Atmospheric water vapour is shaped by both sustained global warming (GW) and El Niño–Southern Oscillation (ENSO), but how these two climate influences are expressed in the water-vapour field and how their coupling changes under accelerated warming remain poorly quantified. Here we comprehensively characterize the links between global GNSS-derived precipitable water vapour (PWV), a vertically integrated measure of atmospheric moisture, and both GW and ENSO by integrating Independent Component Analysis (ICA) and Convergent Cross Mapping (CCM). Warming-related and ENSO-related PWV signals are extracted from the GNSS PWV field using ICA and are denoted as <em>PWV<sub>GW</sub></em> and <em>PWV<sub>ENSO</sub></em>, respectively. <em>PWV<sub>GW</sub></em> is closely associated with GW (<em>r</em> = 0.79), and CCM supports dominant <em>GW</em> → <em>PWV<sub>GW</sub></em> coupling. By contrast, <em>PWV<sub>ENSO</sub></em> responds to ENSO with an approximately 12-month lag (<em>r</em> = 0.77), and extended CCM indicates bidirectional coupling between ENSO and <em>PWV<sub>ENSO</sub></em>. After the transition to accelerated warming around 2013, the sensitivity of <em>PWV<sub>GW</sub></em> to temperature increases markedly, consistent with enhanced thermodynamic moistening. At the same time, the <em>ENSO</em> → <em>PWV<sub>ENSO</sub></em> coupling weakens despite stronger ENSO variability. These findings suggest that accelerated warming shifts the balance between long-term thermodynamic moistening and interannual ENSO-related variability in the atmospheric water vapour field.