How to verify geolocation information related to EUDR Due Diligence Statement- 1/5

Task 1: Check the Alignment of AOIs with Administrative Boundaries

In Step 4: Registering Basic Information About the Forest Source of my guidance for timber suppliers, I emphasize the importance of collecting comprehensive data for EUDR compliance, including geolocation information recorded by forest managers on the ground. In "Smallholders Using Google Earth to Draw Their Operation Area Map," I briefly introduced how such data collection can be conducted. Now, in this detailed article series, I will outline my approach to verifying geolocation data from five key perspectives:

• Check the alignment of AOIs with administrative boundaries and extract the most detailed available administrative information for each AOI.
• Verify forest status as of December 31, 2020, using the GFC2020 and GFT2020 datasets.
• Conduct NDVI analysis to identify anomalies before and after the logging date. Track NDVI changes over time to monitor potential land use conversion.
• Track canopy changes in the tropics, and identify conversions using other datasets.
• Combine satellite imagery analysis with ground-based verification, such as drone surveys, DNA testing, and field audits.

This guidance involves Google Earth Engine (GEE) and QGIS, requiring some level of coding and GIS analysis. However, as data verification is typically conducted by more resourceful entities further down the supply chain, my approach ensures that no high-end computing power or advanced coding skills are needed. The methods outlined here are accessible and practical, and I will walk you through the steps with screenshots and detailed explanations, ensuring that anyone can follow along. This approach is particularly useful for companies managing bulk procurement or sourcing from multiple forest units, allowing them to efficiently verify geolocation data for numerous logging sites.

Task 1: Check the Alignment of AOIs with Administrative Boundaries

For this task, we use publicly available datasets and free tools such as QGIS and GEE to obtain administrative boundary information for an Area of Interest (AOI). We then compare this data with official documents provided by the forest unit, such as logging licenses or land tenure records, to ensure that the geolocation information corresponds accurately to legal documentation. The results can also help categorize AOIs based on their locations for further analysis in other tasks.

A similar approach can be applied to determine whether an AOI is located within or near critical areas, such as conservation zones, high conservation value areas (HCVs), or indigenous cultural landscapes, by referencing relevant boundary datasets.

Step 1. Get familiar with GEE and its user interface

Google Earth Engine (GEE) is a cloud-based platform that allows users to analyze geospatial data efficiently. To begin, visit the GEE Code Editor: https://code.earthengine.google.com/

If you don’t have an account yet, follow the on-screen instructions to register and create a project with a name of your choice. Once registered, you’ll see the GEE user interface, which consists of the following key elements:

Uploading Your Own Data as Assets

To analyze your own geolocation data, follow these steps to upload it to GEE:

1. Click on the "Assets" tab (found on the left panel).
2. Click the red "NEW" button to add a new dataset.
3. Choose the appropriate data format from the list.
4. If your data is in KML or KMZ format (as commonly submitted by forest managers), you will need to convert it to a Shapefile (SHP) or GeoJSON format before uploading. i will show the approach in step 2.

Running a Script in GEE

After successfully uploading the dataset, you can now paste and edit a script in the Script Editor. Once ready, click "Run" to execute the script and analyze the data.

Debugging & Viewing Results

If there are any errors in the script, they will appear in the Console window below the editor. Use this area to troubleshoot issues and refine the script. Once the script runs successfully, you will see the results displayed on the map or in the Console, depending on your script’s output.

Step 2. Prepare your AOI data

To ensure accurate analysis, you need to properly format and upload your Area of Interest (AOI) data. You can either combine multiple AOIs into a single asset locally before uploading to Google Earth Engine (GEE) or upload each AOI separately and merge them in the script.

Option 1: Combine Multiple AOIs into a Single Asset locally

This method is ideal if you have multiple AOI files (e.g., Shapefiles, GeoJSON, KML) and want to merge them into a single dataset before uploading to GEE.

1. Merge AOIs Using QGIS

Load Multiple AOI Files

• Open QGIS and import all your GPS or vector files (Shapefiles, KML, GeoJSON).
• You should see all layers loaded on the left panel.

