Rendition of the Atlantic Warm Pool in Reanalyses

The Atlantic Warm Pool (AWP) is located in the western tropical North Atlantic (TNA) and the Intra-Americas Sea (IAS) (the Gulf of Mexico and the Caribbean Sea). The AWP is an area of warm sea surface temperatures (SSTs) greater than 28.5°C that exhibits a strong seasonal cycle, with the AWP disappearing in boreal winter (December-January-February) and reaching a maximum in late boreal summer (August-September-October (ASO)). In addition, the AWP exhibits interannual variability, with large AWPs nearly three times larger in area than small AWPs. Because there is a lack of surface and subsurface observations in the IAS and the TNA, this study uses reanalysis products to examine the AWP. It is of interest to ask what is the quality of these reanalysis products in their examination of the AWP. Thus, the main objective of this study is to examine and intercompare different reanalysis' renditions of the AWP. Three ocean reanalysis products are examined in this study and they are the following: (1) the Global Ocean Data Assimilation System (GODAS), (2) the Climate Forecast System Reanalysis (CFSR), and (3) the Simple Ocean Data Assimilation (SODA). In addition, because GODAS and SODA are not coupled to the atmosphere whereas CFSR is, two atmospheric reanalysis products are also used and they are the following: (1) the NCEP–National Center for Atmospheric Research (NCAR) reanalysis (hereafter R1), and (2) the NCEP–Department of Energy (DOE) Atmospheric Model Intercomparison Project (AMIP-II) reanalysis (hereafter R2). R1 is used in conjunction with SODA, and R2 is used in conjunction with GODAS. The period of interest for this study is 1980−2006. Each ocean reanalysis is used to examine the annual and interannual variability of the AWP in ASO. In addition, GODAS with R2, CFSR, and SODA with R1 are used to perform a detailed analysis of the SST tendency equation of the AWP. Furthermore, GODAS with R2, CFSR, and SODA with R1 are used to examine the impact of the AWP on the tropical tropospheric temperature and see if the impact is similar to the impact seen by the El Niño Southern Oscillation (ENSO). All three ocean reanalyses depict similar annual variability of the AWP. They all exhibit the AWP to be non-existent in boreal winter (December-January-February) and at a maximum in boreal summer. However, the reanalyses are inconsistent in their initial appearance of the AWP. All three ocean reanalyses depict similar interannual variability of the AWP; however, they are inconsistent in the relationship between the AWP area and subsurface temperatures. From the SST tendency equation, the fluxes contribute more to the AWP SST tendency in GODAS and CFSR, and the fluxes act to remove heat from the AWP. In SODA, the advective terms contribute more to the AWP SST tendency. Additionally, the reanalyses show that the AWP warms the tropical troposphere contemporaneously (ASO) but not three to six months after, as in the case with ENSO.