EXPLORE

Comparisons

Choose a dataset, location and pollutant above, then explore hourly patterns and historical comparisons.

A year, hour by hour

Actual final hourly concentrations; choose how colors represent the concentration range.

The seasonal rhythm

How closely does the seasonal profile agree?

Normalization: each location’s monthly mean is divided by its own full 2015–2024 reference mean. These are local monthly means, not the across-location median curves in the manuscript. Spatial IQR bands are intentionally omitted for this single-location view.

Where each month comes from

Hover for the final F–S rank and persistence decision.

Historical-reference comparisons

How do the distributions compare?

An empirical cumulative distribution shows the fraction of hours at or below each concentration.

Normalized Wasserstein distance = W1 divided by the local reference interquartile range. Lower values indicate closer agreement with the full 2015–2024 reference distribution. Undefined when the reference IQR is zero. Historical records include the project’s gap-filled values.

Across-location perspective

Each point is a site or county in the selected dataset; orange marks the selected location.

Points below the 1:1 line indicate that TAPY is closer to the long-term reference than the median of the ten individual historical-year distances. This is in-sample representativeness, not independent validation of the filling models.