One thing we've got in temporal bias right now is the link with the source of reporting for each event. It could be interesting to take a few hazards types (some they've been tracking a long time, like tornadoes, then some intermediate, like hail, then some newer ones, like one added in 1996) and see for each how the distribution of sources for events has changed with time. I'm thinking something like the plots shown here: https://www.r-graph-gallery.com/136-stacked-area-chart.html
This way we could both show that the number of reports of that event type has increased with time and also that (maybe) the distribution across different sources has also changed.
I suggest we try this starting with tornadoes to see what it looks like. We can talk about how we could write the code to check this when we meet.
One thing we've got in temporal bias right now is the link with the source of reporting for each event. It could be interesting to take a few hazards types (some they've been tracking a long time, like tornadoes, then some intermediate, like hail, then some newer ones, like one added in 1996) and see for each how the distribution of sources for events has changed with time. I'm thinking something like the plots shown here: https://www.r-graph-gallery.com/136-stacked-area-chart.html
This way we could both show that the number of reports of that event type has increased with time and also that (maybe) the distribution across different sources has also changed.
I suggest we try this starting with tornadoes to see what it looks like. We can talk about how we could write the code to check this when we meet.