From what I saw of Google Scholar previously, they have quite a flexible model of what is considered a citation, in comparison to something like Web of Science. The result of this is that stats such as h-indexes of individual researchers may be slightly higher as measured by Google Scholar than by WoS for instance.
You could argue that WoS is giving a stricter interpretation based on more curated results and 'gold standard' citations only, or you could argue that Google Scholar is more representative of true impact, considering a wider variety of impacts.
If two journals relative ranks differ between the two methods, then one possible interpretation might be that one journal has a substantially wider impact outside of strict peer-reviewd citations than the other.
Another possible interpretation might be temporal. The two journals' impacts may be on different trajectories (one up, one down, for instance), and Google Scholar may pick up such signals earlier than more formal curated indices.