Variations of the global sea level pressure (SLP) field reflect atmospheric and oceanic influences and have a profound influence on temperature, precipitation and the global carbon cycle. The impact of various forcing factors on this field was investigated mainly based on numerical simulations. Alternatively, here we identify and quantify the influences of various forcing factors on observational, reanalysis and simulated SLP fields. By applying canonical correlation analysis (CCA) on the aforementioned data sets, we separated and quantified the impact of increase CO2 concentration, El Niño–Southern Oscillation (ENSO), Atlantic Multidecadal Oscillation (AMO), Arctic Oscillation (AO) and solar forcing on the global SLP field, based on their associations with known footprints on the sea surface temperature (SST). Together, their corresponding SLP spatial structures explain ~ 60% of the observed variance. Whereas the atmospheric CO2 concentration has the most prominent impact on the global SLP field, explaining 28% of variance, ENSO and AO account for 9% each. The solar forcing and AMO explain 7%, respectively 6% of global SLP variance. Similar spatial structures corresponding to the same forcing factors are identified based on the reanalysis SLP data. CCA applied on simulated SLP fields derived from six CMIP5 model simulations captures only the spatial structures of atmospheric CO2 concentration, ENSO, AAO and AO. Such a decomposition of the global pressure field based on a linear combination of coupled SST-SLP pairs provide a reference against which one could validate the performance of general circulation models in simulating the lower atmosphere dynamics.