Detect Outlier Climate Models Using RMSE Thresholding

Iteratively removes climate model columns whose RMSE from the ensemble mean exceeds a user-defined threshold. Assumes input is a matrix or data frame with time series in rows and models in columns.

Usage

detect_outliers(
  data,
  start_index,
  end_index,
  threshold_fn = mean_sd_threshold,
  threshold_dial = 2.5
)

Arguments

data

A numeric matrix or data frame with model outputs.

start_index

Integer, starting row index (e.g., corresponding to 2015).

end_index

Integer, ending row index (e.g., corresponding to 2100).

threshold_fn

A function taking a numeric vector of RMSE values and returning an RMSE threshold. Default is mean_sd_threshold().

threshold_dial

Numeric multiplier passed to threshold_fn. Default is 2.5.

Value

A character vector of column names identified as outliers.