Merge AOI Layers

• In QGIS Processing Toolbox, search for "Merge Vector Layers" and select it.
• Choose the AOI layers you want to merge.

Convert Multipart to Singlepart

• After merging, you may need to convert multipart geometries into singlepart geometries for better compatibility.
• Use the "Multipart to Singleparts" tool in QGIS to do this.

Export as a Shapefile

• Export the merged AOI dataset as a Shapefile (SHP).
• Compress all generated files (.shp, .shx, .dbf, .prj) into a ZIP file for easy upload to GEE.

2. Upload to Google Earth Engine

• Open GEE Code Editor (https://code.earthengine.google.com/).
• Go to the "Assets" tab.
• Click New Asset > Table Upload.
• Upload the compressed ZIP file containing the AOI shapefile.

3. Reference the Combined AOI in Your Script

Once the AOI file is uploaded to GEE, reference it in your script:

var AOIs = ee.FeatureCollection(“users/your_username/combined_AOI_asset”); // Replace with your AOI asset ID

Option 2: Add Multiple AOIs as Separate Assets on GEE

If you prefer to keep AOIs as separate assets, follow this approach.

1. Upload Each AOI Individually

• Upload each AOI file separately in GEE Assets.
• If your data is in KML or other unsupported formats, use QGIS to convert it as instructed above before uploading.

?2. Merge AOIs in Your GEE Script

After uploading multiple AOIs separately, you can combine them programmatically using merge() or union().

Example Script to Merge Multiple AOIs

// Load individual AOIs from GEE assets

var AOI1 = ee.FeatureCollection("users/your_username/AOI_1");

var AOI2 = ee.FeatureCollection("users/your_username/AOI_2");

var AOI3 = ee.FeatureCollection("users/your_username/AOI_3");

// Merge all AOIs into a single FeatureCollection

var AOIs = AOI1.merge(AOI2).merge(AOI3);

Map.addLayer(AOIs, {color: 'blue'}, "Merged AOIs");

For large datasets, Option 1 (QGIS Merge) is recommended as it simplifies the number of files to manage in GEE. If flexibility is needed (e.g., adding AOIs frequently), Option 2 (GEE Merge) is a better choice.

Step 3: Prepare the Administrative Boundary Data

To align AOIs with administrative boundaries, we need reliable boundary datasets. Different sources provide varying levels of granularity, allowing us to select the most appropriate dataset for verification.

Available Data Sources

1. FAO GAUL (Global Administrative Unit Layers 2015) – Available on GEE

• This dataset provides up to level 4 administrative boundaries, corresponding to municipality level in China.
• It can be accessed directly within Google Earth Engine (GEE).

2. OpenStreetMap (OSM) – Downloadable for Finer Details.

• OpenStreetMap is a free, community-driven world map that allows access to finer administrative levels.
• For China, it can provide up to level 6 boundaries (county-level).

Download Administrative Boundaries from OpenStreetMap (OSM)

1. Register and Choose an Access Plan

• Go to OSM Boundaries and register for a free account.
• If needed, choose the Explorer Plan for additional features.

2. Download the Target Country’s Administrative Boundaries

• Select the country (e.g., China) and the administrative levels you need (e.g., Level 1 to Level 6).
• Download the dataset in GeoJSON format.

3. Convert the GeoJSON Data to Shapefile (SHP) Using QGIS

Since GEE requires Shapefiles (SHP) for vector data, follow these steps to convert GeoJSON to SHP:

1. Open QGIS and import the GeoJSON file.
2. Follow the same method used in Step 2 to convert the AOI data:Go to Processing Toolbox → "Vector General" → "Convert format"Select GeoJSON as input and Shapefile (SHP) as output.
3. Once converted, compress all generated SHP files (.shp, .shx, .dbf, .prj) into a ZIP file for upload to GEE.

Step 4. Alignment

In this step, we verify whether the Areas of Interest (AOIs) intersect with the corresponding administrative boundaries. This helps ensure that geolocation information aligns with official legal documents such as logging licenses or land tenure records.

