Why does the Dakar series show such high volatility — swings of 20-30% year-on-year — while other West African cities are much steadier? The volatility could reflect genuine underlying variation, or it could indicate that Dakar's data comes from a less stable source or uses a less standardized collection method. Should we add a data-quality assessment or investigate the source?
The 2009 break in population
Looking at the monthly data, January always shows a spike, March always shows a dip, and summer shows a valley—these are such strong seasonal patterns that deviations from them stand out. But if the seasonality changes (e.g., from 20% deviation to 25%), is that a real change or just noise around a persistent pattern? Should we surface the seasonal factors explicitly? See /u/plg_hugo_marende/p/plot-0000.
Will pick this up after the refresh.