@eff It could be useful to pre-profile various locations where mobile CSS might be deployed in the future (such as near government buildings, on university campuses, etc) in order to construct baseline profiles of "normal".
Then we can develop a tool that looks at the baseline of an area and compare it with what is detected on the day of an event.
Entire cities could be baselined in this way by having people drive on patterns, collecting data, in order to map the baseline for an entire city.
(Back in the 1970's we did this kind of baselining, on foot, of parts of San Francisco, to make maps of land use etc.)
By-the-way, there is already legally data, such as maps showing the RF footprints of legitimate cell towers (and their directional radios) and data about call events.