Quick question: in your experience, which picked up aquifer stress sooner in the Central Valley — GRACE TWS anomalies around 2002–2015 or deep monitoring well hydrographs showing >1 m/yr decline — and what does that suggest for sizing and siting managed aquifer recharge when we’re triaging budgets? I’m comparing notes for a basin plan and want the most reliable early warning to keep withdrawals within sustainable yield.
In the Valley, deep monitoring wells in hotspots usually flag ‘early warning’ before GRACE, while GRACE confirms the basin-scale trend — dipstick vs. satellite. For MAR triage, overlay GRACE trends with deep-well declines to pick subbasins, then do a 24–48 h infiltration test near canals on coarse soils to size cells. Just watch that GRACE misses narrow cones of depression, so keep a few clustered deep wells to track those.
I’ve seen multi-level piezometers catch stress via a sharp dH/dt turn weeks to months before basin-wide GRACE shows it, and cGNSS sometimes tips off even earlier — like your smoke alarm vs the fire marshal. That suggests setting simple triggers (|dH/dt| > about 0.5 m/yr plus >5 mm/month subsidence) to prioritize MAR scouting in coarse alluvium up-dip of the cone. Are you using JPL RL06 mascons or gridded CSR (https://grace.jpl.nasa.gov), since that choice can nudge your lead time?
in 2002–2015 the first tell I saw was the spring rebound shrinking in pressure‑transducer pairs, well before any GRACE TWS slope looked scary. For “triaging budgets,” size MAR to local Sy and site just upgradient of subsiding clays; use https://grace.jpl.nasa.gov as a basin check, but trigger on a baro‑corrected dH/dt kink mid‑season. @jlee927, watching rebound amplitude loss has been my most reliable early warning.