Long-term variability of heat waves in Argentina and recurrence probability of the severe 2008 heat wave in Buenos Aires
Abstrakt
Heat waves are one of the main concerns related to the impacts of climate change, because their frequency and severity are projected to increase in a future climate. The objectives of this work are to study the long-term variability of heat waves over Argentina and to estimate recurrence probability of the most severe 2008 heat wave in Buenos Aires.
We used three definitions of heat waves that were based on (1) daily maximum temperature above the 90th percentile (MaxTHW), (2) daily minimum temperature above the 90th percentile (MinTHW) and (3) both maximum and minimum temperatures above the corresponding 90th percentiles (EHW). The minimum length of a heat wave was 3 days, and the analysis was performed over the October-March period.
Decadal values in Buenos Aires experienced clear increases in heat waves according to MinTHW and EHW, with the highest frequency for both in the 2001-2010 decade, but at other stations, combinations of different trends and decadal variability resulted in some cases in a decrease of extreme heat waves. In the north-western part of the country, a strong positive change in the last decade was found, mainly due to the increment in the persistence of MinTHW but also accompanied by increases in MaxTHW.
In general, other stations show a clear positive trend in MinTHW and decadal variability in MaxTHW, with the largest EHW cases in the last decade. We also estimated recurrence probability of the longest and most severe heat wave in Buenos Aires (over 1909-2010, according to intensity measured by the cumulative excess of maximum daily temperature above the 90th percentile) that occurred from 3 to 14 November 2008, by means of simulations with a stochastic first-order autoregressive model.
The recurrence probability of such long and severe heat wave is small in the present climate but it is likely to increase substantially in the near future even under a moderate warming trend.