Factors affecting land surface model spin-up behavior in the high resolution North American Land Assimilation System (NLDAS)

The study focuses on developing an effective spin-up strategy for supporting the NOAA/NCEP's and NOAA/OHD's operational hydrological and land surface modeling missions, as well as furthering their support to the National Integrated Drought Information System (NIDIS), NOAA Hydrology Test Bed, and the NOAA Climate Test Bed. Accurate initialization of land surface and hydrological models is critical for correct hydrological and land surface predictions, because the process of a model adjusting to its forcing can severely bias land surface simulations. A thorough investigation into the factors impacting model spin-up is conducted in order to develop a sophisticated technique to generate optimal initial conditions for the 33-year (1979-2011) 4km Noah and SAC-HT retrospective simulations, and to provide guidance for real-time simulations.

It is important to recognize and reduce the required spin-up time so that as much usable land surface model output as possible can be obtained from the retrospective simulation. Therefore, we have designed and undertaken numerical experiments to evaluate the production and evolution of model initial conditions (ICs) through model spin-up methods. Previous studies approached this issue by spinning up a model over a single recursive year of forcing or over multiple-year forcing climatology. We seek to build on these previous efforts by concentrating on an effort to increase spin-up efficiency as our research priority. Factors affecting the spin-up efficiency (forcing data, land surface type) will be fully evaluated in this study, and we expect to develop an effective and efficient spin-up strategy for obtaining an optimal initial condition and quantify the spin-up time needed for land surface modeling. This enhanced spin-up methodology will have broad applicability, with findings benefiting both operational and research-type simulations conducted at both OHD and EMC, as well as in the broader land-hydrology research community.