We could have the sources on the y-axis, years on the x-axis, and color could show number of reports from that source in that year. This might help with temporal bias to show how new sources get added (like social media and CoCoRaHS) as new technologies / systems develop.
We could either show for all hazards---in which case some temporal changes will result from more types of hazards being added to the database---or focus on one like tornadoes or hail that's been reported for a while.
Something kind of like this plot:
From here: https://benjaminlmoore.wordpress.com/2015/04/09/recreating-the-vaccination-heatmaps-in-r/
We could have the sources on the y-axis, years on the x-axis, and color could show number of reports from that source in that year. This might help with temporal bias to show how new sources get added (like social media and CoCoRaHS) as new technologies / systems develop.
We could either show for all hazards---in which case some temporal changes will result from more types of hazards being added to the database---or focus on one like tornadoes or hail that's been reported for a while.