# Welcome Welcome to the Earth Index Docs! Here you'll get an overview of all the amazing features [Earth Index](https://app.earthindex.ai) offers to help you search and monitor the Earth. Our team at Earth Genome is here to help you every step of the way. We're always looking for ways to improve our public documentation, so please email us at , message us on [Discord](https://discord.gg/UZZSuxaS7P) with any feedback or submit a bug or feature request through [this form](https://app.earthindex.ai/feedback/). ### Jump right in


Getting Started	Create your first project	1.png	Create an account
Concepts	Learn the concepts of Earth Index	2.png	Concepts
Case Studies	Explore some sample uses	3.png	Case Studies

# Create an account Earth Index is live at: [app.earthindex.ai](https://app.earthindex.ai) {% hint style="info" %} Already logged in? Head to the [Starting a new project](/getting-started/starting-a-new-project) section to get started. {% endhint %}

To log in, you can "Continue with Google" or enter your email and password to receive a one-time code to verify you and create your account. ## Open Access vs Full Access: **Open Tier:** When you first sign up, you will default to being an 'Open' user. This type of user can create a project, add a reference layer and run up to 50 searches a day. {% hint style="info" %} While you are an Open user, you will see a banner and a limited project creation flow. {% endhint %}

**Full Access Tier:** If you require a more feaures that open allows, you can request to become a 'Full Access' user. Your request is reviewed by the team for impact and mission alignment and will be considered on a case by case basis.

### Join Discord: Along with logging into Earth Index, we invite you to make an account on Discord to engage with the Earth Index community. **Join the** [**Discord here**](https://discord.gg/UZZSuxaS7P)**!**

{% hint style="warning" %} If you receive an error that your email "is not allowed to access this application", please contact our team at or send us a message on Discord. {% endhint %} # Starting a new project Once you’ve finished setting up your account, you can start a new project! On the homepage of Earth Index, click on the :heavy\_plus\_sign: to start a new project.\

\ A window will pop-up allowing you to enter details about your project. There are four fields to complete before creating your project:

1. **Name**: Give the project a name relevant to the work you'll be doing. 2. **Description:** Optionally, add a description that briefly covers a little more context about your work in this project. 3. **Area of Interest (AOI):** Use the map to draw an area you want to run your search in. 1. Click the '**Define Area of Interest**' button. 2. Follow the on screen instructions but start by clicking '**Draw a Rectangle**' button then click on the map and drag to draw a bounding box of your AOI. 3. If your area is too large (max 900,000 sq km) then you can click '**Reset Drawing**' to draw again. 4. Once you have a valid AOI, you can click '**I'm Finished**' to finish the drawing step 4. **Time Period:** Select between 2024 or 2025 global imagery to search against. 5. After you've entered all the details, click on '**Create Project'** in the bottom right to start exploring! Watch this video if you need help! {% embed url="" %} # Feature Requests and Bug Reports You can use the form inside the app ([full link here](https://app.earthindex.ai/feedback/)) to submit bug reports AND/OR feature requests. Please limit to one bug or one feature per submit of the form. Thank you for helping improve the Earth Index User Experience!

# Brainstorming Think of searching with [Earth Index](https://app.earthindex.ai) as an iterative process. You need to teach, through trial and error, Earth Index to find what you are looking for. You will provide examples by adding labels to a map. Earth Index will then search for similar features across your area of interest. Before adding labels, you need to think about what you want to find with Earth Index. Once you have identified what you want Earth Index to find, think about what data it needs to assist you in the search. Before you begin, brainstorm the key features of what you will want to identify:

This exercise is intended to help you think about positive and negative associations in imagery. When perusing the satellite imagery to find a desired object, you want to think about what is unique about the object compared to other similar-looking objects. Some objects might look similar (for example, a blue swimming pool and a blue basketball court), which can affect the predictions in Earth Index. Therefore, positively associating with correct features and negatively associating with incorrect features will filter the results for a better outcome. Learn more about positive and negative labels on the [Assigning labels](/concepts/assigning-labels) page. {% hint style="info" %} Still struggling with how to brainstorm? Ask questions on the Earth Index Discord to chat with others who might be able to help! Many of our users have been at this same stage and are happy to share tips & tricks. {% endhint %} Once you've spent some time brainstorming, you're ready to jump in! # Navigating the toolbar Before you begin using Earth Index to search the planet, it is important to learn about its interface and how to navigate. In Earth Index, you can use your cursor to navigate N/S/E/W on the map and to zoom in/out of the selected region, using standard map navigation (click, drag, and use mouse wheel to zoom). {% embed url="" %} ### **Using the Toolbar** Once you've entered your project, you'll see a small toolbar on the right side. This toolbar has a series of 5 options that help you navigate and label the map.

