On the surface, I'd say no. You seem to have Incidents as the middle of your design. In other words, many things cannot exist in the database without an incident. That is not the way the real world works.
One example from your design: you have a table of resources. In the real world, a resource may be a patrol car or a tool or a police officer. Any one of those things can exist in the real world all by themselves, without any reliance on an external entity (table). However, in your design, a single incident in time and space helps to define a resource. I don't know of any patrol car that is defined by an incident (except maybe the event in the factory that built it). Also, this design doesn't allow a resource to "cooperate" in more than one incident, though the same incident may help to define the existence of one or more resources. So, the incident of John Hinckley (sp?) shooting President Ragan may be responsible for defining a patrol car, a pistol, and a bullet proof vest. However, those three items can no longer take place in another incident again, according to your design.
This is besides the broken rules of normalization. There are too many fields that "could" be null to come anywhere close to 1NF according to one definition.
Also, you have some 1:1 relationships defined. 1:1 are rare and the only good time I know of using them is to implement polymorphic behavior in your database (as in C++/Java classes or interfaces).
Your design seems to be saying that there cannot be an "event" without there being a "weather".
You also have a lot of orphaned tables. I'm sure that a "narrative" is in some how related to some other entity in the database (possibly many) but there is no indication of this.
Hope this helps.
