Introducing the Regional Community Earth System Model, R-CESM

Dr. Yun Liu
Department of Oceanography, Texas A & M University
This presentation centers on the Regional Community Earth System Model (R-CESM), a high-resolution, fully coupled regional Earth system model system built upon the Community Earth System Model version 2 (CESM2) framework and incorporating the WRF CLM, and ROMS models. R-CESM is designed to improve understanding of climate variability, future projections, and extreme events at regional scales, offering flexibility in its design and operation, including various air-sea flux schemes. We showcase here a few preliminary studies that illustrate its novel aspects and value. High-fidelity simulations: A 9-year regional coupled simulation of the Gulf of Mexico demonstrated R-CESM's accuracy in reproducing mean Sea Surface Temperature (SST), precipitation, and ocean eddy kinetic energy (EKE). Tropical Cyclone (TC) insights: Simulations revealed that the WRF air-sea flux scheme, which caps drag coefficients at high wind speeds, yielded stronger TC intensities and more realistic surface pressure-wind relationships compared to the CESM scheme. Loop Current deep dynamics: R-CESM has been used to investigate Loop Current (LC) deep dynamics and eddy shedding in the Gulf of Mexico, identifying previously uninvestigated 8–16-day midwater mixed waves. These waves are stimulated by the interaction between the penetrating LC and steep topography, defining five distinct wave train scenarios. Crucially, a novel "west necking-down region" around 88.5°W, characterized by enhanced 8–16-day variance, was identified as being associated with LCE separations, offering new insights for predictability from lower ocean layers. Advanced Data Assimilation: An "online" ensemble coupled data assimilation (ECDA) capability has been implemented within R-CESM, integrating an Ensemble Kalman Filter (EnKF) algorithm for high computational efficiency by processing data in memory. Observing system simulation experiments (OSSEs) demonstrated that assimilating Sea Surface Height (SSH) observations is more effective than SST in constraining R-CESM's ocean states, leading to much smaller and more consistent analysis errors. This system successfully enabled a 12-day lead-time prediction of a Loop Current eddy-shedding event and a 25-day lead-time prediction of a reattachment event. Courtesy to: Dan Fu, Xiao Ge, Justin Small, Ping Chang, Steven Dimarco, Gokhan Danabasoglu, Jaison Kurian, Brian Kauffman, Abishek Gopal, Sanjiv Ramachandran, Zhi Shang, Katherine Thayer-Calder, Mariana Vertenstein, Xiaohui Ma, Hengkai Yao, Mingkui Li,,Zhao Xu, Xiaopei Lin, Shaoqing Zhang and Lixin Wu.

Видео Introducing the Regional Community Earth System Model, R-CESM канала Comunicación OceFísica