Seminar: Sujan Koirala

The terrestrial biogeochemical models used to represent vegetation-water-carbon interactions differ in complexity and parameterization, often dictated by the target mechanisms and processes and spatial scale of model application. Typically, model complexity and process representations reduce when the spatial scale increases. Such simplification involves abstractions of processes that introduce empirical model parameters, that are often under-to-unconstrained, leading to large model uncertainties in simulations of water and carbon cycles, as shown in several climate model intercomparisons. The current era of rapid growth of satellite Earth Observation, observational networks, and well as observation-based estimates, therefore, provides unprecedented opportunities to improve the model representation and understanding of the key carbon-water cycle processes. Exploiting this key data-centric opportunity, the SINDBAD model-data-integration (MDI) framework was developed for the seamless integration of diverse observational data to constrain terrestrial water and carbon models of varying complexity. This presentation will focus on the recent applications and developments of SINDBAD. In particular, we will answer key questions on the importance of river water storage in model parameterizations at the global scale, and the process representations of carbon cycle processes from local to global scales. Additionally, the recent technical developments will be summarized with a peek into the potential SINDBAD-based opportunities to improve the understanding of the global water and carbon cycles.