Tutorial Ch 4: Multiple mechanisms in the same host application

[Copilot]

Adds Tutorial Chapter 4 covering how to create and run independent MICM solver instances for different chemical mechanisms within a single host application — a common pattern in multi-domain atmospheric models.

Python tutorial (tutorials/4. multiple_mechanisms.ipynb)

• Defines two mechanisms in-code: A→B→C (Arrhenius) and a reversible D⇌E+F system
• Creates separate MICM solver + state per mechanism; advances both in the same time loop
• Side-by-side matplotlib visualization of both systems

solver_1 = musica.MICM(mechanism=mechanism_1, solver_type=musica.SolverType.rosenbrock_standard_order)
solver_2 = musica.MICM(mechanism=mechanism_2, solver_type=musica.SolverType.rosenbrock_standard_order)
state_1, state_2 = solver_1.create_state(), solver_2.create_state()

while curr_time < sim_length:
    solver_1.solve(state_1, time_step)   # 3-species A→B→C
    solver_2.solve(state_2, time_step)   # 3-species D⇌E+F
    curr_time += time_step

Fortran tutorial (fortran/test/tutorial/test_multiple_mechanisms.F90)

• Solver 1: configs/v0/analytical (6-species with Arrhenius + user-defined rates)
• Solver 2: configs/v1/analytical/config.json (3-species pure Arrhenius)
• Registered as tutorial_multiple_mechanisms in fortran/test/integration/CMakeLists.txt

Fortran docs (docs/source/user_guide/fortran/chapter4.rst)

• New chapter explaining independent solver instances, separate states, simultaneous time-stepping, and memory management
• Added to Fortran tutorial toctree

Tutorial renumbering

Existing tutorials 4–11 shifted to 5–12 to place the new chapter in its logical position (after user-defined reactions, before parallelization). Updated all cross-references and doc index links accordingly.