tonychengu / adopt-a-stream-georgia

citizen science + data science project

Adopt-a-Stream Georgia data science projects

Local water quality impacts health, well-being, and the local economy in many ways. Data about water quality in Georgia has been collected for many years by Adopt-a-Stream (AaS), and is made publicly available through their web portal. This is a great beginning to promoting public access and involvement in monitoring the impact of poor quality water on communities, and in finding causal connections between water quality outcomes and economic or political decisions about land use, industrial regulation, etc.

The water quality dataset hosted by AaS-GA was created through the volunteer efforts of many citizen scientists throughout the state over many years.

Data sets

We currently have two datasets from AaS-GA. One is a direct database export sent by them in December, 2021 in XLS format. The other is scraped from their web portal in September, 2021, and contains significant differences from the direct export that are not yet well understood. The two raw datasets are recorded here.

The raw and processed data are stored in this repository using git's Large File Storage. They are stored as zip files and should be uncompressed in your working environment before you run any of the existing scripts. If the repo is over its (small) monthly bandwidth limit, you can download the individual zip files from this GDrive shared folder.

Steps to get started with this repo

Data files

This repo contains some large data files that are managed using Git's Large File System (LFS). Follow the instructions in the link to install the LFS CLI tool before you clone.

Library dependencies

Make sure to run pip install -r requirements.txt in the repo's main directory to ensure you have the python library dependencies.

Geodata-specific dependencies

Mapping geodata with plotly will also require npm and node installed in your OS, which support the jupyterlab-plotly extension. For instance, on OS X you might use homebrew to install the node dependencies:

brew install npm node
jupyter labextension install jupyterlab-plotly

and there are other ways to do that for other operating systems.

Project maturity

As of Jan 2022, we are in the earliest stages of gathering and preparing a dataset for further research.

The current project goals are the basic validation and preparation of the available data.

Other resources used

This API provides an on-demand lookup for local information based on FIPS identifiers. utils.py provides code to use this, but this data is already pulled for all the recorded data points and cached in the repo.

Plotly's own record of basic US demographics that we use at the state level for GIS county information from utils.get_state_map, but it could be useful for scientific exploration at a later date.

The watershed boundary data was downloaded directly from the USDA data gateway. In this case, the "Geographic NAD83 projection" was chosen (arbitrarily) and the output format was ESRI Shapefile.