Global change predictions rely on process-based mechanistic models representing the Earth system and its different components, as is the case of land surface models, which simulate processes occurring in terrestrial ecosystems. However, the development of such models has been traditionally biased towards mesic and humid ecosystems, and towards the mechanisms driving their functioning. As a result of this, dry ecosystems (i.e. ‘drylands’) have been underrepresented in land surface modeling, despite constituting a large fraction of emerged lands, which extension could also increase under climate change. Moreover, due to increasing droughts, even the functioning of non-dryland ecosystems can shift beyond some thresholds, making them resemble drylands. Therefore, it is expected that some mechanisms that are prevalent in dry ecosystems (i.e. ‘dryland mechanisms’) will increase its relevance worldwide.
How suitable are global land surface models to incorporate dryland mechanisms? In order to tackle this issue we face many challenges, from data availability to the current design limits of those models. Here a perspective will be discussed for several dryland mechanisms (e.g. photochemical degradation, soil hydrophobicity), the current state of their incorporation into different models and, as a particular example, the case of ORCHIDEE, one of the most widely used global land surface models. Future advances to update those models with dryland mechanisms could contribute to improve the accuracy of predictions of ecosystem responses to global change.
