The publication record traces how the questions and methods have developed.

The method maps which places and forecast times most influence a tropical-cyclone intensity prediction, helping researchers judge whether the model relies on physically sensible signals.

The analysis keeps grout heat storage in a thermal response test and separates early-time grout effects from later-time ground-conductivity information.

Two groundwater processes can respond at different rates. Accounting for both reconciles the observed amplitude attenuation and phase shift of periodic head signals at the two field sites.

A simplified fault interface reproduces the full two-dimensional drawdown closely for most tested conditions when the fault is not too wide, making it useful for rapid analysis.

Pumping-induced pressure changes at the two sites are predicted to bend the overlying formation by less than routine geodetic instruments can detect.

At the tested sites, allowing groundwater flux and hydraulic gradient to adjust at different rates improved the pumping-test interpretation and changed the inferred response times.

A stream does not supply a pumping well instantly or indefinitely. Including finite stream storage and delayed exchange improves the field interpretation, especially for smaller streams and certain time periods.

This analytical solution uses image wells to represent pumping, recharge, and coastline geometry in island aquifers.

In two field tests, allowing thermal dispersion to increase with groundwater velocity gave a more realistic account of heat movement than a constant-dispersion model.

A compact fault model represents both resistance through the fault core and rapid flow along fractured damage zones.

Using drawdown from several observation wells, the method recovered a nearby hydraulic boundary together with aquifer transmissivity and storage; three wells performed about as well as four.

Accounting for short well screens and low-permeability layer boundaries produced more plausible aquifer properties at the field site than assuming fully penetrating wells.

The water-table condition keeps a time scale for gravity drainage, which helps avoid treating unsaturated flow as instantaneous by default.

This technical reply shows that the published transport solution gives the wrong long-time concentration and identifies the equations that require correction.

A physics-informed neural network reconstructed the tested wave fields accurately from incomplete and noisy data after its spatial and temporal scales were optimized.

For a highly permeable well skin, head remains continuous but flux can change across the interface; the usual low-permeability skin condition is therefore not valid.

A simplified interface captures the main drawdown behavior of a fully resolved high-permeability zone around a well while requiring less computation.

For the tested borehole thermal energy storage systems, image wells matched finite-element temperatures at much lower computational cost and revealed how boundaries alter heat storage.

A lagging-response model inferred plausible unconfined-aquifer properties from groundwater levels even when detailed pumping rates and well locations were unavailable.

Because a stream stores only a finite amount of water, its stage changes during pumping and its support to the well weakens with time.

A fast analytical calculation estimates freshwater-lens shape and pumping vulnerability for multiple wells and changing recharge.

Including both near-well skin and rate-dependent well loss reproduces the sharp peaks in the field data that simpler well models miss.

Hourly groundwater levels were used to reconstruct regional pumping of 1.71-2.05 billion cubic metres per year, with patterns consistent with agricultural demand and land subsidence.

In the dual-porosity model, stronger exchange shifts more of the response into fractures and less into the rock matrix near the coast.

A standard pumping-test model can estimate aquifer storage from observation-well drawdown even when the individual pumping wells are unknown.

A signal-decomposition method recovered where pumping occurred and how pumping rates changed from dense groundwater-level records in the synthetic test.

A two-time-scale transport equation reproduced the two-stage breakthrough behavior observed in both laboratory columns.

Soil type and the lower thermal boundary increasingly affect temperature profiles over time; ignoring the lower boundary can bias groundwater-flux estimates.

For three rainfall patterns, the analytical model reproduced measured infiltration rates and matched the numerical temperature profiles.

When flux responds more slowly than hydraulic gradient, pumping can produce wave-like drawdown and temporary recharge overshoot; when storage responds more slowly, the behavior resembles delayed drainage or dual porosity.

Temperature profiles provide time-varying estimates of vertical groundwater flux instead of imposing a constant-flux condition.

For the tested wells, the combined model best explained the data; at LA-87B, non-Darcy well loss mattered more than formation damage.

A generalized model let the data determine the relative influence of two limiting behaviors and fit both pumping records better than fixing either limit in advance.

The analytical transport solution matched the laboratory concentration data and estimated transport parameters without requiring the extra batch tests used by the comparison method.

A delayed stream boundary approximates an explicitly modelled streambed efficiently, although the difference depends on pumping time, well distance, and streambed properties.

Applying the fitted well-loss criterion indicates that both test wells were inadequately designed or developed.

During the tested injection period, the coupled heat-transfer model reproduced the measured temperature distribution at an aquifer thermal energy storage site.

Allowing groundwater flux and hydraulic gradient to lag by different amounts fitted drawdown at three sites and produced specific-yield estimates consistent with laboratory ranges.

A pumping-test model is used to examine whether radial flux and drawdown gradient need to respond at the same instant.

An analytical constant-head model that includes unsaturated drainage and leakage from a lower layer agreed with numerical simulations and field observations under the tested conditions.

In a finite two-zone aquifer, the outer boundary has little effect early in a pumping test but controls late-time drawdown and well flow.