Development of Hybrid 3-D Hydrological Modeling for the NCAR Community Earth System Model (CESM)
Collaborative Institutional Lead(s):
It is widely recognized that global land processes play a significant role in the terrestrial water cycle and biogeochemical cycle (e.g., CO2, CH4 fluxes). For this reason, global earth system models (ESMs) for climate studies include sophisticated land models. However, the governing equations in the vertical one- dimensional (1-D) direction only are solved for these land processes, and none of the current global ESMs includes 3-D hydrological processes. The project objective is to develop and evaluate the first hybrid 3-D hydrological model with a horizontal grid spacing of 1 km for the National Center for Atmospheric Research (NCAR) Community Earth System Model (CESM).
To solve the 3-D soil moisture Richards equation at a grid spacing of meters to tens of meters in ESMs would be 10-20 years away due to a lack of computing power and observational data. Therefore this project takes the initial step to develop a hybrid 3-D hydrological model for ESMs. Major model development tasks include:
- solve the horizontal pseudo 2-D hillslope storage Boussinesq equation for groundwater and separate equation for surface water with a grid spacing of 1 km over global land
- directly couple the (horizontal) 2-D governing equations with the vertical 1-D computation of soil moisture in the CESM land component [i.e., the Community Land Model (CLM)] – this coupling is named "hybrid 3-D" to differ from the full 3-D Richards equation which is much more expensive computationally
- develop the required global data of soil thickness and permeability
To substantially increase the computational efficiency, major computational tasks include:
- maintain the current structure of CLM by doing vertical soil moisture and temperature computations in land units and plant functional types (i.e., don't do CLM computations in each 1 km2 cell)
- develop efficient numerical schemes for the vertical 1-D and horizontal 2-D equations
- pre-process global 1 km land cover, plant functional type, elevation, and bedrock data
Major model evaluation tasks include the use of extensive observational data and high-resolution numerical solutions of 3-D Richards equation over small domains. All these tasks are based on this research team's extensive research efforts in the past 20 years. The proposed hybrid 3-D hydrological model will be the first one for global ESMs. It will require the combination of better understanding, high-resolution modeling, high performance computing, and an interdisciplinary team of experts. Hence this represents a project with high risk/high reward potential. The benefits of this project include: more realistic treatment of land processes that will feedback to the atmosphere; more realistic treatment of river discharge to global oceans; dynamic prediction of wetlands that are crucial for CH4 and CO2 fluxes; transferability to other regional and global climate models; and direct support of the Long Term Measure (in model component development) of the DOE Climate and Environmental Science Division (CESD).