nyc park ranger response to Animal Condition calls

Charts & Graphs, Lab Reports, Visualization


Park rangers respond to hundreds of calls regarding the health and safety of wild and domestic animals in New York City parks. These calls involve dozens of species in various states of health that affect the outcome of these responses.


The dataset used to make these charts and graphs was found on NYC Open Data. I originally intended to try to find data specifically about Prospect Park. There were no datasets that only contained data about Prospect Park and trying to limit other datasets to a single park left them too small to use. While searching for a dataset regarding NYC Parks in general, I found NYC’s Urban Park Ranger Response to Animal Condition and chose to work with it because I was interested in the topic and it related to my experience working at the Animal Care Center of NYC.

The dataset was relatively clean and did not require much manipulation. I was able to change everything I needed to in Tableau Public, the same program I used to make the graphs. A drawback to this dataset was that it only contained a year’s worth of data. This limited what I was able to convey about the data without being redundant. Especially considering the numbers were dominated by responses to a single species.

Tableau Public was the program used to create the graphs and charts.


The first graph made was a line graph to visualize the number of park ranger responses each month over the course of the year. The data went from mid May of 2018 to the end of June 2019. I decided to exclude the data from the month of May so that the graph presented a 12 month period. I thought this was appropriate because only 3 responses were included in the data from May 2018 and wouldn’t skew the representation.

While verification is still necessary, the spikes in calls in August and November is presumably because there are more visitors to the parks during these months not because there are more animals, healthy or sick, during these months.

When working with the data, I noticed that raccoon was the species that dominated the amount of responses park rangers made. I created two groups within the Species Description dimension, Raccoons and Other Species. I used a table calculation to graph the percentage of the total Sum(# of Animals). As clearly displayed in the pie chart raccoons make up 53% of the species responded to by NYC Park Rangers.

To visualize which species rangers were responding to most in each borough, I created a highlight table. I grouped the species into broader categories similar to what the Animal Care Center use. This helped condense and better visualize the data. Making sure to analyze the percentages based on the column so that they were calculated by borough rather than species, I used 7 steps of color. Despite being somewhat arbitrary in number groupings (29% is grouped with 36% rather than 25%) it best visualized the top 2 types of animals were responded to most in each borough, the groupings still made sense visually and most of the species were in a single group under 10%.

While this table also reinforced the prevalence of responses to raccoons it clearly puts birds as number two and highlights the differences in the Bronx and Staten Island with their higher rates of responses to birds and deer respectively.

I chose to visualize the resulting action of the ranger’s response because I thought it was an important aspect of the data and worked as the final step in the when the Rangers responded, what they were most often responding to and what their response narrative.

The simple bar graph shows that the most common result of a response was the animals being brought to the Animal Care Center of NYC or the call was unfounded and action unnecessary.


Compiling the graphs into a single document, I tried create a cohesive image that would make a clear statement. Hopefully, the user would follow the progression from how many responses the rangers make to what they’re most likely responding, how that is influences by where they are responding and the response is.


The visualization process was limited by the dataset only containing a year’s worth of data. Data covering a longer period of time, two to five years, would have allowed for a more in depth if not entirely different view of the park ranger response to calls regarding the condition of animals in the cities parks.

While the dataset didn’t require cleaning up, I would have like to manipulate it more to try and produce different stories to tell. Because the numbers on the raccoons were so high compared to other species, it was hard to create graphs that didn’t express the fact that responses to raccoons were common. I also would have made more graphs visualizing the data of different species within broader categories, i.e. the type and number of snake species responded to by rangers. By the time I realized I wouldn’t be getting better graphs by looking at the data as a whole I felt locked into that narrative