A great question came when we deal with numerical modeling in modern times or in the last 5 years, and that question is how much programming skills should a hydrogeologist - numerical modeler have? This question is above this one: Should a hydrogeologist - numerical modeler need to program in any language?

As a part of an socratic method attempt to answer this, we will split the question into some subquestions.

Why is programming a needed skill for a hydrogeologist - numerical modeling?

Wrong question: Because you evaluators / clients will be impressed with the code you have developed to build / run / analyze a groundwater flow model

Right question: We will deal with bigger baselines not only from piezometric heads but also from temperature, potential evapotranspiration, infiltration recharge, and this data need to be filtered and formatted to a model input. Numerical models require the interaction between other geological models or surface water models. We would need piezometric trends instead of just time series binary arrays, we would need to assess the interaction of certain boundary condition instead of total water balances, and we would need powerful 3d visualizations instead of 2d contour visualization.

Why hydrogeologists - numerical modelers don’t program, or why they don’t program that much?

I will speak in a general sense, as people that will read these are from all parts of the world. On a generic scope, on a normal higher education in hydrogeology, numerical modeling is offered through limited courses and programming isn't a regular course. The hydrogeologist - numerical modeler as a behavioral model will find programming as an interesting but foreign topic. Most of the hydrogeologists that program have been self-taught and we lack a common approach to program in hydrogeology.

Are hydrogeologists - numerical modelers required to learn Python?

We should be careful on the selection of programming languages. Currently Python is the most developed language with great tools for data processing, machine learning, raster processing, model construction and output visualization, however Python relies heavily on C and the performance on heavy matrix calculation is limited. We need to be open to current or future programming languages that can both solve the groundwater flow matrices and process the model input/output.

What will happen if hydrogeologists - numerical modelers keep not programming?

For most of the modelers nothing would really happen, since the current workflow of hydrogeological evaluations with groundwater modeling is kind of reactive to change, or the disbelief from stakeholders on numerical simulations limit the scope of the numerical simulations.

If hydrogeologists - numerical modelers keep not programming they will rely more on commercial software that offers a specific tool, or they will spend several hours on data processing that can be automated through coding.

By sure, if modelers keep not programming we won't make the next level of fully featured groundwater flow software where more hydrogeologists can create better evaluations. If we keep not programming we won’t make the software that will preserve our groundwater resources for future generations.