NPS Discover
A fully automated ETL pipeline that extracts data from various data sources, transforms the data, and visualizes visitation and economic trends of national park resources.
Purpose
The National Park Service is primarily funded by Congress through the annual appropriations cycle, mandatory funds, user fees, and private philanthropy. With the large majority of NPS funding coming from federal dollars, it has become increasingly important to the protection of the NPS budget to communicate the positive return on that funding.
As this is an exhibition, the project was intentionally designed to be overly complex by incorporating a variety of tools and expertise that may not necessarily be required while being designed in a way that functions longterm with 0 cost. Several capabilities that were coded manually are available for purchase through PowerBI Pro.
Architecture
- Windows Task Scheduler is set to run Python Scripts on schedule
- Data collection:
- Data is requested from the NPS API and saved as a CSV. API Call
- Economic Impact Reports are manually collected from the Department of the Interior, and tables are extracted and compiled PDF Reader
- Public Use Statistics Reports are collected from the NPS Visitor Usage Statistics Report, compiled, and saved to CSV Visitors
- Weather and Climate data is scraped from Wikipedia and saved as a CSV Wiki Scraper
- Data is connected to PowerBi service On-Premises Data Gateway
- PowerBi Dashboard is connected to dataset
- PowerBi Service is set to Refresh and Publish at pre-determined schedule
- Finalized Powerbi report is published
User Analysis
An automated data collection survey that enables users to submit feedback directly via dashboard interaction.
On submission, user feedback data is pushed to google forms and then analyzed within the dashboard at the next refresh.
Dashboard
The final product is an interactive BI report that visualizes 20 years of national park data to help understand the economic impact of resources managed by the National Park Service.
Dashboard refreshes hourly from 9-4PM. 2022 Visitatation data has been released with finalized 2022 Economic data expected 06-2023.
Automated Briefing Export
Python automation enables scheduled refresh and PDF export of an executive briefing based on data loaded into the dashboard.
