Case Study: Regional Groundwater Analysis¶
Overview¶
This case study demonstrates a comprehensive regional groundwater analysis across nine major U.S. Metropolitan Statistical Areas (MSAs) using the pyGWRetrieval package.
Study Parameters¶
| Parameter | Value |
|---|---|
| Temporal Coverage | 1970-01-01 to 2025-11-17 (55 years) |
| Data Source | USGS NWIS Field Measurements (gwlevels) |
| Buffer Radius | 25 miles per zip code |
| Metropolitan Areas | 9 |
| Zip Codes Analyzed | 99 |
Metropolitan Areas Analyzed¶
| MSA | Primary State(s) | Wells | Records |
|---|---|---|---|
| New York | NY, NJ | 6,375 | 3,025,035 |
| Miami | FL | 1,718 | 1,408,186 |
| Washington | DC, VA, MD | 8,184 | 1,342,912 |
| Houston | TX | 2,283 | 651,355 |
| Boston | MA | 2,273 | 583,494 |
| Philadelphia | PA, NJ, DE | 11,177 | 496,196 |
| San Francisco | CA | 1,073 | 320,785 |
| Chicago | IL | 956 | 91,506 |
| Dallas | TX | 140 | 76,458 |
Total: 7,995,927 records from 33,018 monitoring wells
Key Findings¶
Regional Trends¶
| Region | Trend (ft/yr) | Direction | Significance |
|---|---|---|---|
| Dallas | -10.64 | Rising | p < 0.001 |
| Chicago | -2.26 | Rising | p < 0.001 |
| San Francisco | -1.03 | Rising | p < 0.001 |
| Philadelphia | -0.23 | Rising | p < 0.001 |
| New York | +0.35 | Falling | p < 0.001 |
| Washington | +1.10 | Falling | p < 0.001 |
| Houston | +0.07 | Stable | Not significant |
| Boston | -0.003 | Stable | Not significant |
| Miami | -0.03 | Stable | p < 0.001 |
Recovery Success Stories
- Dallas: Remarkable recovery (+10.6 ft/year rising), likely due to effective groundwater management
- Chicago: Significant improvement (-2.3 ft/year rising), possibly due to reduced industrial pumping
- San Francisco: Moderate recovery (-1.0 ft/year rising), reflecting drought response measures
Areas of Concern
- Washington DC: Only region with significant declining trend (+1.1 ft/year deepening)
Sustainability Index¶
Composite sustainability scores (0-100 scale):
| Region | Index | Risk Level |
|---|---|---|
| Dallas | 73.0 | Low |
| Chicago | 68.1 | Low |
| Philadelphia | 68.0 | Low |
| Miami | 66.9 | Low |
| Boston | 65.9 | Low |
| New York | 63.3 | Low |
| Houston | 62.7 | Low |
| San Francisco | 60.0 | Medium |
| Washington | 40.4 | Medium |
Future Projections (10-Year)¶
| Region | Projected Change | Confidence (R²) |
|---|---|---|
| Dallas | -115.0 ft (rising) | 0.73 (Good) |
| Chicago | -43.5 ft (rising) | 0.73 (Good) |
| San Francisco | -16.8 ft (rising) | 0.59 (Moderate) |
| New York | +7.4 ft (deepening) | 0.56 (Moderate) |
| Washington | -5.2 ft (rising) | 0.71 (Good) |
Generated Visualizations¶
The analysis produces 15 publication-ready figures:
- Regional Trends (
regional_trends_by_msa.png) - Trend analysis by MSA - Data Quality (
data_quality_analysis.png) - Coverage and density metrics - Distributions (
regional_distributions.png) - Statistical distributions - Temporal Patterns (
regional_temporal_patterns.png) - Decadal changes - Monthly Boxplots (
monthly_boxplots_by_region.png) - Seasonal patterns - Annual Boxplots (
annual_boxplots_by_region.png) - Inter-annual variability - Correlation (
regional_correlation_clustering.png) - Regional relationships - Extreme Events (
extreme_events_analysis.png) - Drought analysis - Rate of Change (
rolling_trend_analysis.png) - Trend acceleration - Geographic (
geographic_grouping_analysis.png) - Coastal vs inland - Dashboard (
regional_summary_dashboard.png) - Summary scorecard - Change Points (
change_point_analysis.png) - Regime shifts - Sustainability (
sustainability_index.png) - Risk assessment - Projections (
future_projections.png) - Water level forecasts - Statistics (
comprehensive_statistics.png) - Summary tables
Output Files¶
| File | Description | Size |
|---|---|---|
all_groundwater_data.parquet |
Complete dataset | ~80 MB |
data_by_zipcode/*.parquet |
Per-zipcode files | ~130 MB |
groundwater_wells.geojson |
Well locations | ~25 MB |
sustainability_metrics.csv |
Risk scores | ~4 KB |
water_level_projections.csv |
Forecasts | ~4 KB |
comprehensive_statistics.csv |
Full statistics | ~4 KB |
ANALYSIS_REPORT.md |
Complete report | ~50 KB |
plots/*.png |
15 figures | ~15 MB |
Total output: ~275 MB
Running the Analysis¶
Or programmatically:
from pyGWRetrieval import GroundwaterRetrieval, TemporalAggregator
# Initialize
gw = GroundwaterRetrieval(
start_date='1970-01-01',
data_sources=['gwlevels']
)
# Retrieve data for multiple zip codes
data = gw.get_data_by_zipcodes_csv(
'AirbnbMSACity_with_ZipCode.csv',
zipcode_column='ZipCode',
buffer_miles=25,
parallel=True
)
# Aggregate and analyze
aggregator = TemporalAggregator(data)
annual = aggregator.to_annual()
trends = aggregator.get_trends(period='annual', parallel=True)