Environmental Science and Engineering Seminar

Despite major advances in Earth system modeling, substantial gaps persist between observed and simulated carbon fluxes across land and ocean, with the largest discrepancies on land. Bridging these gaps requires improved process understanding, tighter integration of observations, and targeted model innovation. First, I examine the strengths and limitations of current process-based land surface models in representing net CO₂ fluxes, highlighting where, and why, they diverge from observational constraints. Second, I propose a strategic framework for advancing these models, emphasizing improved representation of critical processes including radiative transfer and nutrient acquisition, guided by remote sensing data and refined benchmarking practices. Finally, I present results from the CliMA Land model demonstrating how observation-informed refinements to model complexity, specifically through the integration of global vegetation trait variability into land surface radiative transfer schemes, substantially enhance simulations of surface albedo, energy balance, and carbon uptake. Together, these efforts underscore how advancing the representation of biogeophysical and biogeochemical processes, guided by observations, can narrow the carbon gap and strengthen the reliability of future climate projections.