The first three modes help you navigate and assign labels, while the latter two describe a type of label.

**Pointer:** Use this to add positive or negative labels to tiles on the map. When a tile is labeled, it will be highlighted in blue.

**Paint Brush:** Use this to designate large areas with positive or negative labels quickly.

**Eraser:** Use this to remove mistakes. As you are making labels with the pointer or paint brush, you can choose between tiles that are looking for (positive) and tiles that you're not looking for (negative). You can use the thumbs up/down buttons to toggle between the **type** of label you're assigning. ### **Setting Location & Switching Basemap** On the bottom of the screen, there are three different tools available.

1. Your longitude and latitude are displayed. If you want to shift to a different known location, click on the coordinates and enter the new location in the pop-up. 2. By clicking the layer symbol, you can select which "Basemap" to use in Earth Index. Currently, we offer Mapbox Satellite, Mapbox Outdoors, and Sentinel-2. Sentinel-2 is the default selection. 3. When zoomed out, a button will appear called "**Zoom in to draw**". When clicked, it will zoom in close enough to allow you to select tiles. You will now see the tiles pop up underneath your cursor. From afar, you are unable to select and label individual tiles, so this button enables you to automatically zoom further in and place labels! ### **Return Home** To return to Earth Index's home screen and find other projects, click the purple and white "e" in the top-left corner. ### **Keyboard Shortcuts** If you prefer to use your keyboard, we've built in some keyboard shortcuts to streamline your navigation of Earth Index. Click on the keyboard symbol in the top right to learn how to toggle between toolbar options and how to undo/redo changes.

{% hint style="info" %} If you're looking for additional keyboard shortcuts, feel free to share any feedback with our team in Discord or at . {% endhint %} # Assigning labels Now that you know how to use Earth Index, let's learn how to actually assign labels. Earth Index works through an iterative process, where users select a few tiles to show examples to the model. Using these, it can search across the entire region for other similar features. In order to find these, we need to assign labels to tiles. ### **How to approach labelling** Think about impactful positive and negative labels. For example, airport landing strips and roads can look similar. Airport landing strips, however, come to an abrupt end while most roads do not. Ensuring that you include positive/negative labels of this feature will help Earth Index learn the difference.

There is no perfect number of labels; you can try a small number of labels and see how good the outputs are. Based on the outputs from the “predictions” feature, refine your labeling process to move closer and closer to what you are looking for. Learn about predictions on the [Running predictions](/concepts/running-predictions) page. You should be able to see the number of labels, positive or negative, in the window on the top right.

### **Using Tools** Using the pointer, clicking will place a label, and clicking and dragging will move the map. With the paint brush and eraser, clicking will place/remove a label, while clicking and dragging will assign/remove labels to all tiles underneath the brush. See below for an example of using the paint brush.

{% hint style="info" %} Keep in mind that more isn’t better; painting an entire geographic area is unlikely to improve results since it inputs similar data over and over. It won’t hurt the results, but it won’t be the most efficient use of Earth Index (or your time!) {% endhint %} Once you feel confident in assigning positive and negative labels, go to the next page to run predictions. You'll quickly see how this process is very iterative. # Running predictions Once you've assigned a few labels, you're ready to run predictions. ### Running predictions To begin, click "Run Predictions".

Soon after, a series of tiles will appear on the Predictions window. These are tiles that are most similar to the positively labelled tiles. On the map, they will appear with a purple outline, clearly distinct from the blue outline (from manually labelled tiles).

After running your first prediction, you can click on the "Config" button to change the number of predictions generated.

### Filter by confidence When you've generated some predictions, our algorithm is more confident about some matches than others, within the number of predictions you desired. As you drag the slider to the right, the confidence filter will increase and the number of tiles displayed will decrease. You'll be shown tiles that our algorithm is more confident are a match as opposed to all 500 predictions, for example. Try dragging the slider to adjust the confidence level!

As you can see, the number of predicted tiles has decreased to 42. Given tiles near the already labeled ones are quite similar, we would expect them to remain, especially when filtering toward higher confidence.

### Updating predictions Once you've added additional positive or negative labels, whether from predicted tiles or other observations, you should update the generated predictions. Given this is a trial-and-error process, being iterative by adding a few more labels and updating predictions is necessary and efficient! To update, click the "Update" button in the Prediction window.

