Undeclared baselines change industrial pollution indicators fivefold: a dual-basis reporting protocol tested on continuous factory monitoring

Environmental indicators divide a measured burden by activity and are read against a reference condition. Ecological assessment and groundwater regulation require that condition to be declared; concentration indicators do not, although infiltration and earlier activity leave a floor persisting once activity stops. The burden must be integrated as measured or above that floor, a choice almost never recorded; what the omission costs is quantified here. Electrical demand, operating state and twelve sensing channels were monitored at 5-min resolution over 150 days at three factories in Istanbul, Türkiye, machining, electronics assembly and moulding, giving 129,603 observations. A validity screen admitted 16 of 36 site-channel combinations, two exclusions resting on a pairwise independence test. Across 13 well-determined combinations the reference choice changed the burden attributed to non-productive operation by a median factor of 5.4, and between 3.9 and 10.0 across baselines from the 1st to the 25th percentile. The same floor biased concentration-load relationships, 15 of 16 appearing sub-linear until the baseline was removed, reversing 9. The effect exceeds the change a one-fifth cut in idle load makes to the non-productive energy fraction. Of 11 comparable studies, none of the five measuring concentration declared its basis. Factory calibration alone leaves absolute levels indicative; the state comparisons carrying the argument are unaffected. Dual-basis reporting answers this in four steps from records already held, costing one column and two sentences of method. Both follow from the algebra of a measurand with a positive floor, transferring to water quality indices, ambient air and soil metrics.