Raindrop Size Distribution and Atmospheric Variables for Machine Learning

A dataset for training and validating machine learning models to retrieve raindrop size distributions. It contains DSD-derived variables and environmental meteorological variables from automatic weather stations and ERA5 reanalysis. The dataset is 71.6 MB in size, was authored by Limin Lin, and was last updated on April 12, 2026.

Use Cases

• Training machine learning models to predict raindrop size distribution shape parameters based on surface meteorological variables.
• Validating retrieval algorithms for rain rate and liquid water content using integrated disdrometer and reanalysis data.
• Analyzing temporal patterns in precipitation microphysics using 5-minute and 10-minute averaged DSD variables.
• Developing models for tropical cyclone case studies using the provided independent test dataset.

Strengths

• Includes a dedicated independent test dataset for tropical cyclone cases, which supports model validation for specific weather events.
• Integrates two distinct data categories: DSD-derived variables and environmental meteorological variables from multiple sources (AWS and ERA5).
• Provides time-averaged values (5-minute and 10-minute) for key DSD integral moments, which may support temporal analysis.

Limitations

• Row count is unknown, which may limit suitability assessment for large-scale model training.
• Column-level documentation is absent; field semantics must be inferred after download.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

figshare, associated with a specific research publication.

Collection Method

Likely contains integrated observations from disdrometers, automatic weather stations, and ERA5 reanalysis data.

Freshness

Last updated 2026-04-12 17:01:21; freshness should be verified.