Watch this video to learn how to run predictions: {% embed url="" %} ### How does this work? Each tile is represented through an embeddings produced by the foundation model. When you select a set of tiles that are similar, we search across all the tiles in this region to find tiles with similar embeddings, or nearest neighbors. If many points are labeled positive, the machine searches for the average of those positive vectors. If you include negative labels, the machine searches for the average positive vector displaced away from the average negative (by the difference between the average positive and the average negative). {% hint style="info" %} Now that you've learned how to label and make predictions, try building a dataset of some object or feature in the imagery! Go through a trial-and-error process and play around with Earth Index. We're looking to improve the tool, so feel free to share any feedback with us over Discord or at . {% endhint %} # Reference Layers We use the term "reference layer" to refer to geospatial data that provides additional context to assist in the labeling process. This could be: * Infrastructure like roads, train lines or hospitals * Administrative boundaries like Indigenous land borders, park boundaries, or other protected land * Previously surveyed or collected datasets * Or any other geospatial information that helps provide context to the project ### Limitations Currently, reference layers have the following limitations: * Only GeoJSON files are supported. Most GIS systems should be able to output this format. You can test your files at [geojson.io](https://geojson.io/) * Files must be less than 10MB in size * You are not able to edit features or geometries that you upload. ## Managing Reference Layers ### Adding a Reference Layer Reference layers can be added by opening the layer panel on the right side of the screen. 1. Click on the "New Overlay" button 2. Select the GeoJson file you want to upload from your computer and submit After upload, there will be a short delay as the file is optimized for visualization on the web. ### Updating Reference Layers You can edit the name of the reference layer by simply clicking on the existing name and updating it in place. The display color for each layer can be changed by clicking on the color next to the name and selecting a new one from the color picker. Both the name and the color choices will be saved as part of your project and will be visible if you make the project public. ### Removing Reference Layers Reference layers can be removed from a project by clicking the 3 dots next to the layer and clicking "Delete Layer". Once removed, a layer cannot be added back without re-uploading. To simply hide the layer instead click the eye icon next to the layer. ## Working with Reference Layers One of more interesting uses of reference layers is using them to identify known examples of what you are looking to find. This allows a user to label a large number of positive points without manual review of each point, quickly building up a power mode. Let's walk through an example below. #### Starting from Scratch We can see in this project that the user has uploaded a dataset from 2017 that outlines known chicken farm buildings. We can also see a new building complex in the middle of the screen that appears to have been built after the data was collected. This is a great example of how you can leverage an existing dataset to find new data in Earth Index.

First, lets convert all of the known buildings into positive points. Click on the 3 dots next to the layer and then click "Convert to labels"

Now the grid cells that overlap with features in the reference layer file have been marked as positive points.

# Sharing a Project Sharing a project is easy, just go to the top right of your project and click '**Share**'. This will open a dialogue box where you can toggle the project sharing on and off. \ \ **Sharing disabled by default:**

**Toggle on to enable and generate share link:**

If you disable public sharing, any users with the old URL will no longer be able to view the project until you enable the same project again. ### Public View After you share the project, a public link becomes available that shows a limited viewer where users can view your labels and predictions. Users can also download those as geojson. {% hint style="info" %} NOTE: Once a project is public, anyone with the URL can view it. {% endhint %}

# Tips & tricks ### **Create multiple projects** To find all examples of what you are looking for, create multiple projects, each with different input labels. While you could label all on one map, the overall utility of search results will decline. For example: 1. Input labels for single logging roads may behave differently than labels for jumbled logging roads (near the actual current/past clearing site vs. road to access). 2. Input labels for roads in a village may produce different searches than roads in a forest. 3. Mining - creating different projects for large mines, small mines, and skinny mines on small tributaries. ### **Quality over Quantity** In looking through the predictions, find a couple instances of positive and negative labels within the predictions you think will help teach Earth Index to better find what you are looking for, instead of going through all of the predictions. Then try the search again and see if it improves. Labelling is a process of trial and error, so don't get discouraged if it takes a while to find the right dataset. ### Basemap note When exploring the toolbar, you have the ability to select which "Basemap" you would like to display on the interface. There are three options for Basemap including: Mapbox Satellite, Mapbook Outdoors, and Sentinel-2. The two Mapbox maps are intended to be a helpful reference for exploring the Earth, but Sentinel-2 is the primary satellite imagery. This imagery has been indexed into embeddings that are the core essence of Earth Index, allowing you to search across different tiles and find objects. Thus, when labeling tiles and running predictions, our algorithms use the embeddings based on Sentinel-2 imagery. # Quick validation Quickly assign labels to your predictions Once you have added some initial labels and ran predictions using your labels, a series of tiles appear in the prediction window that might be similar to your chosen labels. In order to quickly assign labels to your predictions, you can select "Quick Validation" at the bottom of the prediction window. Quick validation will step through a series of about 50 tiles in a row, with the same pattern of "observe and label" for each tile. This feature aims to help you focus on assigning labels to individual tiles, without worrying about zooming in to each predicted tile. You can quickly label the predictions and help refine subsequent searches. {% embed url="" %} ### **Entering Quick Validation** Selecting "Quick Validation" launches a separate window to assist with quick labelling, where predicted tiles are sorted into a queue. Earth Index will zoom in directly to the first unlabeled tile, allowing you to closely observe its features.

