Published evidence pumping testsmodel factordecision propagation

Transformation Uncertainty

A measured response becomes a design number after a model translates it.

Definition

Groundwater studies often treat aquifer-test parameters as inputs to later uncertainty analysis. Transformation uncertainty concerns the step that converts field responses into inferred parameters.

The measured object is the drawdown or recovery response. Transmissivity and storage are interpreted with an aquifer-test model: Theis, leaky aquifer, lagging Darcy, dual-porosity, a numerical inverse model, or another response model. Each analytical model carries its own bias, dispersion, identifiability limits, and support scale.

Drawdown, recovery, and temperature are measured. Aquifer and thermal properties are inferred with analytical or numerical interpretation models. Transformation uncertainty follows the chain from measured record to apparent parameter and then to an engineering decision variable.

Role of Data Methods

Data methods can support record organization, candidate-model comparison, sensitivity checks, diagnostic review, and evidence organization. Their results still require analytical-model checks, identifiability analysis, held-out prediction, and decision-level uncertainty propagation.

Evidence Boundary

Current evidence supports the following cautious statements:

pumping-test transmissivity and storage are inferred parameters produced by an interpretation model;
the study reviews front-end drawdown-to-parameter transformation uncertainty;
a 10,000-scenario numerical benchmark provides conditional transformation model factors for four aquifer-test interpretation models;
field cases provide diagnostics, not proof of true aquifer parameters.

This work complements established groundwater uncertainty methods. It does not treat one analytical model as a replacement for established hydrogeologic interpretation.

Why It Matters

Engineering decisions inherit interpretation choices. A fitted parameter can look precise while still carrying transformation error from model form, time window, boundary simplification, and field support.

Applied work identifies where interpretation uncertainty changes allowable pumping, response time, recovery planning, or a design threshold.

Calculation-backed concept animation

One normalized LDL response can lead to different aquifer-test interpretations.

This explainer uses a simplified Lin-Yeh (2017) lagging-flow special case for every plotted curve. The animation shows normalized response shape, but the lines are still generated from model calculations.

Scene 1

The curve is a normalized LDL response.

The browser first computes drawdown from a simplified line-source special case of the Lin-Yeh lagging-flow equation, then normalizes the response to reveal the transition shape.

Normalized LDL response computed, not sketched

τ_q = 0, τ_s = 0

Classical Darcy

τ_q = τ_s

Equal lag

delayed transition

τ_s > τ_q

larger τ_s − τ_q separation

Decision case

Calculation Lin-Yeh special case

All response curves are generated in the browser, then normalized to compare response shape.

Transformation analytical model

The curve becomes apparent parameters through assumptions about lag times, radius, and leakage length.

Decision effect threshold shift

Allowable pumping, recovery time, or design threshold changes when inferred parameters change.

The plotted curves are calculation-backed teaching cases: the Classical Darcy limit, equal-lag cancellation, separated lag with τ_s greater than τ_q, and a stronger decision-sensitive case. They are not hand-drawn concept curves.

Values are normalized for communication and use a public teaching implementation of the no-wellbore-storage, no-skin, line-source special case of the Lin-Yeh (2017) leaky-confined lagging-flow equation with a 12-term Gaver-Stehfest inversion. This is not the full finite-radius, wellbore-storage, and skin solution used for calibrated field analysis, and it is not an engineering recommendation. Symbols follow Lin and Yeh (2017): τ_q is the water-flux lag and τ_s is the drawdown-gradient lag.

Collaboration entry

Bring the aquifer-test or TRT record that remains unresolved under its current interpretation model.

A useful starting point is a specific issue: delayed recovery, phase-amplitude disagreement, TRT drift, boundary-condition sensitivity, or residual structure that may change a pumping limit, recovery schedule, design threshold, or energy design.

Send the forcing history, the measured drawdown, recovery, head, or temperature record, the suspected model-assumption issue, and the decision that depends on the interpretation.

Send a data/model brief Choose a collaboration mode