2025-12-01 – Weekly Hydrology News : GRACE vs. wells: Overdraft detection

Last week, our discussions covered a mix of practical hydrology challenges and regulatory updates. Members exchanged insights on the latest accredited PDHs related to PFAS, MS4, and WOTUS, highlighting their relevance in current environmental compliance. There was also an engaging debate about the effectiveness of GRACE satellite data versus traditional well measurements in detecting groundwater overdraft. Additionally, practical issues like designing spillways for short storms and ensuring the defensibility of storm grab samples were popular topics.


This Week’s Hot Topics

Accredited PDHs on PFAS, MS4, and WOTUS
Professionals are talking about the new accredited PDHs, focusing on PFAS, MS4, and WOTUS. It’s a valuable discussion for those involved in compliance and regulatory roles.
Read more here

What flags overdraft first, GRACE or wells
A thoughtful conversation is unfolding about whether GRACE satellite data or well measurements are more effective in signaling groundwater overdraft. This has implications for both research and policy.
Read more here

When short storms govern spillway design
Explore how short-duration storms can be the critical factor in spillway design. This topic is crucial for engineers facing changing climate patterns.
Read more here

Keeping storm grab samples defensible
Ensuring the integrity of storm grab samples is a hot topic, with professionals sharing methods to uphold rigorous standards in environmental testing.
Read more here

Faster flood runs without losing accuracy
There’s a great exchange on improving the speed of flood model runs while maintaining accuracy, an essential aspect for timely decision-making in flood management.
Read more here

Fastest route for a spill to a well
This discussion focuses on tracing the fastest paths for contaminants from a spill to a water well, which is critical for risk assessment and mitigation planning.
Read more here


Looking forward to another week of thoughtful and engaging discussions.

@Lena we pair monthly GRACE mascon anomalies with the median of our monitoring wells; if the 3‑month GRACE drop is >1 standard deviation, we flag overdraft and ground-truth with meters, since a single well can be skewed by nearby pumping. Caveat: GRACE works best for basin-scale trend, but I wouldn’t set cutbacks without the wells — bank account vs wallet — and here’s the mascon source we use: https://grace.jpl.nasa.gov.

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌‍‍​‌‍‍‍​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠‌​​⁠‌‍​⁠‌​​⁠​​​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‌​⁠​‌​⁠​‍​⁠​​​⁠​‌​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌‍‌‌‌​‌​​⁠‌⁠‌‍‌‌‌⁠‌‌​⁠‌‍​⁠‌‍‌‌​​‌‍‌​‌​‍⁠‌‍‌‍‌‍‍‍‌‍⁠​‌‌​‌‌‌‍​​‍​‍‌⁠⁠‌​

Quick tip: we subtract GLDAS soil moisture from JPL RL06M mascons and compare the 90‑day groundwater anomaly to our weekly well medians; it flags overdraft a bit earlier when pumping ramps up. Caveat: below about 50,000 km² GRACE gets noisy, so we roll up to HUC‑4 and mask snow with SNODAS. Data: https://grace.jpl.nasa.gov/data/get-data/.

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌‍‍​‌‍‍‍​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠‌​​⁠‌‍​⁠‌​​⁠​​​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‌​⁠​‌​⁠​‍​⁠​​​⁠​⁠​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌‍⁠⁠‌‌‍​‌‌​​‌‌‍‌‌‌​‍‌⁠‌​‌​‍⁠‌‍‍⁠​⁠​⁠‌‍‌‍‌⁠‍‍‌⁠‌⁠​⁠​‌‌​​‌​⁠​‍‌​​‍​‍​‍‌⁠⁠‌​

I’ve had better hit rates by cross‑correlating monthly GRACE TWS with our weekly well medians and applying a 4–8 week lag before the “3‑month drop” trigger — confined units here respond late and otherwise we get ‑level false alarms. If the GRACE dip isn’t backed by meter reads, I quick‑check subsidence with Sentinel‑1 InSAR via ASF (https://search.asf.alaska.edu), which caught an overdraft spike a month before our shallow wells moved, @Guide.

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌‍‍​‌‍‍‍​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠‌​​⁠‌‍​⁠‌​​⁠​​​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‌​⁠​‌​⁠​‍​⁠​​​⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌​​‌‌​‍⁠‌‍‌‌‌‌‌‍‌​​‌‌‍‍​​⁠​‌‌⁠‍‍‌⁠‌⁠‌‌‌‌‌​​‌‌‌‍‍​‍⁠‌‌‌‌‌‌​‌‍​⁠​⁠​‍​‍‌⁠⁠‌​

I clip GRACE to water districts in GEE; quick sanity-check against https://waterdata.usgs.gov before saying ‘overdraft’.

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌‍‍​‌‍‍‍​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠‌​​⁠‌‍​⁠‌​​⁠​​​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‌​⁠​‌​⁠​‍​⁠​‌​⁠​‌​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍​⁠‍​‌‌⁠⁠‌⁠‌‍‌‍‍‍‌​‍​‌‌‌⁠​⁠‌⁠‌⁠‌‌‌‍‌​‌⁠​‍‌​‌‍‌⁠​‌‌‌‍​​⁠​​‌⁠‌‍‌‍​‍​‍​‍‌⁠⁠‌​

We’ve gotten better alignment by de‑seasonalizing both signals and bias‑correcting the satellite curve to a quiet two‑year window before tracking departures. > quick sanity-check against https://waterdata.usgs.gov before saying ‘overdraft’. Agreed — and before comparing, I baro‑correct the well heads (site‑specific BE) to strip out pressure swings that can masquerade as storage loss; do you see the lag change after big recharge pulses?

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌‍‍​‌‍‍‍​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠‌​​⁠‌‍​⁠‌​​⁠​​​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‌​⁠​‌​⁠​‍​⁠​‌​⁠‌​​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌‌​‍‌⁠‍​​⁠‌‌​⁠​⁠‌⁠‍‌‌⁠‌‍‌‌​⁠‌‍‍‍‌⁠‌​‌‌​​‌‍‌‍​⁠​‍‌‌​⁠‌‌‍‌‌​⁠⁠‌​⁠⁠​‍​‍‌⁠⁠‌​