Introduction
For lab four I choose to investigate storm surge zones in New York City and the disproportionate effects hurricane events might have on the BIPOC population. As a co-facilitatory for Pratt’s Disaster Resiliency Network (PDRN) I am particularly interested in learning about natural disasters in NYC, the effects they could have on communities, and the ability for the community to recover after an event. Recently, I have been learning about the aftermath of Hurricane Katrina in Louisiana and thinking about the devastating effects Super Storm Sandy had in this area. In New Orleans, the majority of the city lies below sea level, with a historically wealthier, whiter community living in the few areas that are above sea level. This distribution was significant in the aftermath of the Hurricane since many of the displaced residents of New Orleans were less wealthy and most were African American. My main question for this lab is what regions of the city would be flooded in the event of another hurricane like Sandy? And within these affected areas, what percentage are people of color? Furthermore, how much average wealth is in the affect areas? I hope that by investigating these patterns I can uncover any inequalities and possible disadvantages to the BIPOC and lower income communities within flood zones. This is the first step in creating an equitable recovery plan after disasters, and to build community resiliency.
Discussion
The NYC Department of City Planning actually has an interactive map showing the flood zones in the event of different category Hurricanes available here. I’m glad this resource is publicly available. However, when looking at it from an information visualization perspective, I wanted to improve upon this basic map. In the screen shot below, you can see that the base map is a dark satellite image and the flood zones appear in a bright, sometimes neon coloring. This stark contrast is useful but a bit too glairing in my opinion. In my map, I choose less vibrant and more contiguous colors. I also used a transparency of 40% to ensure visibility of the census tract boundaries and PIBOC dot density information.
Materials
The primary tool used in lab four is Carto. The website for this tool can be found here: https://carto.com The tool is designed to be open to the public however, the free version has limited data uploads, maps, and layers per map. This tool is set up to easily import data in a variety of formats including excel doc and shapefiles. The actual manipulation of the data requires more knowledge of how the tool works.
Methods/Process
Learning about Carto
To learn about Carto I watched the youtube tutorials and demo video that Professor Sula posted on our canvas class page. In addition, I spent several hours trying different functions of the program and created a variety of different versions of this map. The demo provided important step by step instructions for many elements of the program. I ran into several challenges importing my data at the beginning, including random error messages and data display issues. Allowing myself ample time to experiment within the program was necessary for troubleshooting, learning more about the program, and improving my visualization.
Obtaining Data
Most of this data I had actually used previously and compiled in my personal data library over the past two semesters. The raw flood data set can be found here and the demographic data was complied from a variety of American Community Surveys 2019 5 year estimates which I obtained from the census data portal here.
Putting it all Together
After learning how to use Carto and choosing my datasets, I edited the datasets slightly in ArcMaps before exporting the data and loaded the datasets into Carto. Originally, I had data labels on the hurricane zones and I was attempting to add a widget to the bottom of my Carto map to better examine the percent of the population per census tract living below the poverty level. This early attempt looked like this:
Ultimately, I abandoned the attempt to use widgets because I did not have enough time to fully learn about and incorporate them properly. Next, I played around with the flood zone data layer and symbolized it with a variety of blues, darkest blue for the event of a hurricane 1 and lightest blue for a hurricane 6. After cleaning the storm zone layer and creating a legend, I tried displaying the demographic data in a variety of ways. I experimented with the graduated symbol size and which attribute to have them display. Initially, I intended to show percent poverty with the graduated symbols but ultimately that visualization did not tell a clear story. I settled on using the graduated symbols to display the percent of the population per census tract that are people of color. I’ve calculated this by dividing the total number of people who do not identify as “white alone” divided by the total population in each census tract. In the legend I liked how it showed the values for each size circle as shown in the picture below.
In my final version, however, I changed this label system to simply label the largest circle as 100% and the smallest as 0%. While this loses the level of detail some may prefer, I think it cleans up the legend and makes it easier to understand at a quick glance.
I also decided to include a color scale for the graduated symbols, with light green representing lower median household incomes and dark green representing higher median household incomes. I choose a continues green scale because I associate green with money and I also felt this color stood out from the base map and the flood zones. Finally, I formatted the popups to include additional demographic data about each census tract.
Results and Reflection
My final Carto map I created can be seen above and can also be found here. I really appreciate the ease with which Carto allows you to make an interactive web map. I have only made one other interactive web map and it was for an undergraduate course I was in. For that project we did not use carto, but rather a combination of ArcMap, leaflet, and another program which I do not remember the name of. That process was really cumbersome and I will definitely be using Carto for all my future interactive web map needs. The only thing I did not enjoy about using carto was the inability to manipulate the actual data tables how I needed to. For instance, the flood zone map came with a few extraneous categories which I tired to symbolize as completely translucent. However, the translucent display can only be set at a layer level, not by the subcategories. So, then I tried to delete the extra rows from the data table but I was not able to edit or alter the data table within Carto. As someone who is familiar with using other mapping platforms like ArcMap and QGIS, I found using Carto to be somewhat limiting in this way. I had to rely on ArcMap for all the data editing I needed to do.
In conclusion, I feel this map does an adequate job displaying three separate but related data sets. It tells the story of community risk of displacement, highlighting areas with majority BIPOC communities (represented by larger circles) and of low income neighborhoods (represented by light green circles. Three specific areas in NYC that seem most at risk of flooding and displacement due to their low income and high percentage of communities of color are the east edge of Manhattan, the south east part of the Bronx, and Brownsville and the neighboring community to the east bellow Jackie Robinson Parkway. These three areas all consist of large circles, indicating high percentages of POC and light green symbols showing lower average median incomes. These communities would likely be least resilient in the case of a hurricane and should be focused on fro recovery and resiliency disaster planning.
Lower Manhattan The Bronx Eastern Brooklyn
References
Adding a dataset to Carto tutorial: https://www.youtube.com/watch?v=5gZGQwcns3Y
Creating a map on Carto tutorial: https://www.youtube.com/watch?v=Yx9LwuDVBe0