Assessing Planetary Boundary Transgressions and Their Causes - Using the FRIDA System Dynamics Model

The planetary boundaries (PBs) framework defines a "safe operating space" based on nine key Earth system processes. Out of these, four are terrestrial and their primary driver of transgression is agriculture. To better understand how agricultural activities might further influence the terrestrial PBs, it is essential to model their drivers and interactions over time. A helpful tool for studying complex dynamic relationships like these are World-Earth models, in particular FRIDA because of its aim to provide an internally consistent representation of many societal and Earth system processes. In this study, we included within FRIDA the PBs for: biosphere integrity, land system change, freshwater use, biogeochemical flows and climate change. This allows us to quantify their temporal trajectories, identify drivers of their transgression, and explore their main interactions. In total, seven different PB control variables are implemented across the five PBs studied, using both directly related variables in FRIDA and proxies related to the calculations using assumed relationships based on literature. By running the FRIDA model in a scenario governed by endogenous model behaviour, the PB quantifications are validated against values documented in the literature. Since FRIDA is still under active development, this study should be seen as a first effort to integrate PB status quantification and analysis into such a model.

The results show strong agreement with independent, earlier, estimates of PB trajectories and in particular whether the PB’s control variables are in the safe operating space, the zone of increasing risk or the high risk zone. However, some notable differences still occur, which may be attributed to the proxies developed to account for some relevant processes not currently represented in FRIDA. We also explore the role of certain drivers of (single or joint) PB transgressions centred around agriculture and associated societal processes and behaviours such as diet. As a part of this, we illustrate that an unambiguous attribution of PB transgressions to any given driver is challenging given that the coupling of drivers leads to non-linear and dynamically evolving feedback processes. Overall, we demonstrate the general suitability of the FRIDA model for simulating PB trajectories, their drivers and interactions. For future studies potentially using the model to inform decision-making, we recommend implementing all PB control variables, if possible in a more spatially explicit manner and without the aforementioned proxies.