Figure 2:

Schematic illustration of the factor c(t), representing the combined effects of NPIs, adaptive behavioural changes, seasonality and viral evolution on the reproduction number. L₁, L₂ and L₃ represent the known durations (and timings) of first, second and third lockdowns, respectively, as imposed by governments. The gray area illustrates the time period included in our analyses, with a note that the third set of restrictions was still in place when the data fitting period ended. T₀ is the number of days in the series prior to gradual contact reductions early in the pandemic (estimated). T₁ (> T₀) is the day first lockdown begins (informed by data). T₂ is the number of days for the ramp of increasing transmission after first lockdown to return to baseline c(t) = 1 (technically, this is estimated and used to define the slope of the linear increase rather than to imply that it will continue to follow the trend beyond the study period). Whether c(t) is allowed to increase beyond baseline (solid) or not (dashed) does not affect our results as we consistently find the factor to remain below baseline throughout the study period.