### **Labelling** If you are able to identify either a positive or negative association, you can select either option to label the tile. To learn about labelling, read this article: [Assigning labels](/concepts/assigning-labels) If it is uncertain or confusing how to label, you can click "Skip" to proceed to the next tile. ### **Filter by Confidence** If you want to curate *which* tiles are being shown in this workflow, you can toggle between various levels of "confidence". Confidence suggests how certain our model is that the current tile is very similar to the labeled set of tiles. You can alter the quick validation workflow to show you "High Confidence", "Medium Confidence", and "Low Confidence" tiles.

# Exporting or deleting labels ### Deleting labels If you're not feeling confident in the labels that you have assigned, and you just prefer to start from the beginning, no worries. You don't need to create an entirely new project, because you can delete all existing labels. Clicking the three dots near "Labels" allows you to "Delete All Labels" and start fresh in this projects.

### Exporting data Earth Index is incredibly useful for building a dataset of locations via satellite imagery that are similar due to similarities in their embeddings. However, it can be challenging to use this dataset for downstream analysis solely within Earth Index. Clicking the three dots near "Labels", as shown above, allows you to "Export GeoJSON" for use elsewhere! Watch this video to learn more: {% embed url="" %} # Using "Compare" When you want to view two different searches side-by-side, you can use Earth Index's **Compare** tool. You might be interested in comparing how a certain region has changed over time, or seeing how two searches differ based on the labels you place. Compare allows you to simultaneously view two different searches, and observe the # Logging roads on Malaita Island Malaita, the most populous island in the Solomon Islands, is currently facing severe threats to its critical ecosystems due to unregulated logging, driven by foreign demand and population growth. To protect its rich biodiversity and fragile environment, we want to use Earth Index to identify and monitor logging roads on Malaita. This will help us assess the extent of the harmful logging activities and inform necessary action. Let’s start by brainstorming key features to label to help Earth Index identify what I am looking for:

### **Positive Labels** Next, I find an example of a logging road and add in positive labels.

I can continue this process with other roads to input more verified labels, or I can try to run predictions to see how Earth Index will do with this amount of data.

### **Running Predictions** By generating a high number of predictions, it is more about its accuracy in aggregate vs. an individual square. Where predictions are clustered, it shows an area that would be worthwhile to investigate. Using the “Config” button, however, allows you to adjust the number of predictions that are generated (1-500). Adjusting the slide bar from low to high confidence can also customize the predictions displayed. ### **Trial & Error** At quick glance, when I look at my results, there are too many non-road objects identified. I will label a few more positive examples and update the predictions.

This time, I spot check some areas with a large number of clustered predictions. Sure enough, the results have improved and other logging roads are identified:

### **Negative Labels** I also find areas labeled that I am not interested in - which I change from predictions to negative labels. As villages, they show the clearing/deforestation that is common with logging roads, which is why they were predicted as something I might be interested in, but it is ultimately not what I am looking for. ### **Updating Predictions** While the change is small, after editing several predictions that identified villages and hitting

the data points continue to refine, with outliers disappearing and clusters refining over the logging roads. As I look closer, I see some dirt roads that are not logging roads were identified. This is where I need to “teach” Earth Index how to differentiate between them. This will involve a bit of trial and error, but I know I want to focus on the color differences (logging roads are reddish) and the deforestation.

### **Quick Validation** Another tool for validating the data is the “Quick Validation” button. Using the “filter by” drop down menu is helpful in targeting the type of data you want to work with to improve the search. High Confidence data will be quicker in multiple positive labels, the reverse for low confidence. Medium allows you to add a human touch to help the predictions with more ambiguous labels. All have their use depending on what you are seeing the tool do.

As you edit your data, use the update features to continue to refine the labels.

### **Export Data** When are you done? It’s up to you and the goals of the project you are working on. When you have data that you feel will be helpful, click on the three gray dots to expand to “Export GeoJSON” data.

Watch this video if you're curious to learn more! {% embed url="" %}