Phy-DL FMF: Global Land Aerosol Fine-Mode Fraction, 2001-2020

A 20-year global land dataset of aerosol fine-mode fraction (FMF) at 500 nm, retrieved by integrating physical and deep learning methods with MODIS satellite data. The data is provided at daily temporal resolution and 1° spatial resolution in GeoTIFF format, spanning from -89.5° to 89.5° latitude and -179.5° to 179.5° longitude. It was produced by Xing Yan of Beijing Normal University.

Use Cases

• Discriminating anthropogenic from natural aerosols based on the fine-mode fraction parameter.
• Analyzing long-term trends in aerosol composition over global land areas based on the 20-year time series.
• Validating and improving satellite-based aerosol retrieval algorithms using the integrated physical-deep learning method.
• Studying regional air quality and climate impacts using the daily, 1° resolution geospatial data.

Strengths

• Covers a 20-year time range from 2001 to 2020.
• Provides daily temporal resolution and 1° spatial resolution.
• Integrates physical and deep learning methods for retrieval, which may reduce uncertainty.

Limitations

• Data may be unavailable for certain days due to missing MODIS source data.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count and file size are unknown, which may limit suitability assessment.

Provenance

Source

Beijing Normal University

Collection Method

Integrated physical and deep learning retrieval based on MODIS satellite data.

Time Range

2001 to 2020

Geography

Global land, -89.5° to 89.5° latitude, -179.5° to 179.5° longitude