Scaling Time Series Workflows: Arrays, Submodels, and Python Integration

This session focuses on methods for building transparent and flexible models by moving data logic out of spreadsheets and directly into the GoldSim workspace. We explore how to organize large climate datasets using array label sets and multiple series within a single element. This approach improves model navigation.

The presentation covers how to transform historical records into predictive tools through probabilistic time shifting. Instead of a simple replay of history, this technique allows the model to sample random starting years for each realization. This method helps you account for temporal variability and uncertainty when making decisions about storage requirements.

For advanced users, the webinar introduces the use of submodels to pre-process time series definitions at the start or end of a simulation. This is helpful for calculating annual means or deriving rating curves from historical data. The session concludes with a look at using the GSPY library to facilitate communication between GoldSim and Python. This integration provides a path for creating custom data exports and performing complex external analysis on simulation results while keeping the core logic within the GoldSim environment.