Analyzing the Instabilities in the Venus Atmosphere Using Bred Vectors

Barotropic, baroclinic, and Rossby-Kelvin instabilities exist in the Venus atmosphere, as revealed by previous studies using numerical models. This study aims to deepen our understanding of the instabilities in the Venus atmosphere using the breeding of growing modes (BGM) method for the first time. We conducted a 1-year model simulation as the control run. First, we conducted identical twin experiments in which random perturbations were added to the control run at the initial time, and they grew freely. Perturbations in the upper layer (UL, 70-100 km altitudes) grow faster at the beginning but saturate earlier at a low value compared to those in the lower layer (LL, 40-70 km altitudes). Next, we implemented the BGM method and conducted multiple breeding cycle experiments with different rescaling amplitudes and intervals, the two parameters of the BGM method. The bred vectors (BVs) from these experiments identified instabilities in various regions. With large rescaling amplitudes, the structure and amplitude of the BVs in the LL closely resembled the deviations in the control run, indicating the growth of BVs due to barotropic, baroclinic, and Rossby-Kelvin instabilities. Composite mean analysis shows larger BV amplitudes in the morning hemisphere at the 60-70 km altitudes in the mid-latitudes, indicating enhanced baroclinic instability by the thermal tide. Finally, we estimated the intrinsic predictability related to baroclinic instability for Venus is >1 month, which would be longer than that for the Earth of similar to 2 weeks. Plain Language Summary The Venus atmosphere has been studied by numerical models that take inputs such as temperature and winds and provide evolution. We applied the "breeding of growing mode" method to a numerical model to deepen our understanding of the dynamics of the Venus atmosphere. We use two independent model simulations starting with different inputs. When the differences become larger over time for some areas, it is a sign of instability in those areas. Instabilities can be studied by investigating how the differences change with time in various regions. We found that in the cloud layer at 40-70 km height, the difference grows slowly at the beginning. However, it becomes larger than the difference above the cloud layer at the 70-100 km height in the later stages, indicating that large-scale instabilities dominate in the cloud layer while small-scale instabilities dominate above. Additionally, we identified barotropic, baroclinic, and Rossby-Kelvin instabilities at high, mid, and low latitudes in the cloud layer, respectively. Thermal tide due to solar heating impacts the instability around the cloud top (similar to 70 km) in the mid-latitudes. Finally, the predictability of the Venus atmosphere would be longer for the Venus atmosphere (>1 month) than that for the Earth's atmosphere (similar to 2 weeks).