- Make_scatterplots.Rmd -- create scatterplots of the observations in the validation set, colored by year and by state; make a map of the locations of the monitors in the validation set (colored by how many days each monitor was collecting data)
- Map_county_averages.Rmd -- show overall and seasonal county averages (maps) for the Di and Reid datasets as well as the differences between them
- Spatial-temporal_plots.Rmd -- plotting spatial and temporal metrics by subset (year, state, season)
- Weighted-summary-stats_validation-set.R -- Generating summary statistics of the validation set
- Metrics_weighted.Rmd -- Calculating the main comparison results (i.e. Di vs Monitor and Reid vs Monitor)
- Spatial-temporal_metrics_weighted.Rmd -- Calculating spatial and temporal metrics by subset (e.g. warm season vs cold season), and comparing the Reid and Di results
- AQI_table.R -- Calculating the AQI classification and binary classification results
- Process_raw_smoke_data.Rmd -- Extracts data from the raw Airsis and WRCC (smoke validation monitor) files.
- Monitor_processing.Rmd -- Combines Airsis and WRCC data. Data points from this validation set are removed if they are over 1,000
$\mu g / m^3$ or if they overlap with the Reid training observations. Later, we determined that this was insufficient for removing all overlapping points, so the final cleaning steps are shown in Removing-overlap_validation-set.Rmd - Reid_processing.Rmd -- Aggregates any duplicated observations in the Reid dataset.
- Nearest points.Rmd -- For each monitor observation, finds the nearest Reid and Di observations.
- Add_covariates.R -- Making sure the spatial and temporal covariates (e.g. "GEOID" and "warm season") were filled in and consistent throughout the main validation set (2008-2016).
- get_Pop-density.R -- obtaining and merging in population density data to enable weighting by population density
- Although the validation monitor observations were obtained using the R package PWFSLSmoke, we actually were sent the data resulting from this query by a member of the AirFire research team.
- The Di data can be accessed on the NASA SEDAC website
- The Reid data can be accessed on Figshare.