Modeling Seasonal Carbon and Water Dynamics in Dryland Ecosystems: Challenges for Process-Based Models

Terrestrial ecosystems play a key role in the global carbon cycle, and drylands cover most of the land area and are very likely to expand as climate change causes evaporation to increase faster than precipitation. At the same time, they are often overlooked, and currently models cannot reproduce seasonal patterns of the vegetation in drylands.

The main reasons for this problem are the complexity and diversity of the processes that govern the terrestrial ecosystems' functioning; they may be missing or under-represented in the models. Among such processes, one can first emphasize the response of vegetation to water availability and its asymmetry, as well as soil and plant hydraulics, fires, and additional groundwater sources.

This study uses the terrestrial biosphere model QUINCY, that simulates the energy, water and carbon balance and vegetation dynamics of global terrestrial ecosystems. Our simulations cover 15 sites in the USA, Australia and Europe, which were chosen as arid sites based on water balance and temperature. The results are validated with data from the FLUXNET database and the Copernicus projects.

It was found that, on average, the QUINCY model overestimates gross primary productivity (GPP) and sensible heat (SH) values in dry areas. Also, water use efficiency (WUE) according to the model is higher compared to observations, and evaporation fraction (EF) is lower at most of the studied sites. Sensitivity analysis improved model performance at only a few sites, suggesting inadequate representation of key dryland vegetation dynamics. These results can be used to understand what processes need to be modified or added to improve performance.