Global w​arming intensifies atmospheric water vapor transport between ocean and land, which increases the likelihood of extreme precipitation and floods. However, accurate estimations of water vapor exchange between ocean and land are difficult due to the lack of available data and effective methods. This study developed a novel eight-direction-vector decomposition algorithm for calculating water vapor flux between ocean and land based on the ERA5 reanalysis dataset, and the results showed that global water vapor exchange between ocean and land had significantly increased in the past 40 years, except for Antarctica. During 1980--2018, the average annual net water vapor inflow from ocean to land (Q_net) was 44.68×10¹⁵ kg/yr and Q_net increased at a rate of 1.48×10¹⁵ kg/yr per decade. The intensified atmospheric water vapor exchange between ocean and land was directly caused by the increase of atmospheric water vapor content, which largely depended on the rising air temperature, and it was found that water vapor flux between ocean and land increased by over 8%/K with the increasing air temperature at the global average. This study also identified El Niño-Southern Oscillation (ENSO) as an important contributor for the global ocean-land water vapor exchange anomalies. A strong El Niño event (MEI=1) can result in a 1.36×10¹⁵ kg/yr (3.03%) decrease in Q_net, and a strong La Nina event (MEI=-1) can increase Q_net by 1.38×10¹⁵ kg/yr (3.09%). The eight-direction-vector decomposition algorithm was effective in the ocean-land water vapor flux estimation and can be used at different spatial and temporal scales, which can provide great insights on the mechanisms of extreme precipitation events.