Mining science workbench
In-app calculators run on prepared CSV/table inputs (optical screening, survey volumes, water balance, freeboard, lab review, trend). Heavy engines — InSAR, groundwater, geochemistry, runout, dust — run as isolated workers on authorised data; this workbench plans, catalogues and reviews them, it does not run them in the browser.
Geometric freeboard
Lowest surveyed crest − pond level, one datum. No wave/settlement/design allowance — those are separate engineering inputs.
Prescribed water balance
Period-by-period inflow − outflow (catchment excludes the pond). Storage is not pond level. Seepage is a measured/modelled input.
Optical change screening
NDVI/NDWI candidate change from prepared same-season L2A cells; SCL cloud mask on both dates. A candidate signature, not a cause.
Survey DEM volume
Cut/fill from after−before elevation on co-registered surfaces + a limit of detection. Apparent surface change, not mass; no integrated-volume uncertainty.
Laboratory review
Censored results stay qualified (never zeroed); medium + unit must match a reference; comparisons are to user references only — not legal limits.
Historical trend
Least-squares description of a monitored series (≥3 times). Never a failure forecast; radar LOS stays LOS, not vertical.