mescla.reactions.residuals

Reading geochemical reactions out of mixing residuals.

A mixing model is a null hypothesis: this water is nothing but a blend of those waters. Where the hypothesis fails, the failure is informative. Species that systematically depart from the mixing prediction are not measurement noise; they are the record of what happened to the water between the source and the sampling point.

The sign convention is the one that matters:

• measured > predicted – the system gains the species: mineral dissolution, desorption, cation-exchange release, mineralisation of organic matter.

• measured < predicted – the system loses it: precipitation, sorption, redox consumption, degradation, cation-exchange uptake.

Tubau et al. (2014) read the Besos Delta aquifer this way: excess Mg, Ca, Br, F, As, Fe and B against the river mixing line (carbonate dissolution and water-rock interaction) and deficits in O2, NO3, TOC, K and PO4 (redox processes plus ion exchange).

Functions

classify_source_sink	Label each species conservative, source or sink.
fit_conservative	Fit mixing ratios on the conservative tracers, carrying the rest along.
measured_vs_predicted	Per-species comparison of measured and mixing-predicted concentrations.
reaction_report	A readable summary of what the residuals imply -- paste it into your notes.