Inverse model/Silicate weathering/Springs

[sandrablanco]

Hello
I am trying to create some inverse models to help understand the evolution of water in springs that discharge into volcanic rocks.

I am hoping to see chemical evolution between a ephemeral spring (solution 1) and a perennial spring (solution 2). Rainwater is a major source of Cl, SO4 and NO3 if they were concentrated substantially by evapotranspiration and were not removed by biological processes.  The NO3 is reduced base on concentrations in the springs, but SO4  and Cl the atmospheric contribution could be the only source for  these ions. Moreover, halite and gypsum are not reported in the study area.

I have been reviewing the forum and understand that I need a  separate source / sink for each item included in the mole balance.  However in the script I do not have a source / sink for Cl, SO6 and  NO3. I thought of defining halite and gypsum as the mineral phases  in order to represent the atmospheric contribution, but this will influence other ions and not only Cl and SO6. Besides, the above  does not resolve the source / sink of NO3 and removing this component  increases the charge balance error.

In general my concerns would be
How can I represent the atmospheric contribution for Cl and SO4?
What approach could I handle with NO3?

Many thanks to anyone who can help me.

[dlparkhurst]

If you assume both Cl and SO4 are conservative, you have a fundamental incompatibility. Cl increases by about a factor of 4, whereas SO4 increase by a factor of about 1.2. So, there is no concentration factor that can achieve both changes in concentration.

Here is an input file that includes Cl, NO3, and SO4 as conservative. To get results you have to make choices about uncertainties. I have applied 0.2 uncertainty to everything except Cl, which has 0.5. This produces some models that you can consider. If you don't like them, you can reassign uncertainties or add additional phases to shift results, but there is no magic answer that will preserve all of the concentrations of both solutions.