CONTROL ID: 1810365

TITLE: Development and Application of a Benchmarking System for Land Models (Invited)

AUTHORS (FIRST NAME, LAST NAME): David M. Lawrence¹, James Tremper Randerson², Mingquan Mu², Forrest M. Hoffman⁴, William J. Riley³, Charles D. Koven³, Katherine E. Todd-Brown², Gretchen Keppel-Aleks²

INSTITUTIONS (ALL): 1. NCAR-CGD, Boulder, CO, United States.
2. UC Irvine, Irvine, CA, United States.
3. LBNL, Berkeley, CA, United States.
4. ORNL, Oak Ridge, TN, United States.

ABSTRACT BODY: There is a widely understood need for improved assessment of terrestrial system models that are utilized in climate, weather, and ecosystem modeling. A more comprehensive and more systematic land model ‘benchmarking’ process should help speed the development of new parameterizations, improve the design of new measurement campaigns, and reduce uncertainties associated with key land surface processes. However, the development of informative metrics is non-trivial. Past data-model intercomparisons have certainly strengthened the representation of key processes in land models, but often this information has not been easily accessible for use by other modeling teams or in future intercomparisons. Further, the development of sophisticated model diagnostics programs—that can fully exploit the richness of large Earth System data sets like satellite or Fluxnet measurements—are outside the scope of any single investigator. Here, we will describe progress in the development of a comprehensive (and extensible) land model analysis system that spans water, energy, carbon, and vegetation dynamics metrics. Several classes of metrics are employed including (1) large-scale state and flux estimates (e.g., soil C, atmospheric CO₂ cycle, surface temperatures, GRACE water budget estimates, river discharge); (2) functional responses of explicitly modeled and emergent processes (e.g., stomatal responses to VPD, near-surface soil moisture responses to precipitation); and (3) experimental manipulation responses (e.g., N additions, FACE, warming). We will demonstrate the application of selected metrics to several versions of the Community Land Model (principally CLM4 and CLM4.5 and variants therein). Additionally, we will discuss how the analysis package, which is open source and modular, is designed to be extensible and improved by many different modeling and measurement teams.

INDEX TERMS: 1847 HYDROLOGY Modeling, 0466 BIOGEOSCIENCES Modeling.
(No Image Selected)
(No Table Selected)

Additional Details

Previously Presented Material: 30% presented at an AMS meeting

Contact Details

CONTACT (NAME ONLY): David Lawrence
CONTACT (E-MAIL ONLY): dlawren@ucar.edu
TITLE OF TEAM: