This is a long overdue roadmap to track progress towards v0.2, which we hope to be the first scientifically "useful' release of Terrarium.

v0.1.x

• Improved documentation Major overhaul of documentation #84, Documentation fixes and improvements  #102
• Coupling LandModel with a SpeedyWeather atmosphere via full SEB Coupling LandModel with full surface energy balance to SpeedyWeather #110
• Iterative solvers for SEB Investigate and fix possible bugs in surface energy balance calculations #135
• Parameter handling Add initial implementation of parameter handling for all processes #123
• Make timesteppers independent from caches Split timestepper and cache and revise Heun #122
• Heterogeneous soil texture Add initial implementation of heterogeneous soil properties #131
• Reactant compatibility for SoilModel Reactant compatibility  #138
• Prescribed LAI/phenology vegetation model (see Towards a fully functional vegetation model #137, Prescribed phenology scheme #146 )
• Simple single-layer snow scheme Simple single-layer snow scheme with SEB coupling #141 Add simple single-layer snow scheme #144
• Finish addressing Issues with bloated model state and long compile times #66
• Add benchmarking suite Add performance benchmarks #163
• Improved integration with Oceananigans grids (Full integration with Oceananigans AbstractGrid interface with support for multi-domain vertical discretizations #178, Make AbstractLandGrid subtype AbstractGrid #179)
• Fix issues with snow (Investigate issues with snow melting too fast in coupled LandModel simulations #177)

v0.2

• Multi-scale timestepping (Add support for multi-scale time stepping #147)
• Working LUE vegetation model enabled by multi-scale timestepping (see Towards a fully functional vegetation model #137)
• Sub-daily vegetation carbon modeling via Farquhar model (see Towards a fully functional vegetation model #137)
• Implicit timestepping for soil energy/hydrology
• Multi-grid support (soil, snow, canopy) (Full integration with Oceananigans AbstractGrid interface with support for multi-domain vertical discretizations #178)

This is a non-exhaustive list; I will add more things as I think of them. Additional suggestions or comments are of course welcome.