Squirrels in the City

Audrey Leong
Data Mining the City 2022
7 min readMay 1, 2022


by Audrey, Stella, Taha

1. Research Report

AbstractThis project aims to highlight the impact of daily human activity on urban wildlife in the metropolitan city, with an emphasis on the interaction between squirrels and home-pets, as well as human amenities, all superimposed upon the ever-present seasonal shifts in nature. By zooming in on the southwest end of Central Park, a 3D simulation model of the squirrels’ movements can illustrate their behavioral patterns throughout different times of day and seasons. Based on the issues of coexistence between human and urban wildlife — that the squirrels’ foraging pattern largely overlaps with paths of human activity — this project aims to tweak human behavior by relocating human life facilities, such as trash cans, in order to better delimitate the squirrels’ area of activity and decrease the potential ‘aggressive’ interactions.

Introduction — In the metropolis of New York City, how we, humans, see and live in the city is only one entry point to observe urban life. If we move our focus on to the green spaces or the unobtrusive corners, it is intriguing to see that the urban wildlife life is also largely affecting the city. Central Park is the home to over two thousand squirrels, living off of the leftovers that humans throw in the trash can and water from the fountains. Their behavioral patterns between food and leisure also dependent on the fact that the home-pets must be taken care of and have their own constrained outdoor activity space, such as dog parks. In other words, the living situation of urban wildlife greatly relies on human’s activity patterns.

During the pandemic, while human activities were limited, the urban wildlife was also facing a critical living situation with less food and running public facilities. News on “Squirrels Attacking People” and “Squirrels Broke Into Human Residential and Stole Foods” came out regularly. The series of Squirrel Attack ended up with parks warning visitors not to feed the squirrels. However, once the ‘balanced’ life pattern of coexistence was broken, problems showed up on both sides of human and urban wildlife, and will slowly accumulate into more serious issues.

This research will analyze squirrel’s behavior throughout both long and short term periods of time in order to see how human interference and amenities affect the urban wildlife populations. By observing the time, place, and causes of these interactions, we can determine what steps need to be taken in order to preserve the compromise and coexistence between urban human life and urban wildlife.

2. Methodology

Population — The research process began with the actual data provided by the squirrel census, in order to get a general idea of how the squirrel population moved throughout the park over the course of a few weeks, giving us a long-term overview of their movement in twelve hour increments. However, in order to simulate their interactions with human amenities, we needed to bring our focus to an hourly scale, which would better align with the activity of food trucks, fountains, and general human movement.

Squirrel Visualization
Grasshopper: animating the squirrels
Synthetic population

Space — As a proof of concept and to limit the scope of the project slightly, we have decided to focus on the SW corner of Central Park alone (with its exact limits being N 67th street / E 6th avenue / S 57th street / W Columbus).

Having established this, the next part involved finding data for each of the amenities we’ve chosen, such as the trash cans (which we’ve found by cross-corroborating OSM map data nodes with the DSNY trash collection schedule), the drinking fountain locations (as provided by Central Park guides), the hours during which pets are allowed off leash, and the locations of nearby food truck businesses. This gave us a general status quo of the current situation (and squirrel affinities of these different objects at different times of the day), from which we could start building alternative layouts and see how human-wildlife interactions alter as a result of this.

Selected Site Area

Time — For the time simulation, we have opted for an activity based model with the following assumptions:

  • Trash gets collected periodically at 7AM, 3PM, 10PM, and 12AM (human activity which drives the squirrels away), as the trash fills up it attracts more squirrels, until human interference comes to empty it again
  • Repelled by dog fountains when dogs are numerous in the park: 6–9AM and 9PM-1AM off-leash hours, and from 12–1PM and 5–8PM when human owners get off work and have the chance to walk their pets.
  • Based off of existing data, squirrels have a 33% chance of being above ground in trees in the morning, and 23.5% in the afternoon or evening.
  • Food trucks generally remain stationary, changing their location at most once or twice a day
Grasshopper: adjusting the date, hour, and other environmental factors

Baseline Model from Existing Data— The simulation demonstrates the movement of the urban wildlife in New York City, specifically squirrels in the southwestern corner of Central Park. This is largely based off of data provided by the Squirrel Census Data (location within the park in half day increments), in order to simulate their movement over the course of two weeks and how it differs in response to changing seasons and general human activity.

QGIS: squirrel sightings from 10/06/2018–10/20/2018 in 3 day increments
3D Model Visualization: squirrel sightings from 10/06/2018–10/20/2018 in 3 day increments

Urban Simulation Model

Baseline Simulation Model — Now knowing the various factors and their frequencies, we can start moving around these obstacles in order to see how they affect the squirrel population’s movement throughout the park.

The Grasshopper process is as follows:

Complete Grasshopper script
Randomly designating start and destination trees
Human amenities, grouped by water, waste baskets, restaurants, and food trucks
Defining squirrel behavior
Main simulation engine
Post-processing visualization
Temporal styling to signify hourly increments
Baseline simulation of squirrel behavior
Baseline squirrel population and behavioral pattern

Comparison Simulation Model — We make some following new assumptions for the optimized simulation:

  • Keeping the food trucks, restaurant, and fountains stationary
  • Start a garbage sorting policy — Dry Trash, such as plastic bottles, tissues, and artificial products; and Wet Trash, such as food.
  • Creating a 50-meter offset from the existing human pathway, and relocating the waste bags for wet trash along the new path.
  • The ratio for dry and wet trash can amount is set as 1:2, more wet trash can will be placed in the green lawn, away from the existing pathway.

Conclusion/Takeaway — By having a new trash sorting system in Central Park, the squirrels will be mainly active in the green spaces and remain distanced from the human pathways. Aggressive actions, such as stealing tomatoes from old lady’s grocery bags, robbing foods from visitors, and attacking pedestrians, will hopefully decrease. The relocation of trash cans in the middle of the green lawns also solve the existing problem of ‘people needing to walk over to throw their waste.’ At the same time, for a further benefit, the separation of dry and wet trash will be more sustainable for waste decomposition.

3. Sources & “New Data”

NYC Open Data — 2018 Squirrel Census Data with documentation and map

NYC Open Data — Mobile Food Vendors

The Squirrel Census

Central Park Dog Owner Guide

Central Park Daytime Operations

DSNY Trash Collection Schedule and Waste Basket Map