Projects / Back Bay

Back Bay Water Quality Analysis

Data analysis project in Python

Dates: May – Jun 2025
Status: Complete
Dataset: 2,400 rows, 1994–2018
Tools: pandas, NumPy, Matplotlib

Bubble chart of average water temperature by year from 1994 to 2018, with bubble size showing the size of each year's percent change — Average water temperature by year. Bubble size shows the absolute percent change for that year.

Bar chart of percent change from 1994 to 2018: water temperature up about 3 percent, air temperature down about 3 percent, water depth up about 34 percent — Percent change, 1994 to 2018

Line chart of average water temperature in degrees Celsius from 1994 to 2018 — Average water temperature (°C)

Line chart of average air temperature in degrees Fahrenheit from 1994 to 2018 — Average air temperature (°F)

Line chart of average water depth in meters from 1994 to 2018 — Average water depth (m)

Overview

In May and June 2025 I analyzed a 2,400-row dataset of water temperature, air temperature, and water depth for Back Bay National Wildlife Refuge in Virginia Beach, Virginia, covering 1994 to 2018. The data came from the Department of the Interior's water quality records for the refuge.

What I built

Cleaned and organized the dataset in pandas, then computed yearly averages for each variable
Wrote a Python script that takes a starting and ending year from the user and reports the percent change in average water temperature, air temperature, and water depth over that range, plus the largest change in water temperature and the year it happened
Plotted each variable over time in Matplotlib, along with a percent-change bar chart and a bubble chart that combines yearly water temperature with the size of each year's change

Results, 1994–2018

Water temperature: +2.83%
Air temperature: −3.03%
Water depth: +33.50%
Largest water temperature change: 28.59%, in 2016

What the data shows

Average water temperature rose even though average air temperature fell slightly, so the two didn't move together over this period. That suggests the water temperature change was driven by something else, possibly water depth or other environmental conditions.

The broader point is that air and water temperatures don't always follow the same trend. Understanding a body of water means looking at several variables at once, which matters for managing coastal areas and planning for climate impacts.

Challenges

Real-world environmental data is noisy and often incomplete, so a meaningful part of this project was deciding how to handle missing or questionable readings without distorting the overall trends.

What I learned

Practical data cleaning with pandas before any analysis is meaningful
How to turn a raw dataset into a clear visual narrative
Communicating quantitative results in a way a non-specialist can follow