Climate scenarios and impacts of droughts and other hydrometeorological events in the Central American dry corridor

Abstract

Predicting rainfall during April-May-June (AMJ), as the first peak of the rainy season in the Central American isthmus, is very important since it has been observed that more or less humid conditions during AMJ tend to be preceded by early or late starts of the rainy season. A late onset of rainfall, for example, followed by drier than normal conditions during MJ and by a subsequent period of intense summer or heatwave, can significantly affect key socioeconomic sectors in the isthmus such as hydropower generation, drinking water supply or agriculture. In this presentation, data from 162 rainfall stations were used to construct predictive models for MJ as the first peak of the rainy season, using Canonical Correlation Analysis (CCA). The aspects to predict during MJ are rainfall accumulation and the Normalized Precipitation Index (NPI) in Central America. The sea surface temperature (SST) anomalies observed in the 63°N - 10°S and 152° E - 15°W domain were used as predictors. The CCA models, using the SST anomalies in February, show a good predictive ability of the accumulations and the NPI during MJ, in an important region of Central America. The results showed that warmer (cold) conditions in the eastern equatorial Pacific SST anomalies, along with colder (warmer) conditions in the North Tropical Atlantic during February, tend to be correlated with drier (wet) periods during the next MJ bimester in virtually the entire isthmus. This suggests that the SST could modulate rainfall during MJ in Central America by influencing the position of the Inter-Tropical Convergence Zone and the magnitude of trade winds.