ek_daily_temperature() creates the temperature-focused daily change-factor
workflow described by Ek et al. (2018). Matching historical and future daily
tasmin and tasmax inputs are required.
The method constructs one climate baseline for each day of the annual cycle, derives daily mean temperature and DTR from the paired extrema, and applies the Ek combined shift-and-stretch equation to every baseline EPW hour. It does not apply an undocumented smoothing window and it preserves the baseline day order and within-day timing.
The paper does not fully reconcile its generic stretch-factor equation, combined temperature equation, and variance description. This implementation uses relative DTR change as the anomaly multiplier because that interpretation gives zero-change identity and the stated daily mean and variance behavior. The selected equation and calendar adaptation are retained in result provenance.
The "paper_faithful" policy changes dry-bulb temperature while preserving
the baseline humidity fields. The "harmonized" policy instead preserves
feasible baseline specific humidity and recomputes relative humidity and dew
point against the projected temperature.
ek_daily_temperature(
reference = NULL,
policy = c("paper_faithful", "harmonized")
)A required historical_reference(),
shift_reference_plan(), or extracted ShiftClimate stage.
Physical execution policy: "paper_faithful" preserves the
baseline humidity fields; "harmonized" applies shared specific-humidity
closure.
A complete ShiftMorphMethod for shift_future_epw().
Ek, M., Murdock, T. Q., Sobie, S. R., Cavka, B., Coughlin, B., and Wells, R. (2018). Future weather files to support climate resilient building design in Vancouver. https://hdl.handle.net/1828/21874