Data and code for estimating dust column mass and surface PM2.5 over Baghdad using machine learning
Description
This dataset contains daily data used to estimate dust column mass and surface PM2.5 over Baghdad for the period 2014–2025. The package includes the processed input data, Python notebooks for hyperparameter tuning and final model evaluation, selected model configurations, and results for 2025. The models use five predictors: air temperature, high cloud cover, aerosol optical depth at 550 nm, precipitation, and dewpoint temperature. The reference targets are gridded aerosol products. Further details about the files and their use are provided in the README file.
Files
Baghdad_ML_Zenodo_Draft_v3.zip
Additional details
Software
- Programming language
- Python