Reference-grade technical material: equations, algorithms, numerical methods, and design rationale. The user guide explains how to use Kalix; this section explains what it does and why.

Planned contents

Scientific reference

• Node equations — per node type, document the governing equations, parameters, units, and assumptions
• Routing methods
• Rainfall-runoff models (GR4J, others)
• Storage dynamics
• Loss and gain processes

Numerical methods

• Timestep semantics and stability
• Solver choices and tolerances
• Mass balance and conservation checks
• Handling of missing data and edge cases

Calibration & optimisation

• Objective functions (NSE, KGE, logNSE, …)
• Optimisation algorithms available in Kalix
• Warm-up, parameter bounds, constraints
• Multi-objective methods

Data & I/O

• Data cache internals