DATASETS
Colocation Maps
Colocation Maps measure the probability that two individuals from two locations are found in the same location at the same time. These maps help researchers analyze the probability that people in areas with disease outbreaks will come in contact with new populations.
Request Access
Explore CMMID's UK Colocation visualization
Key Features
Granular estimates
Colocation Maps are typically constructed at the scale of counties or local equivalents.Frequent updates
Colocation Maps predict colocation events as short as five minutes and are refreshed weekly, providing a window into how population mixing patterns are evolving in near-real time.Epidemiologically focused
These maps are critical in helping epidemiologists understand the probability that a disease will be spread from one region to another by human-to-human contact.
Methodology
Step 1: Define area of interest
We define an area of interest based on the pandemic area or research goals.Step 2: Identify home area
We then assign individuals to home administrative regions (counties in the USA) based on their typical location.Step 3: Analyze location trajectories
We simplify the GPS data of users over both time and space. We divide each week into 5-minute time intervals. We split the surface of the earth into grid tiles roughly 600 meters on each side.Step 4: Calculate colocation events
We count the number of colocation instances where two users are present in the same grid tile during the same 5 minute window. We then aggregate all colocation events over a week to the regional level (counties in the USA).Step 5: Calculate colocation probability
We surface the probability of colocation between regions by looking at the number of observed colocation instances over the total number of potential colocation events in a week.RESOURCES
Using Colocation Maps
Colocation Maps are available to nonprofits and researchers who sign data sharing agreements.
Request Access
RESOURCES