We use the .intersection() or .contains() functions in Google Earth Engine (GEE) to check alignment. The results can be printed for review, and non-aligned AOIs can be highlighted for further investigation.

Two Methods for AOI Alignment

Below are two script options for this alignment task:

1. Using the uploaded administrative boundary data from Step 3 (e.g., OpenStreetMap data).
2. Using the FAO GAUL dataset (available directly in GEE).

You can choose the one that best fits your dataset and project needs.

// Script 1: Using Custom Assets

// Define AOI and boundary data paths

var AOIs = ee.FeatureCollection("projects/ee-malichao/assets/zzxt3"); // Replace with your AOI asset ID

var boundaryData = ee.FeatureCollection("projects/ee-malichao/assets/level7"); // Replace with your boundary asset ID

?

// Function to buffer points (50m) and handle polygons as is

var bufferAOI = function(aoi) {

? return (aoi.geometry().type() === 'Point') ? aoi.buffer(50) : aoi;

};

?

// Apply buffer if necessary

var processedAOIs = AOIs.map(bufferAOI);

?

// Add AOIs and boundaries to the map

Map.addLayer(processedAOIs, {color: 'blue'}, "Processed AOIs");

Map.addLayer(boundaryData, {color: 'green'}, "Custom Boundaries");

?

// Function to check intersection with boundaries

var checkAlignment = function(aoi) {

? // Filter boundaries that intersect the AOI

? var intersectsBoundaries = boundaryData.filterBounds(aoi.geometry());

? return aoi.set({

??? intersects: intersectsBoundaries.size().gt(0), // True if intersects

??? boundaryNames: intersectsBoundaries.aggregate_array("name").distinct() // Get unique boundary names

? });

};

?

// Map the function over all AOIs

var alignedAOIs = processedAOIs.map(checkAlignment);

?

// Print aligned AOIs to the console

print("Aligned AOIs:", alignedAOIs);

?

// Export results to Google Drive

Export.table.toDrive({

? collection: alignedAOIs,

? description: "Custom_Boundary_Alignment",

? fileFormat: "CSV"

});

?

// Highlight non-aligned AOIs on the map

var nonAlignedAOIs = alignedAOIs.filter(ee.Filter.eq('intersects', false));

Map.addLayer(nonAlignedAOIs, {color: 'red'}, "Non-Aligned AOIs");

?

// Center the map on the AOIs

Map.centerObject(processedAOIs, 6);

// Script 2: Using FAO Data

// Define AOI and FAO level

var AOIs = ee.FeatureCollection("projects/ee-malichao/assets/zzxt3"); // Replace with your AOI asset

var useFAOLevel = 2; // Set FAO level: 0 for countries, 1 for regions, 2 for districts, etc.

?

// Load FAO datasets dynamically

var FAODataset0 = ee.FeatureCollection("FAO/GAUL/2015/level0");

var FAODataset1 = ee.FeatureCollection("FAO/GAUL/2015/level1");

var FAODataset2 = ee.FeatureCollection("FAO/GAUL/2015/level2");

var FAODataset3 = ee.FeatureCollection("FAO/GAUL/2015/level3");

?

// Select the dataset for the desired level

var selectedFAODataset = useFAOLevel === 0 ? FAODataset0 :

???????????????????????? useFAOLevel === 1 ? FAODataset1 :

???????????????????????? useFAOLevel === 2 ? FAODataset2 : FAODataset3;

?

// Filter FAO dataset to the AOI extent to reduce processing

var filteredFAODataset = selectedFAODataset.filterBounds(AOIs.geometry());

print("Filtered FAO Dataset:", filteredFAODataset);

?

// Function to buffer points (50m) and handle polygons as is

var bufferAOI = function(aoi) {

? return (aoi.geometry().type() === 'Point') ? aoi.buffer(50) : aoi;

};

?

// Apply buffer if necessary

var processedAOIs = AOIs.map(bufferAOI);

?

