African scenarios and map fixes

[nicolepaul]

• Clean up and rerun the previously included African scenarios: 2003 Boumerdes (Algeria), 2004 Al Hoceima (Morocco), 2016 Bukoba (Tanzania), 1992 Cairo (Egypt)
• Integrate content from the 2003 Marrakesh-Safi (Morocco) from the marrakesh branch that @raoanirudh was working on; add new data available from USGS (finite fault, stations)
• Take 1989 Salima (Malawi) out of 'DRAFT' mode (unit test failure due to Tests fail when USGS doesn't have any station data #26)
• Correct some broken code and maps in the src functions (e.g., addreses Stamen tiles no longer functional in contextily #23)
• Added 2009 Karonga (Malawi) (unit test failure due to Tests fail when USGS doesn't have any station data #26)

[nicolepaul]

Some notes about the 2023 Marrakesh-Safi event, which likely could be improved.

• USGS now has 3 seismic stations (see here). One of these stations has a PGA of ~0.79g, which can significantly increase out estimated ground shaking when we condition to the seismic stations.
• Previously, a point rupture from USGS was used, but now there is a finite fault estimate. Should we consider keeping both for reference? This version only includes the finite fault estimate and not the point rupture.
• In Slack, @raoanirudh shared some fatality estimates by Adm2, but I didn't add those since I didn't know the source
• I've seen some presentations that included impact at the subnational level, but would need to reach out to the presenters to find the source

[CatalinaYepes]

LGTM