☀️ Impact of Utility-Scale Solar Facilities on Land Values in Virginia

This repository contains data, scripts, and a Shiny app developed for the 2025 Data Science for the Public Good (DSPG) program at Virginia Tech. Our research investigates how proximity to utility-scale solar facilities affects land values across Virginia at both the county and parcel levels.

---

📁 Repository Structure

solar_facility_dspg/

├── All_project_codes/ - All data cleaning, merging, and regression scripts
├── clean_data/ - Cleaned datasets (CSV, shapefiles, RDS) for analysis
├── raw_data/ - Raw, unprocessed datasets from public/state sources
├── ShinyApp/ - R Shiny dashboard for interactive exploration
├── README.md - You are here
└── DSPG_Git.Rproj - R project file

---

Reproducibility Guide

1. Required R Packages

Install the following R packages:

tidyverse, sf, plm, broom, readr, ggplot2,
tigris, leaflet, shiny, shinythemes, shinyWidgets,
DT, lubridate, stringr, readxl, rgdal

---

Data Overview

Final Datasets for Regression

• clean_data/county_level_cleaned_model_data.csv — County-level dataset for panel regression
• clean_data/final_parcel_data.rds — Final parcel-level dataset (merged attributes, distances, zoning)

Raw Input Files (raw_data/)

• Solar_facility_VA - Copy.xlsx — Raw solar facility info with coordinates
• va_county_pop_all.csv, va_avg_housing_age.csv — County-level ACS/demographic data
• New_Corn_Data.csv, New_Soy_Data.csv — Raw crop yield data
• old_Corn_data.xlsx, old_Soy_data.xlsx — Deprecated crop data
• va_housing_by_year.csv, va_housing_status.csv — Raw housing status metrics

Cleaned Data (clean_data/)

• parcel_level_variables.csv, final_parcels_with_all_attributes.csv — Intermediate merged parcel datasets
• parcels_urban_dist.csv, parcels_water_dist.csv — Parcel-level distance to urban/water
• parcel_land_cover.csv — Land cover classification per parcel
• solar_facility.csv — Cleaned solar facility data
• va_population.csv, va_housing_year.csv — ACS demographic inputs
• final_va_parcels_clean.gpkg, merged_data.shp — Spatial layers
• model_results.csv — Regression model output

---

Running the Regression Models

County-Level

Run in R:

source("All_project_codes/county_level_reg_mods.R")

Optional supporting scripts:

• county_level_merge_code.R
• county_level_variables.R

Parcel-Level

Run in R:

source("All_project_codes/parcel_level_reg_mods.R")

Supporting scripts:

• parcel_level_mod_distance_variables.R
• parcel_level_variables_merging.R
• parcels_transactions.R

---

🌐 R Shiny Dashboard

To launch the dashboard:

setwd("ShinyApp")
shiny::runApp()

The dashboard includes:

• Parcel and county map layers
• Solar facility locations
• Transmission line overlays
• Regression model outputs
• Team and methodology tabs

---

Acknowledgments

This research was conducted as part of the Virginia Tech Data Science for the Public Good (DSPG) Summer 2025 program.
Supported by the USDA NIFA DATA-ACRE Grant No. 2022-67037-36639.

---

Project Team

Aziza Altyyeva & Collin Holt
Mentored by Dr. Michael Cary, Anbin Rhee
Contact: altyyevaa@berea.edu