Are total volcanic emissions worldwide constant, or do they rise and fall? We know there are large fluctuations on a scale of tens of millions of years, but is the total volume of emissions stable on a scale of tens of thousands, the last ten thousand in particular? The last major increase that I know of was shortly before the extinction of the dinosaurs and according to some geologists was the cause of the extinctions at the end of the Cretaceous, but I suspect there have been further significant episodes of volcanism since then.
If you have access, I would recommend you to read this short chapter of the Encyclopedia of Volcanoes. In summary:
- At ocean ridges, studies show a steady rate of oceanic crust production during the Cenozoic. There have been fluctuations before, maybe due to the breakup of Pangea or other major tectonic events. The present volcanic production rate is ~20 km$^3$ per year.
- At volcanic arcs, it is harder to constrain because of the poor preservation of volcanic material. Tephra layers found in deep sea drilling cores show some peaks in explosive activity "in the Quaternary (0-5 Ma), the middle Miocene (c. 13-17 Ma), and potentially the Eocene (c. 37-42 Ma)". This is not well explained. There is also a shorter term variability due to deglaciation episodes: mass unloading depressurize magmatic systems. A conservative estimate of the modern output is 0.5 km$^3$ per year.
- Finally, Large Igneous Provinces (LIPs) are very episodic (long-term average of one LIP every 10-20 Ma) but they emit a large volume (generally > 1 million km$^3$) in a relatively short period of time (a few million years). LIP emplacement coincide with major mass extinction events.
My thinking is that average scale of volcanic eruptions and resultant emissions should be the most stable on the scale of hundreds and thousands of years.
- Yearly and decade scale: Major variations where a single large eruption will result in a year being an outlier
- 100-1000 year scale: Generally quite consistent. No major changes in areas of volcanic activity and random outliers won't have a drastic effect on an average activity
- 10,000-100,000 year scale: This is the time scale close to the average life time of a volcano within a subduction zone due to movement of plates and source of magma moving away from the current active area. As a result on this time scale we should expect some older volcanoes to become dormant and new ones to pop up but there won't be much consistency
- Scale of millions of years: Major changes in geometry and activity of subduction zones, potentially new areas of rifting opening up, so overall not very stable