shape_calibration: Calibrates the shape of the night interpolating curve
In Interpol.T: Hourly interpolation of multiple temperature daily series

Description Usage Arguments Value Note Author(s) References See Also Examples

Calibrates the shape of the night interpolating curve, either horizontal-axe parabola or line, by changing the exponent z (see reference). It functions according to the comparison of the daily thermal range and the climate (reference) monthly one.

  shape_calibration(meas, date.format = "ymd",
    cal_times_list, band_min = 0:23, band_max = 0:23,
    ratio_dtr_range = c(0, 6), nr_cycles = 10,
    min_mo.length = 21, full.24.hrs.span_min = TRUE,
    silent = FALSE)

`meas`	measured hourly values file (table), where the first column is the series' ID
`date.format`	input date format (formats for function `chron`)
`cal_times_list`	calibration list of "time" parameters (output of `par_calibration`)
`band_min`	band of hours of occurrence of day minimum in the daily series (continuous). See `Note`
`band_max`	same for maximum time
`ratio_dtr_range`	range for seeking the optimal value of `ratio_dtr`
`nr_cycles`	number of calibration trials within the calibration ranges (all)
`min_mo.length`	minimum number of days to calculate any monthly values of dtr (is passed to function `Mo.Th.Ra.`)
`full.24.hrs.span_min`	logical, if set to `FALSE` does not allow to shift minimum time to the late hours of the day
`silent`	logical, if set to `TRUE` suppresses any warning issue

A list containing the optimum values of ratio_dtr

meas must be organized as 4-field records, all series in the same file, no headers. Column order: station ID, date, time (hour), T, [others fields, if any...] separated by spaces. This field order is mandatory.

Default date format is "ymd" (yyyy/mm/dd). Different combinations can be passed to function with date.format, but separator must be "/"

band_min and band_max are the time bands according to which the minimum and maximum temperature were calculated in the daily series to be interpolated. In general, they range from 0 to 23, unless the series has had some restriction in the calculation of minimum and maximum values. Hence, these bands can be different from those used to calibrate the most frequent occurrence of min and max.

The optimal value of ratio_dtr (k, eq. 7, in the quoted reference Eccel (2010a)) is chosen as the one with the (absolute) minimum value of the bias (irrespective of its sign). ratio_dtr is the ratio between the daily thermal range of the day to be interpolated and the mean monthly value for that series. The corresponding values of mean absolute error and RMSE can be checked in the resulting list.

min_mo.length (passed to function Mo.Th.Ra.) refers to the sum of days along all the series for each specific month, not for any single month in one year.

full.24.hrs.span_min is TRUE as default. If must be set to FALSE only if minimum values of the daily series have been calculated on a restricted time band, which is included in band_min (see function par_calibration). If this is the case, the minimum of the interpolated curve will always fall within band_min (early hours of the day). If this option is erroneously chosen, errors as large as 0.6 deg C can arise in the average of mean daily T.

Emanuele Eccel, Emanuele Cordano emanuele.eccel@iasma.it

Eccel, E., 2010: What we can ask to hourly temperature recording. Part II: hourly interpolation of temperatures for climatology and modelling. Italian Journal of Agrometeorology XV(2):45-50 http://www.agrometeorologia.it/documenti/Rivista2010_2/AIAM%202-2010_pag45.pdf,www.agrometeorologia.it

Original algorithm from: Cesaraccio, C., Spano, D., Duce, P., Snyder, R.L., 2001. An improved model for determining degree-day values from daily temperature data. Int. J. Biometeorol. 45: 161-169. http://www.springerlink.com/content/qwctkmlq3tebthek/

See also: Eccel, E., 2010: What we can ask to hourly temperature recording. Part I: statistical vs. meteorological meaning of minimum temperature. Italian Journal of Agrometeorology XV(2):41-43. http://www.agrometeorologia.it/documenti/Rivista2010_2/AIAM%202-2010_pag41.pdf,www.agrometeorologia.it

par_calibration, Th_interp

library(Interpol.T)
data(Trentino_hourly_T)

stations <- c("T0001","T0010","T0129")

calibration_shape <- shape_calibration(meas = h_d_t[h_d_t$V1 %in% stations,],
						cal_times_list = calibration_l[stations],
                     band_min = 0:23, band_max = 0:23, ratio_dtr_range = c(0,4),
						min_mo.length=21)