Building Damage Assessment from the 2015 Gorkha Earthquake in Nepal

A dataset of over 760,000 buildings across 11 districts in Nepal, compiled by Nripesh Poudel for research on the 2015 Gorkha earthquake. It includes structural and geometric attributes used to predict multi-class damage states with an interpretable machine learning framework. The dataset was last updated in May 2026.

Use Cases

• Predicting building damage states based on structural attributes like age, height, and plinth area
• Identifying key vulnerability drivers, such as mud-mortar stone superstructures and plan irregularity, using SHAP analysis
• Clustering regions into distinct vulnerability signatures based on local construction practices and topography
• Supporting targeted retrofitting and policy development for disaster risk reduction

Strengths

• Includes data for over 760,000 buildings, providing a substantial sample size
• Covers 11 districts, enabling regional analysis
• Features multiple structural attributes, including age, height, plinth area, and types of roofs, foundations, floors, and superstructures
• The analysis framework integrates predictive modeling with structural engineering principles for transparency

Limitations

• Column-level documentation is absent; field semantics must be inferred after download
• Row count is unknown, which may limit suitability assessment
• Data may reflect geographic bias inherent to the specific earthquake event and surveyed districts

Provenance

Source

figshare

Collection Method

Likely compiled from post-earthquake surveys or assessments for the 2015 Gorkha earthquake.

Time Range

Associated with the 2015 Gorkha earthquake.

Freshness

Last updated 2026-05-09 07:43:32; freshness should be verified

Geography

11 districts in Nepal.