Real-world hydrology use cases built with AquaScope. Self-contained, reproducible Jupyter notebooks on real data — every headline result checked against a published reference.
| # | Title | Data | Methods | Validated against |
|---|---|---|---|---|
| 01 | Bulletin 17C flood frequency — Potomac at Little Falls | USGS gauge 01646500 · 1931–2025 · n = 80 | LP3 · GEV (L-moments) · Q-Q & P-P | FEMA DC FIS (2010) Table 4 — within ±10 % |
| 02 | Baseflow & hydrological signatures — French Broad at Asheville | USGS gauge 03451500 · 1995–2025 daily | Lyne-Hollick + Eckhardt filters · 22 signatures | Wolock (2003) USGS OFR 03-263; Santhi et al. (2008) JoH |
| 03 | FAO-56 ET₀ and rice crop water — Bangkok | FAO-56 Example 18 + Open-Meteo daily, 2024 wet season | Penman-Monteith · Kc curves · soil water balance | FAO-56 Example 18 (5.0 mm/day) and CLIMWAT Bangkok climatic norms |
| 04 | Mann-Kendall trend & Pettitt change-point — Red River at Grand Forks | USGS gauge 05082500 · annual peaks 1882–2025 | Mann-Kendall · Sen's slope · Pettitt · PELT | Ryberg et al. (2014) J. Hydrol. Eng.; Vecchia (2008) USGS SIR |
| 05 | Bivariate flood copula — Potomac peak/volume | USGS gauge 01646500 · daily 1950–2025 | Gaussian/Clayton/Gumbel/Frank · joint return periods | Salvadori & De Michele (2004) WRR; Genest & Favre (2007) JHE |
| 06 | Taiwan RPI & DO trend — Tamsui River basin | Taiwan MOENV dataset AQX_P_07 · multi-station WQ | AI recommender · Taiwan RPI · Mann-Kendall | Taiwan EPA RPI methodology; Chen & Liu (2003) EMA; Liu et al. (2019) Water |
Every case fetches its own data at run time, runs the analysis end-to-end, generates the committed plots in outputs/, and prints a side-by-side comparison against the published reference.
AquaScope is the engine behind every case. Install it from PyPI:
pip install aquascope # core — collectors + hydrology
pip install "aquascope[viz]" # add matplotlib, seaborn, folium (used by the demos)
pip install "aquascope[all]" # full stack — ML, viz, spatial, dashboardFor the complete feature list, documentation, and roadmap, see the main AquaScope repository.
Each case folder pins its exact dependencies (AquaScope version + extras) in requirements.txt, so cases stay reproducible as the library evolves:
cd 01_potomac_flood_frequency # pick a case
python3 -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt # installs the pinned aquascope[viz] + case extras
jupyter notebook notebook.ipynbPython. AquaScope requires Python ≥ 3.10. On Python 3.14+ some scientific wheels may not yet be published; if installation is slow or fails, fall back to Python 3.13.
GitHub renders .ipynb files natively. Open any case folder, click notebook.ipynb, and the full analysis — code, prose, plots, verified results — appears inline.
NN_<slug>/
├── README.md ← scenario, methods, verified results
├── notebook.ipynb ← the runnable analysis end-to-end
├── requirements.txt ← pinned dependencies
├── data/ ← static inputs or a fetch script for live data
└── outputs/ ← generated artifacts — CSV tables, PNG figures
Code: MIT — same licence as AquaScope.
Data referenced by these notebooks is fetched from public-domain U.S. federal sources (USGS NWIS) under the USGS data policy.
If a case here informs your published research, please cite AquaScope itself — the canonical BibTeX entry lives in the Citation section of the main repository.