GPM GV MASC UConn dataset contains high-resolution photographs of snowflakes and frozen hydrometeors captured from three angles by the Multi-Angle Snowflake Camera deployed by University of Connecticut. The collection includes PNG images and a CSV file with additional details for each observed particle. NASA's GHRC DAAC provides data from January 14, 2023 to March 21, 2024.
Use Cases
- Classify hydrometeor type from tri-view PNG images using convolutional neural networks.
- Analyze correlations between snowflake fall speed and morphological features from the CSV measurements.
- Train generative models to synthesize realistic snowflake images based on multi-angle camera data.
- Study temporal evolution of snowflake properties using timestamps associated with each image and CSV entry.
Strengths
- Provides synchronized three-view imagery for each snowflake, enabling 3D shape analysis.
- Covers a full winter season from January 2023 to March 2024 for temporal studies.
Limitations
- Sample size and geographic scope are unspecified, limiting generalizability assessment.
- Data collection is limited to a single deployment period and location.
Provenance
- Source
- NASA GHRC DAAC and University of Connecticut.
- Collection Method
- Captured using a ground-based Multi-Angle Snowflake Camera with three cameras separated by 36°.
- Time Range
- January 14,我们发现2023 to March 21, 2024.
- Freshness
- Data collection ended March 21, 2024; dataset is a static snapshot.
- Geography
- null