// Add AOIs and filtered FAO boundaries to the map

Map.addLayer(processedAOIs, {color: 'blue'}, "Processed AOIs");

Map.addLayer(filteredFAODataset, {color: 'green'}, "Filtered FAO Boundaries");

?

// Function to check intersection with FAO boundaries and include higher-level names

var checkAlignment = function(aoi) {

? var intersectsLevel0 = FAODataset0.filterBounds(aoi.geometry());

? var intersectsLevel1 = FAODataset1.filterBounds(aoi.geometry());

? var intersectsLevel2 = FAODataset2.filterBounds(aoi.geometry());

? var intersectsLevel3 = FAODataset3.filterBounds(aoi.geometry());

?

? var level0Names = intersectsLevel0.aggregate_array("ADM0_NAME").distinct();

? var level1Names = intersectsLevel1.aggregate_array("ADM1_NAME").distinct();

? var level2Names = intersectsLevel2.aggregate_array("ADM2_NAME").distinct();

? var level3Names = intersectsLevel3.aggregate_array("ADM3_NAME").distinct();

?

? var allNames = ee.List([]);

? if (useFAOLevel >= 0) allNames = allNames.cat(level0Names);

? if (useFAOLevel >= 1) allNames = allNames.cat(level1Names);

? if (useFAOLevel >= 2) allNames = allNames.cat(level2Names);

? if (useFAOLevel >= 3) allNames = allNames.cat(level3Names);

?

? return aoi.set({

??? intersects: allNames.size().gt(0), // True if any intersection

??? FAONames: allNames.distinct() // Combine all names without duplication

? });

};

?

// Map the function over all AOIs

var alignedAOIs = processedAOIs.map(checkAlignment);

?

// Print aligned AOIs to the console

print("Aligned AOIs:", alignedAOIs);

?

// Export results to Google Drive

Export.table.toDrive({

? collection: alignedAOIs,

? description: "FAO_Boundary_Alignment_Hierarchical",

? fileFormat: "CSV"

});

?

// Highlight non-aligned AOIs

var nonAlignedAOIs = alignedAOIs.filter(ee.Filter.eq('intersects', false));

Map.addLayer(nonAlignedAOIs, {color: 'red'}, "Non-Aligned AOIs");

?

// Center the map on the AOIs

Map.centerObject(processedAOIs, 6);

Step 5 Accessing and Interpreting the Results

Once the alignment analysis is completed, it’s essential to review, interpret, and use the results effectively. This step will guide you through:

1. Submitting the Task in GEE

After running the script, GEE does not automatically generate a downloadable file. You must manually trigger the export by following these steps:

Steps to Submit the Task

1. Locate the "Tasks" panel on the right side of the GEE interface.
2. Find the unsubmitted task labeled "Custom_Boundary_Alignment" (or the name you defined in the script).
3. Click the "Run" button next to it.
4. In the pop-up window, choose the destination folder in Google Drive and confirm by clicking "Run" again.?

?2. Downloading the Results from Google Drive

Once the export task is completed:

1. Open Google Drive (https://drive.google.com).
2. Navigate to the folder you selected during the export process.
3. Find the exported CSV file (e.g., AOI_Boundary_Alignment.csv).
4. Download the file to your local device.

3. Understanding the Results

The CSV file contains key information about AOI alignment with administrative boundaries. For script 1, the alignment results are included under the item "boundaryName", which in Chinese say"China, Mainland China, Henan Province, Nanyang Municipality, Tongbai County".

Similarly the result from script 2 is under the item "FAONames", which in English and Chinese Phonics say “China, Henan Province and Nanyang”

4. Identifying & Troubleshooting Issues

If an AOI does not align properly, consider these troubleshooting steps:

? Check if the AOI coordinates were correctly formatted

? Verify that the correct administrative boundary dataset was used (FAO GAUL vs OSM)

? Confirm that the AOI does not fall outside mapped administrative areas

If misalignments persist, consider:

• Using alternative boundary datasets (e.g., refining OSM or FAO GAUL datasets)
• Re-examining the source data (ensuring AOIs were properly drawn in Step 2)
• Consulting with local authorities or experts for verification