Application and Evaluation of a Novel Python‐Based Ensemble Data Assimilation Framework NEDAS in Whole Atmosphere Community Climate Model (WACCM)

Abstract This study presents a new Python‐based whole‐atmosphere ensemble data assimilation (DA) framework that integrates the Next‐generation Ensemble Data Assimilation System (NEDAS) with the Whole Atmosphere Community Climate Model (WACCM). The system applies a Local Ensemble Transform Kalman Filter to assimilate multi‐platform temperature and wind observations from the troposphere to the lower thermosphere. We performed whole atmosphere cycling assimilation experiments from December 2018 to February 2019 using WACCM + NEDAS. Validation against independent temperature observations shows that the global root‐mean‐square error in WACCM + NEDAS temperatures was reduced to less than 5 K in the stratosphere and less than 15 K in the mesosphere and lower thermosphere (MLT), representing error reductions of over 70% relative to Freerun‐WACCM and over 50% relative to specified dynamics simulation of WACCM (SD‐WACCM). For dynamical features, the results indicate that WACCM + NEDAS successfully reproduced the temperature and zonal wind evolution during a major sudden stratospheric warming (SSW) event with elevated stratopause in the Northern (winter) Hemisphere, which is in close agreement with the Modern‐Era Retrospective analysis for Research and Applications version 2 (MERRA‐2) and the Japanese Atmospheric General circulation model for Upper Atmosphere Research with Data Assimilation System (JAGUAR‐DAS) Whole neutral Atmosphere Reanalysis. Notably, WACCM + NEDAS shows substantial improvements in mesospheric cooling during SSW compared to SD‐WACCM. In addition, the quasi‐two‐day wave activities in the Southern (summer) Hemisphere MLT region are also well reproduced by WACCM + NEDAS, which cannot be captured in either Freerun‐WACCM or SD‐WACCM.