Temporal and Spatiotemporal Modeling: More Basics

Happy new year, everyone. I hope you enjoyed the turning over of a new year. This week was a short one, since the lab reopened today for business, so I thought I would continue the modeling basics blogs and talk this week about temporal and spatiotemporal modeling. If you haven't read the last blog on spatial modeling, check that one out before you continue here, as some of the information will resurface in the spatiotemporal modeling discussion.

Temporal modeling--or modeling that involves aspects of time--is the Ecosim portion of our ecological modeling program. Rather than model simple ecosystem interactions with the organisms in our model, we can choose to evaluate changes in our model groups over time. Running the simulations across time allows us to evaluate fluctuations in biomasses, which become interesting as some model groups increase in size while others decrease. Imagine a scenario where a key organism gets dramatically reduced in the model. We can evaluate how this change persists throughout time and if the model can remain balanced or if the system crashes because of the change in biomass. Additionally, modeling the ecosystem across time allows us to evaluate how different magnitudes of internal (trophic dynamics, fishing) or external (environmental) pressures destabilize ecosystem dynamics, which can provide helpful insights to management and policy efforts. 

Spatiotemporal modeling takes the temporal modeling and adds in the spatial modeling I mentioned a few weeks ago. As a refresher, spatial modeling uses a map of the area of interest that is made of environmental conditions (temperature, salinity, etc.) and models how both the interactions between model groups and their responses to the environmental conditions drive the health of each model group. If a model group is under a lot of predation pressure and is also stressed because of environmental conditions, it likely won't survive in the current grid space (cell) it occupies. For spatiotemporal modeling, we create a map of the environmental conditions for each time step of our model. If we want a really coarse or general model, we can provide an average of the environmental conditions for each year of the model time series and the model will grab a new map of the area of interest every year. If we want a more precise model, we can provide more maps per year to truly model how the environment is changing over time. As I have said many times, the more complex we want our models, the more computing power we need to run our calculations, so a modeler often has to weigh the costs and benefits of their model complexity.

Hopefully this rather brief overview of the types of models helps you understand how we evaluate changes in ecosystems across space and time. I'm hoping that this year I will be able to share some of the models I'm working on, though I will need to wait to get clearance to share some pictures from this work. In the meantime, I will continue to write about small portions of this work and how I am working to fill in missing data gaps to improve our models' predictive efforts. Stay tuned.