Characterizing monthly temperature variability states and associated meteorology across southern South America

Key spatiotemporal patterns of monthly scale temperature variability are characterized over southern South America using k-means clustering. The resulting clusters reveal patterns of temperature variability, referred to as temperature variability states. Analysis is performed over summer and winter months separately using data covering the period 1980-2015. Results for both seasons show four primary temperature variability states. In both seasons, one state is primarily characterized by warm temperature anomalies across the domain while another is characterized by cold anomalies. The other two patterns tend to be characterized by a warm north-cold south and cold north-warm south feature. This suggests two primary modes of temperature variability over the region. Composites of synoptic-scale meteorological patterns (wind, geopotential height, and moisture fields) are computed for months assigned to each cluster to diagnose the driving meteorology associated with these variability states. Results suggest that low-level temperature advection promoted by anomalies in atmospheric circulation patterns is a key process for driving these variability states. Moisture-related processes also are shown to play a role, especially in summer. The El Nino-Southern Oscillation and the Southern Annular Mode exhibit some relationship with temperature variability state frequency, with some states more common during amplified phases of these two modes than others. However, the climate modes are not a primary driver of the temperature variability states.