Methods 1 Final Project | December 2023
Zoe Moskowitz

Purpose:

This research hopes to identify discrepancies in temperature between Manhattan County and Bronx County utilizing Hyperlocal Temperature Data collected from the summer of 2019.

Introduction & Background:

Temperatures in New York City are rising every year. As greenhouse gas emissions continue to warm our planet, dense urban cities are experiencing more frequent heatwaves and longer periods of extreme heat. According to the Mayor’s Office of Climate & Environmental Justice, the average number of days above 90℉ will likely triple by the 2050s and quadruple by the 2080s.

Due to a multitude of factors, extreme heat is not felt evenly throughout New York City. Marginalized populations and low-income neighborhoods experience higher temperatures as a result of the urban heat island effect and pervasive inequalities. In areas like the South Bronx, which is considered an urban heat island, temperatures can report anywhere from 1℉ to 7℉ warmer in the daytime and 2℉ to 5℉ warmer in the evening (compared to Manhattan).

An urban heat island is defined by Resources for the Future as “the increase in temperature caused by the built environment.” In the South Bronx there is a major lack of parks, shade coverage and green spaces, a large amount of industrial/polluting facilities (12+ in total), and the presence of the Cross Bronx Expressway - which exposes the population to air pollution and increases asthma rates.

Another reason that low-income neighborhoods and populations suffer from a disproportionate rate of extreme heat is due to decades of redlining and disinvestment. These areas lack adequate housing and typically have poor maintenance, making it difficult to avoid heat and properly cool ones home. Residents are often left with the choice to independently put in expensive cooling measures such as air conditioning or rely on external resources.

Data Source:

The dataset I will be utilizing for this report is available on NYC Open Data and is titled “Hyperlocal Temperature Monitoring.” This dataset was created as a part of the Cool Neighborhoods Initiative and facilitated by the NYC Parks Department, Office of Resilience and NYC Department of Health and Mental Hygiene.

The purpose of this data was to monitor street level temperature on a subset of city blocks in certain neighborhoods with the highest heat mortality risk during the summers of 2018 and 2019. For the purpose of my analysis, I decided to use only the most recent data collected from the year 2019 and narrow the data collection period to 06/15 - 09/21.

Research Questions:

Results

The scatterplots below (Fig. 1-4) show the maximum hyperlocal air temperature by date. The blue colored points represent data taken from Bronx County and the red colored points represent data taken from New York County. When I was done filtering my data, there were 99 data points for each county that I wanted to work with. Due to this, I thought it best to split the scatterplots up by month.

In order to create these visuals, I filtered the NYC Open Data by “borough” and “year”. Due to each date having multiple temperature readings, I then used the group_by and summarise functions to find the maximum hyperlocal temperature per day.

Fig. 1

Figure 1 represents data collected from June, 2019. Out of the 16 datapoints shown, nine show higher max temperature in the Bronx, while seven show higher max temperature in Manhattan.

Fig. 2

Figure 2 represents data collected from July, 2019. Out of the 30 datapoints shown, sixteen show higher max temperature in the Bronx, while fifteen show higher max temperature in Manhattan.

Fig. 3

Figure 3 represents data collected from August, 2019. Out of the 30 datapoints shown, ten show higher max temperature in the Bronx, while twenty show higher max temperature in Manhattan.

Fig. 4

Figure 4 represents data collected from August, 2019. Out of the 21 datapoints shown, seven show higher max temperature in the Bronx, while thirteen show higher max temperature in Manhattan.

Fig. 4

Since this project was focused in the Bronx, I wanted to use ACS data to find which race holds the largest population in the county. The first step was to individually create dataframes for each race using ACS variables (Black, White, Asian, and Hispanic or Latino) and then use this data to create summary statistics.

Once data was found for each individual race, I combined the information into one dataframe. Through summary statistics, I found that the Hispanic/Latino population holds the majority and makes up an estimated 56% of the total population in Bronx County.

Fig. 5

After finding that the Hispanic or Latino made up a majority of the population in Bronx, I wanted to create a borough based map of where this population lives within the county. To make the map, I was able to utilize a dataframe created when creating summary statistics from Figure 4.

Additionally, it was of interest to see where the monitors that collected the Hyperlocal Temperature from NYC Open Data were located. My presumption was they would have spanned a large part of the county - but this was not the case. The sensors are represented by the black colored groupings on the map.

Discussion

to be written

Next Steps

to be written

Methods Appendix

NYC Open Data

The main datasource I used came from NYC Open Data. Hyperlocal Temperature Monitoring was published on August 20th, 2021 through the Cool Neighborhoods Initiative and was created and monitored by multiple teams within the NYC government. These teams include: NYC Parks Department, Mayor’s Office of Climate and Resiliency, and NYC Department of Health and Mental Hygiene.

Within this dataset there were ten columns for each point of data:

  • Sensor.ID = Unique identifier of sensor

  • AirTemp = Average hourly air temperature, degrees; Fahrenheit

  • Hour = Hour of day, military time

  • Latitude = Latitude of sensor

  • Longitude = Longitude of sensor

  • Year = Year

  • Install.Type = Type of mounting - street tree/street light

  • Borough = Borough

  • ntacode = Neighborhood Tabulation Area (NTA) Code

One of the pitfalls of this data is that the sensors used to collect data are not spread across counties. When I made the map that showed Hispanic or Latino population in Bronx county with the overlay of where the sensors were, it was enlightening to see that the sensors lived only in three pockets of the county. If one wanted to have a more accurate read on hyperlocal temperature throughout the entire county, there would be additional sensors.

ACS Data

ACS Data, or American Community Survey, was used to find which race had the highest population in Bronx County. ACS provides annual estimated data in multiple categories including (but not limited to) education, income, employment, and housing.