The only meteorological data available for Greenland in the 18th century come from its SW part, and more specifically from the area where the present capital of Greenland, Nuuk (older used names: Godthaab or Godthab) is located. The dataset contains three available series of meteorological observations: (1) for Neu-Herrnhut (1 Sep 1767 to 22 Jul 1768) and two for Godthaab (2) (Jan-Jun 1787) and (3) (Sep 1784 to Jun 1792). The first series is the oldest long-term series of instrumental measurements of air temperature. In addition, the weather register (Moravian Archive in Herrnhut, catalogue number R.15.J.a.13.) provides more measurements such as wind direction (from 8 directions) and force (on a scale from 1 to 6), as well as a very short weather description. Meteorological observations were made by Christopher Brasen (1738-74) usually two times a day – at 8 a.m. and 2 p.m. The second series of measurements, although not continuous, is the greatest and most reliable available for Greenland for the study period. Observations were made three times a day (7 a.m., 2 p.m. and 9 p.m.) by the Danish Reverend Andreas Ginges (1754–1812) using a methodology and instruments provided by the Meteorological Society of the Palatinate. The sub-daily or daily air temperature data exist for the following periods: Sep 1784 to Jun 1785, Jan-Jun 1787, Nov-Dec 1788, Jan 1790 to Jun 1792 and are available in the manuscript entitled “Astronomiske og meteorologisk Iagttagelser, anstillede i Godthaab i Gronland 1782–1792” and in the society’s yearbook Ephemerides Societatis Meteorologicae Palatinae, which contains data only for 1787. All available historical data were quality controlled and corrected. For example, air temperature in the period 1767–68 was measured using a Fahrenheit thermometer, and we therefore converted all original measurements to the presently used unit, i.e. degrees Celsius. Earlier measurements in this time were made mainly at 8:00 and 14:00, but there were also days when three measurements a day were taken (the third at 22:00 or 23:00). On the other hand, stable hours of measurements (7:00, 14:00 and 21:00) were used in the period 1784–92. As a result, we calculated mean daily air temperatures (MDAT) for the historical periods using the following formulas:
1. MDAT= (T8+T14)/2;
2. MDAT = (T8+T14+2*T22)/4;
3. MDAT = (T8+T14+2*T23)/4;
4. MDAT= (T7+T14 +2*21)/4.
Furthermore, the scale of the influence that the different measurement times had on the values of MDAT was checked using hourly air temperature data taken from 4250 Nuuk station for the contemporary period (2001–10). Corrections needed to obtain “real” MDAT (calculated according to formula MDAT= (T1+T2+T3+ … T24)/24) were calculated for each month separately.