Abstract The Antarctic Peninsula is one of the regions in Antarctica that has experienced notable and extensively studied warming since the mid-twentieth century. Meteorological data, although limited and mostly dating back to the International Geophysical Year (1957–1958), reveal dramatic climate changes in both the Antarctic Peninsula and West Antarctica, with the Antarctic Peninsula showing the highest warming trends on the continent. One major manifestation of this warming is the increasing frequency and intensity of extreme warm events. This study utilizes temperature data from the Artigas Antarctic Scientific Base provided by National Weather Service, Uruguayan Institute of Meteorology (INUMET), which have not previously been used or quality-controlled. Data were processed and analysed, and a quality hourly database was built for the period 1 January 1998 to 11 December 2016 for the surface temperature, constructing a complete time series interpolating the data with two other nearby Antarctic stations (C.M.A. Eduardo Frei Montalva and King Sejong). Temperature presented an annual cycle with marked interannual and seasonal variability, with the cold season being the one with the highest variability and the largest anomalies, both positive and negative. No significant trend was found for the monthly mean temperature. A study of warm temperature events was conducted at three points on King George Island, defining warm events and extreme warm events as those in which the mean daily temperature was above the 90th and 99th percentile, respectively, for at least 3 consecutive days. A high frequency of occurrence was found (all years except 2015) mostly in the cold season of the year, and with large interannual variability. In turn, it was found that certain atmospheric dynamics favour the generation of these extreme warm events, and that their occurrence is higher during La Niña years. Although the majority of the extreme warm events occurred during positive phases of the Southern Annular Mode, the statistical correlations were not significant.
Balay et al. (Mon,) studied this question.