Tonga Arterial Road Surface and Passability Metrics (2020-2024)

HeiGIT generated this geospatial dataset of road surface, width, and passability metrics for approximately 100 km of arterial roads in Tonga. Using PlanetScope satellite imagery from 2020 and 2024, the analysis provides deep-learning predictions for segments where OpenStreetMap data is incomplete.

Use Cases

• Evaluating heavy vehicle access using 'Road Width Class' and 'Humanitarian Passability Index'
• Identifying infrastructure upgrades via the 'Road Surface Change' indicator
• Supplementing missing OSM 'surface' tags with AI-predicted paved/unpaved labels

Strengths

• 89.2% surface model accuracy
• Multi-temporal surface predictions for 2020 and 2024
• Standardized 6-point Humanitarian Passability Index

Limitations

• Geographic scope limited to Tonga
• Excludes minor and residential road classes
• AI-derived predictions may contain errors inherent to satellite imagery analysis

Provenance

Source

Heidelberg Institute for Geoinformation Technology (HeiGIT)

Collection Method

Deep learning analysis of PlanetScope satellite imagery integrated with OpenStreetMap road geometries.

Time Range

2020 - 2024

Freshness

Contains imagery-based analysis from 2020 and 2024; metadata last updated March 2026.

Geography

Tonga