Terrestrial Carbon Dynamics through Phenology Modeling and Data Assimilation
A GMU proposal submitted to NASA-ROSES-2012 A45: CMS: Carbon Monitoring System Program (NNH12ZDA001N-CMS), $378,122, 3 mo/yr, 07/01/2012-12/31/2013.
Knowledge of carbon cycle dynamics is a basic ingredient for characterizing and predicting the behavior of the Earth’s changing climate system and to inform policy development and planning. Predicting these carbon dynamics requires understanding how the various components of the carbon cycle functioned in the past and present. Recent innovation in remote sensing techniques, intensive in-situ carbon observation campaigns, terrestrial carbon and phenology modeling, and land surface data assimilation provides the tools we need to make significant advancement towards better understanding and predicting terrestrial carbon and phenology dynamics. Therefore, I propose to use innovative new terrestrial modeling, data assimilation (DA) and data fusion techniques to bring together diverse carbon-cycle observations into an ensemble modeling framework that will provide a consistent terrestrial carbon-water-energy budget.
Specifically, the Land Information System (LIS) framework will be upgraded to integrate remotely-sensed phenology observations, ensemble modeling, analysis and data assimilation capabilities to quantify the contemporary North American carbon budget and associated uncertainties. The Community Land Model Version 4 (CLM4) will be incorporated into LIS, and implemented in an ensemble reanalysis mode to produce a 1km, hourly resolution terrestrial product suite for the period of satellite record (~1980 to present). The model ensemble will be constrained with key remotely sensed vegetation observations, including albedo, leaf area index (LAI), fraction photosynthetically active radiation (FPAR), gross primary product (GPP), surface moisture, wetlands delineation, and drainage classes. The project will take advantage of NASA remotely-sensed databases on land cover, land use, land management, urban areas, vegetation type, age classes, soil properties, topography, climate, atmospheric deposition and natural disturbances for improved model parameters. Key biomass and flux data available from the NASA Carbon Monitoring System (CMS) will be used for calibration and validation and for assessing data assimilation product uncertainties. The project will utilize well-established standards and services for land surface data product dissemination, documentation and quality assessment to disseminate all research results in a timely manner in accordance to the NASA Earth Science Data Policy.
The resulting terrestrial LIS-CLM4 reanalysis will provide validated, continuous gridded time series of carbon, water and energy fluxes and storages for 30+ years of record. Data assimilation analysis increments and ensemble errors will also be provided for budget closure and uncertainty assessments. These products will provide a credible historical baseline that will help to inform carbon management and policy decisions.
This work will also directly contribute to the goals of the NASA Carbon Monitoring System, by improving the (1) monitoring and prediction of terrestrial carbon storage, (2) further integrating NASA products and capabilities in a data assimilation context, (3) establishing product uncertainties and accuracies, (4) harnessing unique capabilities of NASA Centers (LIS, DA, CMS and satellite products), and (5) rapid generation and distribution of products. Further, the proposed framework will provide regionally-relevant information while being readily scalable to future NASA missions, collaborative CMS products, and to global land surfaces. The proposed framework will provide an excellent integration platform for shaping and contributing to the next phase of development of the NASA CMS.