The Impact of Differing Land Surface Models and Water Isotopic Parameterizations to the Distribution of Water Isotopes in a Coupled Atmosphere-Land Global Climate Model.

Abstract:

Isotope-enabled Global Climate Models (GCMs) are becoming important tools in facilitating the synthesis of disparate isotope proxy data, allowing for uncertainties in proxy-based reconstructions to be tested in a way not possible with inversion methods. They also provide a means to test processes and parameterizations in the GCMs themselves, as new in-situ and remote sensing systems now can measure water isotopes at the spatial and temporal scale needed to validate global models. However, one issue with isotope-enabled GCMs is that much of the past focus and development has been on the atmosphere and ocean, which means other components of the earth system are poorly understood in comparison. Newly developed isotope-enabled GCMs with fully-functional land surface models, along with new observational platforms, allow for one to examine the importance of the land surface on the distribution of water isotopes in the earth system. We report here on experiments using the new NCAR isotope-enabled Community Atmosphere Model version 5 (iCAM5) and the isotope-enabled Community Land Model Version 4 (iCLM4), as well as a growing number of measurements of isotopic ratios in precipitation and water vapor. In particular, iCAM5 is used to simulate the modern isotopic climate coupled to a. a simple bucket model for isotopes, b. iCLM4 with equilibrium fractionation only, and c. fully-fractionating iCLM4. Along with the use of iCLM4, numerous variations in the representation of kinetic fractionation are examined, as well as different parameterizations for the impact of dew and frost on the isotope ratios in the surface water vapor, snow, and soil moisture. Results show that having a fully-functioning land surface model has a large impact on the simulated isotope ratios, and is necessary if one wants to simulate water isotopes in the earth system accurately. Accurately simulating d-excess and O17-excess requires having a kinetic fractionation factor that properly accounts for the influence of both turbulent and kinetic diffusion in the surface layer. Finally, matching observed isotope ratios in the polar regions requires an adequate simulation of the influence of dew and frost. These findings show that with the observational constraints now available, models of water exchange between the atmosphere and land can be improved.