In order to assess the bias and variability of various approaches to estimating DMT, we can look at data collected at different frequencies. The finest resolution available is minute data, which we have available for a station in Sweden (although not collected at the standard 2m altitude above ground) and for several North American stations. Data sets will be made available at the course page.

Questions of interest include

- How well does the average of daily min and daily max approximate the daily mean temperature?
- What is the effect on bias and variability of different observational schemes (such as hourly, 3-hourly, 6-hourly etc.)
- Does daily min and daily max add value to measures of daily mean temperature, taken several times a day (such as the Swedish Ekholm-Modén formula as described in Nordli et al, 1996, Appendix II)
- Does it matter much when using the average of min and max whether you use actual min and max or min and max of hourly readings?
- If daily temperatures were described by a sine curve, the average daily temperature would indeed be the average of min and max. How well is daily temperature described by a sine curve?

Visby minute data for July Minute data from Red Oak, TN

NCDC five minute temperature data page

Here
are the Ekholm-Modén coefficients.