daily_temperature() creates a future-EPW method from matching future and
historical daily CMIP temperature data. It requires frequency = "day".
The configured workflow retains the selected registered recipe identity and
"harmonized" execution policy while preserving the public method name.
Daily tas changes are estimated with a circular climatology on a common
365-day phase grid. When paired tasmin and tasmax are available for both
periods, each baseline 24-hour profile is constrained to the requested daily
mean, minimum, and maximum while retaining its hourly ordering. Otherwise,
the daily mean change is applied additively and the baseline daily range is
inherited.
reconstruction = "power" uses epwshiftr's monotone power projection.
"btws" instead selects the bounded temperature weighted stretch published
by Eames et al. (2024). The latter remains a composite comparison: only its
hourly reconstruction comes from Eames, while the daily CMIP6 signal,
calendar mapping, humidity closure, and output policy remain epwshiftr
components.
EPW fields outside dry-bulb temperature and its coupled humidity state remain unchanged. After dry-bulb temperature is projected, baseline specific humidity is retained and relative humidity and dew point are recomputed; moisture is clipped only when necessary to avoid supersaturation.
daily_temperature(
reference = NULL,
window_days = 31L,
reconstruction = c("power", "btws")
)A required historical_reference(),
shift_reference_plan(), or extracted ShiftClimate stage.
Odd circular climatology-window width in days.
Hourly temperature reconstruction component:
"power" uses the default constrained power projection; "btws" uses
bounded temperature weighted stretch and requires paired daily
tasmin and tasmax.
A complete ShiftMorphMethod for shift_future_epw().
Eames, M. E., Ramallo-González, A. P., and Wood, M. J. (2024). A revised morphing algorithm for creating future weather for building performance evaluation. doi:10.1177/01436244231218861