How to Work with Temporal Data in ArcGIS Enterprise?

[Nadia_Matsiuk]

Let’s consider an example of the Police Department Dashboard (https://www.arcgis.com/apps/dashboards/9ef296f66f724c36bcaf01fc69768ecd).

We have a point layer containing crime incidents. The data updates every hour and includes attributes such as location, date, district, incident type, etc.

My goal:
I want to create analytical charts that go beyond just showing individual incidents and instead provide summarized insights, such as:

• the average number of incidents over a given period;
• distribution by crime type;
• crime rate trends over time.

I have seen charts like Last 14 and 15-28 Day Comparison in the dashboard. How are they built?

Specific questions:

• How should I prepare data for such charts in ArcGIS Enterprise?
• Should the data aggregation be done at the database level?
• Or can this be handled in ArcGIS Pro or other tools?
• How can I make charts automatically update based on the user-selected time filter (e.g., last 3 days, last month, or a custom range such as 01/01/2025–01/21/2025)?

From what I understand, my chart will be using a precomputed table with already aggregated data.
I’m not sure how this setup would work with time-based filtering. Is this the right approach, or am I missing something?
What are the best practices and tools in ArcGIS for implementing this?

Sorry, my wording might not be entirely clear. This topic is quite complex for me, and I’m trying to understand it as thoroughly as possible. Any relevant help or advice would be greatly appreciated.🙏

[MuneerMajid]

You need a historical layer for the incidents in your map. You already mentioned that your data has attributes like location, timestamp, etc. 

Within your Dashboard, you can play around with the infographic of your choice, lets say a list or a chart. Adjust the filters on the timestamp to get the analytics you wish to display. Its difficult to accurately answer your questions without looking at the data model, but my tried and tested approach is to always reverse engineer you way from the desired output, back to the right data model. Start with answering what your goal is for the dashboard, and work your way back to what kind of data you need to curate with your feature layers. If you only have aggregated data over time in your layer, you wont be able to slice out specific info in your dashboard like count of specific crime incidents by zip/county over a time period. So try to get your data cleaned and prepped to a level where a filter over timestamp AND crime type etc. can get you the desired results.