Forest Foresight provides a user-friendly web interface for those who need predictions for specific areas of interest. This interface combines ease of use with powerful analysis capabilities, making it accessible to users with varying levels of technical expertise.

Access Point: http://forestforesight-wwf.shinyapps.io/interface_forestforesight/

1. Input Requirements

The web interface accepts the following input formats:

• Shapefile (.shp and associated files)

• KML (.kml)

• GeoJSON (.geojson)

Geographical Constraints:

• The area of interest must be located between 30 degrees North and 30 degrees South latitude

• This constraint ensures that the analysis focuses on tropical and subtropical regions where Forest Foresight's predictions are most relevant and accurate

2. Processing

Once a user uploads their area of interest, the web interface performs the following steps:

1. Validates the input file and checks if it falls within the specified latitude range

2. Extracts the Forest Foresight predictions for the specified area

3. Combines the predictions with data from the Forest Foresight open data repository

4. Applies the standard classification and styling to the predictions

3. Preview

Before generating the final outputs, the web interface provides an interactive preview:

• Utilizes a Leaflet-based interactive map

• Displays the user's area of interest overlaid with the styled prediction data

• Allows users to pan and zoom to explore the predictions in context

4. Output Options

Users can choose from three output formats, each serving different needs:

1. PDF Report

   • Includes a base map of the area of interest

   • Overlays the prediction raster, styled according to deforestation probability

   • Color scheme: Yellow (0.5 probability) to Red (1.0 probability)

   • Provides a legend and basic metadata about the prediction

   • Suitable for quick visual assessment and sharing with non-technical stakeholders

2. Vector Format (GeoJSON)

   • Polygonizes individual raster cells with a prediction value above 0.5

   • Each polygon represents an area of potential deforestation

   • Attributes include the original prediction value (0.5 to 1.0)

   • Ideal for further analysis in GIS software or for integration with other vector datasets

3. Raster Format (GeoTIFF)

   • Provides the raw prediction data as a GeoTIFF raster

   • Pixel values range from 0.5 to 1.0, representing the confidence of deforestation prediction

   • Maintains the full resolution and precision of the original predictions

   • Suitable for advanced spatial analysis, custom visualization, or integration into existing raster workflows

5. Prediction Classification

The standard classification for Forest Foresight predictions is as follows:

• Values below 0.5 are not included in the outputs, as they represent areas with low probability of deforestation

• Values from 0.5 to 1.0 represent increasing confidence in the deforestation prediction:

  • 0.5 - 0.7: Moderate risk of deforestation

  • 0.7 - 0.9: High risk of deforestation

  • 0.9 - 1.0: Very high risk of deforestation

This classification is reflected in the styling of both the interactive preview and the PDF output.

6. Use Cases

The web interface is particularly useful for:

• Conservation organizations focusing on specific protected areas

• Researchers studying deforestation patterns in particular regions

• Local governments assessing deforestation risks in their jurisdictions

• Companies evaluating deforestation risks in their areas of operation

• Journalists or educators creating visualizations of deforestation threats

By providing this user-friendly web interface alongside the more technical S3 bucket access, Forest Foresight ensures that its valuable deforestation predictions are accessible to a wide range of users, regardless of their technical background or specific needs.

[Remainder of the document remains unchanged]