New method predicts permafrost changes under climate change
A hybrid modeling approach provides fast and stable forecasts for permafrost behavior.
Researchers developed a way to simulate and forecast permafrost behavior. This method is fast and reliable. It helps address high computation times and energy costs.
The study used a numerical cryohydrogeological study on water and energy transfer. This was done at a forested, continuous permafrost site. The study used parallel high-performance computing.
A spectral network was used to create a surrogate model. A surrogate model is a model used for simulations to forecast over long periods of time. The researchers also used a hybrid modeling framework. This framework assimilates experimental data to enhance the surrogate.
The hybrid model improves physical simulation accuracy. It remains stable during integration and forecasting. The model was tested against a test set of measurements.
Predictions for the next decade were made using CMIP6 climate change scenarios. The results show a higher-than-expected temperature increase in the permafrost. The method demonstrates stable forecasting of climate change effects.
Key facts
- The hybrid modeling approach improves simulation accuracy.
- Permafrost shows a higher-than-expected temperature increase.
- The method provides stable forecasting over long periods of time.