Experts clash over airborne EM and groundwater models; safe-yield estimates diverge

A technical portion of the Indian Wells Valley workshop centered on how airborne electromagnetic (AEM) data and groundwater models are being used to map subsurface clay and fault structures and to estimate sustainable inflow to the basin.

Dr. Ryan Smith, an associate professor at Colorado State University who independently reviewed the AEM ("Skym") dataset, explained how time-domain electromagnetic surveys infer electrical resistivity and why interpretation requires careful calibration to well logs and salinity data. "These electric fields then create their own magnetic fields... it looks like complete nonsense to someone who's never worked with these data," he said, describing inversions and rock-physics transforms that relate resistivity to lithology.

Smith and other consultants said AEM is particularly strong at mapping clay units and fault traces where resistivity contrasts are clear; however, where groundwater salinity is high or where a large, near-surface clay layer exists, AEM depth penetration is limited and non-unique signals ("multipath") can cause misleading deep responses. "When you have a basin like the Indian Wells Valley... the skyam is not going to tell you anything," TAC member Don Decker warned, arguing that multi-path and strong salinity bands limit deep interpretation.

Groundwater-model dispute: The workshop also featured competing numerical models. Dr. Todd Concincaid summarized historical estimates and model calibration practices. He reviewed earlier mass-balance approaches (Conklin & Chase 1969), a DRRI/DGA-derived recharge estimate of about 7,860 acre-feet per year used in recent GSP updates, and the alternative higher inflow numbers the water district has proposed (roughly 14,000'15,000 af/yr). "If the total inflow to the basin is the higher number, then definitively everybody before was wrong," Concincaid said, adding that a proposal that large requires a model that demonstrably and substantially improves fit to observed groundwater levels and other observations.

Concincaid described how modellers build grids, assign hydraulic properties and calibrate by matching simulated heads to observed hydrographs; he argued that a radically different conceptual model must yield a markedly better match to field data to justify overturning previous estimates. He also warned that relying on "mining" stored groundwater (drawing beyond sustainable inflow) would have predictable consequences: widespread shallow well failures, declining water quality as saline or lower-quality water moves vertically, and rising pumping costs.

Why it matters: The recharge/safe-yield number used in policy and management determines allowable pumping and whether the GA must pursue imported water. The disagreement between modeling teams is therefore central to near-term decisions about conservation, acquisitions and whether the community faces steep costs to secure imported supplies.

What comes next: Presenters and attendees urged continued integration of AEM with high-quality well data, transparent model comparisons and public access to model inputs and calibration diagnostics. Several speakers encouraged additional targeted well or geophysical work in areas where AEM and well logs disagree to resolve key uncertainties.