Assimilating Leaf Area Index and Aboveground Biomass into the Community Land Model to Constrain Carbon Dynamics in the Arctic and Boreal Region
Abstract
Climate change in the Arctic and Boreal Region is unfolding faster than anywhere else on Earth as indicated by the arctic amplification that air temperatures in the ABR have been increasing 2 to 3 times faster than globally. Process-based land models such as the Community Land Model (CLM) can be utilized to help us understand and predict the terrestrial carbon cycle under climate change in this region. However, CLM overestimates leaf area index (LAI) and above ground biomass which are important in estimating terrestrial carbon fluxes, allocation and storage. In our study, we assimilated 8-day MODIS LAI and annual aboveground biomass observations into CLM5.1 from 2011 to 2019 using an Ensemble Adjustment Kalman Filter in the Data Assimilation Research Testbed (DART) system. We compared the results from the assimilation run with a free run experiment in which CLM was driven with the same 40 reanalysis forcing ensembles but no observations were assimilated. The results show that assimilating observations of LAI and biomass reduces CLM estimates of LAI and aboveground biomass by 59% and 72%, respectively. This improves model estimates of carbon and water fluxes including gross primary production, ecosystem respiration, transpiration, and sensible and latent heat compared with the validation data sets in the International Land Model Benchmarking (ILAMB) system. Although all vegetation carbon pools decrease as a result of the assimilation, the decrease is not proportional as shown by relatively less leaf and stem carbon but more root carbon, indicating a change in carbon allocation within the vegetation. The reduced CLM estimate of canopy height is in good agreement with an independent canopy height dataset derived from ICESat, suggesting the change of aboveground biomass in CLM is reasonable. The smaller vegetation carbon pools, especially those with slow turn-over, are expected to provide good initial conditions for Earth System Models, potentially improving long-term carbon dynamics and climate projections.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2022
- Bibcode:
- 2022AGUFM.B46B..02H