Precipitation Variations Using Satellite Data during the Recent Sea Surface Temperature Warming Period in the Intra-Americas Region, 1982–2012

A recent warming trend of sea surface temperatures (SST) has been detected in the Intra-Americas Region (IAR) for the past 30 years (1982-2012), with the warmest trends occurring during the latter half of the period. Warming SSTs in the IAR, defined as the geographical region that includes the Caribbean, Mexico, Central America and parts of North and South America, can have potential implications for Tropical precipitation variation. This study investigates precipitation trends for a similar period (1982-2014) to determine variations as a result of warming SSTs. The IAR experiences distinct, bimodal precipitation periods, which define the seasons – the Early Rainfall Season (ERS) from April to June, the Late Rainfall Season (LRS) from August to November and the Dry Season from December to March. Increases in SSTs during the LRS are of particular importance because temperatures during this season reach 26°C and above, which is the threshold for deep convection. Within this temperature range (26°C – 29°C), rainfall dependence on SSTs increases by a factor of 5 and the atmosphere becomes the principal modulator of thermal convection. In this way, SSTs influence Tropical precipitation.

The precipitation analysis, using the Global Precipitation Climatology Project (GPCP) monthly precipitation dataset and selected ground stations, shows that the past 15 years (1997-2012) average precipitation for the LRS has been slightly above the climatology calculated from the years 1982 to 2012. Results suggest that the observed increases in SSTs may be connected to the changes in precipitation for the same period. SST daily anomalies reflect a warming trend over the past 15 years, which further suggests the significant influence that SSTs potentially have on precipitation. Preliminary analyses show that during the ERS, warming SSTs in the region north of South America and above the threshold for vertical convection were found to be highly cross-correlated (0.79) with precipitation. For the LRS, cross-correlation with precipitation and the SSTs near South America also show a high, positive cross-correlation (0.78), similar to ERS results.

Linear and nonlinear analyses of precipitation trends, using the 5% significance level, reveal that regionally averaged accumulated rainfall is decreasing for all seasons except the LRS, but were not determined to be statistically significant for the 30-year period. However, per-grid analyses show that accumulated rainfall per season has been increasing in the southern part of the region, with high statistical significance primarily for the LRS. Conversely, the Dry Season (DS), from December to March, reflects the steepest decreasing trends associated with statistically significant p-values. Furthermore, the areas determined to be statistically significant are primarily over ocean versus trends observed over land. Future work includes characterization of sub-regions within the IAR based on SST trends, which will be tested for correlations with precipitation trends corresponding to the given sub-region. Regional correlation results will be compared to sub-regional results and large-scale climate phenomena will be considered, including the El Niño Southern Oscillation (ENSO) and the Atlantic Multidecadal Oscillation (AMO).