td>Raw data downloads </td><td>✅ </td><td> “View how this was generated” gives access to raw, chart data and the code used to generate the visualization </td></tr><tr><td>Custom maps and data visualizations </td><td>❌ </td><td>Not supported </td></tr></tbody></table>

[^1]: Produces the same result every time given the same inputs

[^2]: AI agent framework that enables Large LLMs to use programming (executable Python code) as a tool to solve problems, rather than relying solely on text or predefined JSON tools


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://help.globalnaturewatch.org/get-started/capabilities-and-datasets.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.