Evaluation of Stratospheric Temperature Simulation Results by the Global GRAPES Model

Abstract

Global final analysis (FNL) products and the general circulation spectral model (ECHAM) were used to evaluate the simulation of stratospheric temperature by the global assimilation and prediction system(GRAPES). Through a series of comparisons, it was shown that the temperature variations at 50hPasimulated by GRAPES were significantly elevated in the southern hemisphere, whereas simulations by ECHAM and FNL varied little over time. The regional warming predicted by GRAPES seemed to be too distinct and uncontrolled to be reasonable. The temperature difference between GRAPES and FNL (GRAPES minus FNL) was small at the start time on the global scale. Over time, the positive values became larger in more locations, especially in parts of the southern hemisphere, where the warming predicted by GRAPES was dominant, with a maximal value larger than 24K.To determine the reasons for the stratospheric warming, we considered the model initial conditions and ozone data to be possible factors; however, a comparison and sensitivity test indicated that the errors produced by GRAPES were not significantly related to either factor. Further research focusing on the impact of factors such as vapor, heating rate, and the temperature tendency on GRAPES simulations will be conducted.