Skip to contents

Structure, print method, and four-panel summary plot for objects returned by streamflowQC_daily, streamflowQC_subdaily, and streamflowQC.

Usage

# S3 method for class 'streamflowQC'
print(x, ...)

# S3 method for class 'streamflowQC'
plot(
  x, max.stations=20L,
  col=c("#2878B5", "#D55E00", "#E69F00"), ...
)

Arguments

x

An object inheriting from class "streamflowQC".

max.stations

Maximum number of stations in the affected-station panel.

col

At least three plotting colours.

...

Additional arguments; the plot method passes them to its final time-series panel.

Details

The plot summarizes station recommendations, missing/review/rejected fractions, counts by active test, and the time distribution of confirmed suspicious observations. Plotting is diagnostic and does not alter the result or write a file unless the caller explicitly opens a file device.

Value

Both methods return x invisibly. A "streamflowQC" object is a list with:

  • accepted.metadata and discarded.metadata: original station metadata plus completeness, evidence, homogeneity, recommendation, and reason fields;

  • accepted.data and discarded.data: original zoo series split by the station recommendation;

  • accepted.corrected and discarded.corrected: the same partitions after the requested correction policy;

  • suspicious: one row per point flagged for review or rejection, including all contributing test names and any spatial estimate;

  • corrections: an audit table containing every changed value;

  • flags, flag.count, and rejected: individual logical flag objects, evidence counts, and confirmed decisions;

  • station.summary: station completeness, review/rejected fractions, breakpoint diagnostics, recommendation, and reason;

  • breakpoint: annual-indicator Pettitt-test diagnostics;

  • spatial.estimate, spatial.score, spatial.support, and neighbours: spatial-test diagnostics;

  • highflow and return.period: component high-flow flags, GEV parameters, and estimated return periods; and

  • settings: all resolved workflow controls.

References

Fileni, F. et al. (2026). UK-Flow15-QC: A quality control framework for better river flow data in hydrological research. EGUsphere preprint. doi:10.5194/egusphere-2026-277.