Variability and trends of extreme dry and wet seasonal precipitation in Argentina. A retrospective analysis

Abstract

RESUMENSe analiza la variabilidad de la precipitación estacional en sus valores medios y extremos para varias regiones de Argentina al norte de 39º S, empleando series mensuales de lluvia que se extienden desde 1860 hasta 2006. Las estaciones seleccionadas pueden considerarse representativas de diferentes regiones climáticas del país. El análisis se focaliza en la distribución de la lluvia, la tendencia estacional, los cambios en la varianza y los valores extremos, a fin de establecer la magnitud de la variabilidad climática estacional de las lluvias en una perspectiva histórica en el centro de Argentina. Los valores extremos se calcularon para una ventana móvil de 40 años definiendo estos valores para los niveles mayores/menores a los percentiles 80/20 (p = 0.80/0.20; 0.90/0.10; 0.95/0.05) de la distribución teórica gamma, seleccionada entre otras cinco distribuciones. Los parámetros de escala y forma de la distribución se calcularon empleando la estimación de máxima verosimilitud (MLE) y el método de remuestreo (bootstraping) para 1000 repeticiones. Las tendencias estacionales se calcularon para cada ventana, al igual que su significación estadística. El uso de una ventana móvil permitió establecer el periodo de cuatro décadas en el cual se verificó el mayor valor de la tendencia, ya sea positiva o negativa. De los resultados se desprende que hubo desplazamientos en las distribuciones de lluvia estacional así como cambios en los valores de retorno (RV) que se calcularon para una ocurrencia de una vez cada 5, 10 y 20 años. En general, las lluvias extremas de verano se incrementaron para toda la región. Los valores extremos húmedo/seco (W/D) de la estación estival correspondientes a los percentiles 90/10 para W/D, respectivamente, sufrieron incrementos significativos de acuerdo con la zona geográfica y el periodo. Sin embargo, se menciona como rasgo particular del verano la existencia de un periodo común a todas las estaciones con tendencias positivas, que se inicia con la ventana 1921-1960 y permanece hasta la última ventana analizada: 1967-2006. Este comportamiento no se observa en la estación de invierno para el oeste y centro de Argentina excepto en la zona representada por Bahía Blanca, para la cual el extremo D (del 10%) se incrementó a lo largo de todo el periodo de estudio. ABSTRACTThe variability in seasonal mean and extreme precipitation is analyzed for several regions of Argentina to the north of 39º S, using long-term monthly time series data which expand from 1860 to 2006. The selected locations can be considered as representative of different climatic regions. This work focuses on the analysis 4 B. Scian and J. Pierini of monthly rainfall distribution, significant seasonal trends, changes in variance and extreme monthly values, in order to establish the magnitude of the seasonal climatic rainfall variability through time for central Argentina. A 40-yr moving window was employed in order to analyze seasonal variability of rainfall extremes. Extremes were computed for different probability levels of ...

Multi‐century tree‐ring precipitation record reveals increasing frequency of extreme dry events in the upper Blue Nile River catchment

“…Percentiles from inferred frequency distributions have been used in different studies to characterize extreme-dry and extreme-wet years (Easterling, 2000;Knapp et al, 2015Knapp et al, , 2016Scian & Pierini, 2013;WMO, 2016). Thus, we defined the threshold values for extremity using a percentile classification approach as described in World Meteorological Organization reports (WMO, 2013(WMO, , 2016.…”

“…Thus, we defined the threshold values for extremity using a percentile classification approach as described in World Meteorological Organization reports (WMO, 2013(WMO, , 2016. Extremedry and extreme-wet years were defined as those with total June-September rainfall <10th percentile and >90th percentile of the total June-September rainfall (Easterling, 2000;Knapp et al, 2015;Scian & Pierini, 2013). Dry and wet years, however, were defined as those with total June-September rainfall between the 10th and 33rd percentile and between 66th and 90th percentile.…”

“…Extremedry and extreme-wet years were defined as those with total June-September rainfall <10th percentile and >90th percentile of the total June-September rainfall (Easterling, 2000;Knapp et al, 2015;Scian & Pierini, 2013). Total June-September rainfall amounts within the threshold values between the 33rd and 66th percentile were classified as normal rainfall periods (Knapp et al, 2015;Scian & Pierini, 2013;WMO, 2016). Total June-September rainfall amounts within the threshold values between the 33rd and 66th percentile were classified as normal rainfall periods (Knapp et al, 2015;Scian & Pierini, 2013;WMO, 2016).…”

“…Dry and wet years, however, were defined as those with total June-September rainfall between the 10th and 33rd percentile and between 66th and 90th percentile. Total June-September rainfall amounts within the threshold values between the 33rd and 66th percentile were classified as normal rainfall periods (Knapp et al, 2015;Scian & Pierini, 2013;WMO, 2016). In addition, the returning periods for events including below/above-average rainfall, wet/dry and extreme-wet/dry years were estimated as the number of years between the end and the beginning of two consecutive events (Martin-Benito, Ummenhofer, Kose, G€ uner, & Pederson, 2016).…”

Regionalization of climate over the Argentine Pampas

“…Frontal rains are generated due to latitudinal and longitudinal movements of air masses. Natural cycles in the atmosphere can affect the climate of the region for weeks, months and decades (Scian et al , ; Scian and Pierini, ). El Niño Southern Oscillation (ENSO) causes inter‐annual rainfall variations with more intensity during autumns and summers (the rainy season in South America) (Grimm, ).…”