The South America Forecast Model Wet Bias

Highlight: The South America forecast model wet bias.

Fig. 1-3: The ECM rainfall forecast bias across South America for the 15-day, 6-10-day, and 11-15-day period for the past 30 days.

Discussion: The wet forecast bias by all models across Brazil is worsening. The maximum error for the 1-5-day, 6-10-day, and 11-15-day period from the past 30 days is 3+ in. of rain (Fig. 1-3). The ECM is cited due to the worst verification although other models are very close, including GFS. The 30-day percent of normal rainfall pattern across South America identifies the mostly dry climate (Fig. 4). Dryness is particularly acute in Uruguay and vicinity plus Northeast Brazil. Most of Brazil and Argentina are drier than normal for the first third of meteorological summer. Consequently, soil moisture anomaly analysis reveals plenty of drought zones especially in Brazil and Argentina dry soils is strengthening (Fig. 5). A partial explanation is related to the global vertical velocity potential observed during DEC-24 (Fig. 6). The blue regions represent ascending air usually related to convection, especially in the tropics. In the tropics, ascending air is generally most likely across warm sea surface temperature anomalies (SSTA) and the presence of El Nino (in the tropical Pacific). Descending air is more common across cool SSTA and La Nina. During DEC-24, despite evolving La Nina, ascending air was widespread across the Pacific tropics and compensated for by descending air across Eurasia and South America. Descending air is subsidence which suppresses connective rains (in the tropics) causing a dry climate. The positive vertical velocity potential values across South America lead to a surprisingly dry climate during DEC-24 in South America poorly forecast by day-to-day operational models.

Fig. 4-6: The South America percent of normal rainfall during the past 30 days, daily soil moisture analysis and DEC-24 global vertical velocity potential observations.