I would be immensely curious to see the raw data given how the point spacing changes (assuming those aren't actual data measurements). The drop and recovery around the eye are so staggeringly smooth.
They are most likely smooth because there are so few data points. The red line has the peak, one data point halfway down, one at the bottom, one halfway up, and then the next peak. 'halfway' could be anywhere between 20% and 80% and it'd still look like a super smooth line.
Air pressure is less bad, but each data point is still moving a huge amount. If there was a lot of variance we'd have difficulties setting it at this resolution.
