Search

Abstract State‐of‐the‐art climate models are extremely useful tools, but they all have biases when simulating tropical cyclone (TC) genesis location. These biases contribute to discrepancies in simulating other characteristics of TC activities (e.g., track, intensity, and duration) and even the surrounding large‐scale circulation. The genesis potential index (GPI) is a useful tool to evaluate the performance of climate models in simulating TC genesis. Here, it is employed to investigate possible causes of simulation biases of TC genesis location in western North Pacific using the downscaled Community Atmosphere Model Version 5 (CAM5) climate model and observations. We find that GPI can reasonably describe the spatial pattern of TC genesis location in either observation or simulation. By having one of GPI‐related variables (i.e., low‐level absolute vorticity, middle‐level relative humidity, vertical wind shear, and potential intensity) change with the others set to stable while testing single variable, the contributions of environmental fields associated with GPI are examined and a possible cause of TC genesis biases is revealed. Due to the overestimated humidity in the east region of western North Pacific, the simulated GPI high value region is to the east of the observed one, and more TCs in the downscaled CAM5 model generate farther east than in observation. These results imply that the biases in simulated humidity and thus TC genesis location may be reduced by correcting the evaporation rate in the deep convection scheme, since the evaporation rate is closely related to environment humidity change as suggested by previous studies.

State‐of‐the‐art climate models are extremely useful tools, but they all have biases when simulating tropical cyclone (TC) genesis location. These biases contribute to discrepancies in simulating other characteristics of TC activities (e.g., track, intensity, and duration) and even the surrounding large‐scale circulation. The genesis potential index (GPI) is a useful tool to evaluate the performance of climate models in simulating TC genesis. Here, it is employed to investigate possible causes of simulation biases of TC genesis location in western North Pacific using the downscaled Community Atmosphere Model Version 5 (CAM5) climate model and observations. We find that GPI can reasonably describe the spatial pattern of TC genesis location in either observation or simulation. By having one of GPI‐related variables (i.e., low‐level absolute vorticity, middle‐level relative humidity, vertical wind shear, and potential intensity) change with the others set to stable while testing single variable, the contributions of environmental fields associated with GPI are examined and a possible cause of TC genesis biases is revealed. Due to the overestimated humidity in the east region of western North Pacific, the simulated GPI high value region is to the east of the observed one, and more TCs in the downscaled CAM5 model generate farther east than in observation. These results imply that the biases in simulated humidity and thus TC genesis location may be reduced by correcting the evaporation rate in the deep convection scheme, since the evaporation rate is closely related to environment humidity change as suggested by previous studies. Key Points: The overestimated humidity contributes to the eastward bias of tropical cyclone genesis location simulated by downscaled CAM5 model [ABSTRACT FROM AUTHOR]