Identification of extreme precipitation events in southeastern South America from their associated synoptic environment

Southeastern South America (SESA), delimited between 38°S–25°S and 64°W–51°W, is characterized as one of the regions in the world with the highest frequency of occurrence of intense storms associated with deep convection, mainly during the spring and summer months (Zipser et al., 2006). These convective storm systems induce extreme precipitation events and produce most of the rain in the warm season (Rasmussen and Houze, 2016), generating significant damage (floods, intense winds, hail) and have a high impact on economic and social activities. Considering that the occurrence of extreme precipitation events in SESA is associated with the occurrence of certain synoptic patterns, the objective of this work was to detect the occurrence of extreme precipitation events from the synoptic patterns that induce these events, in spring (October to December, OND) and summer (January to March, JFM).

Daily data from the ERA5 reanalysis was used to detect recurring synoptic patterns associated with extreme precipitation events in the 1979-2013 calibration period. In order to identify a variety of precursor synoptic patterns of extreme precipitation events, the classification obtained from the principal component methodology (PCA) in orthogonally rotated T mode was used (Huth, 2000). To carry out the classification, the geopotential height was used at the 850 hPa level of the day prior to the occurrence of the extreme events (detected from the 95th percentile of the distribution of daily precipitation of the CPC Global Unified Gauge-Based Analysis of Daily Precipitation from the NOAA Climate Prediction Center). This classification resulted in two dominant synoptic situations for spring and summer. With the days obtained for each main component, compositions of the anomalies were made: the meridional component of the wind at the 850 hPa level, geopotential height at 850 and 500 hPa, and 200 hPa wind.

Based on the compositions made, an analog method was developed that was used to detect the occurrence of intense precipitation events in the verification period 2014-2021. In this methodology, two detection criteria were used, on the one hand, that the correlation coefficient between the fields of daily anomalies and the compositions are greater than a threshold and also that there is consistency between the sign of the daily meteorological variables with that of the compositions on certain grid points called hotspots regions.

For the evaluation of the analogue method, the F1 index developed by Gao et.al 2017 was used, which takes into account the number of true positives (TP), false positives (FP; type I error), and false negatives (FN; type II error).