Treffer: Intrinsic Predictability From the Troposphere to the Mesosphere/Lower Thermosphere (MLT).

The atmosphere's flow becomes unpredictable beyond a certain time due to the inherent growth of small initial‐state errors. While many research studies have focused on tropospheric predictability, predictability of the middle atmosphere remains less studied. This work contrasts the intrinsic predictability of different layers, with a focus on the mesosphere/lower thermosphere (MLT, 50–120 km altitude). Ensemble simulations with the UA‐ICON model for an austral winter/spring season are conducted with a gravity‐wave‐permitting horizontal resolution of 20 km. Initially small perturbations grow fastest in the MLT, reaching 10% of saturation after 5–6 days, compared to 10 days in the troposphere and 2 weeks in the stratosphere. A saturation level of 50% is reached only after about 2 weeks in the MLT, similar to the troposphere. Saturation times are overestimated in a coarser resolution model (grid size 160 km) by up to a factor of two, highlighting the need for gravity wave‐resolving models. Predictability in the MLT depends on horizontal scales. Motions on scales of hundreds of kilometers are predictable for less than 5 days, while larger scales (thousands of kilometers) remain predictable for up to 20 days. This scale‐dependent progression of predictability cannot be explained by simple scaling for upscale error growth. Vertical wave propagation plays a significant role, with gravity waves transmitting perturbations upward at early lead times and planetary waves enhancing long‐term predictability. In summary, the study shows that MLT predictability is scale‐dependent and highlights the necessity of high‐resolution models to capture fast‐growing perturbations and assess intrinsic predictability limits accurately. Plain Language Summary: The atmosphere's movements are difficult to predict after a certain time because small errors in the initial conditions grow over time. While predictability of weather close to the surface is much studied, less is known about how predictable the atmosphere is at higher layers. This study explores how predictability differs in the atmospheric layers, with a focus on the mesosphere/lower thermosphere (MLT), located 50–120 km above Earth, using advanced numerical simulations that resolve smaller‐scale atmospheric waves. The findings show that initially small errors grow fastest in the MLT, making it less predictable than lower atmospheric layers. Typical predictability horizons in the MLT are 5–6 days, compared to 10 days in the lower atmosphere. However, motion in the MLT on larger horizontal scales of few thousand kilometers can remain predictable for up to 3 weeks. Simulations with lower resolution models are shown to overestimate predictability by up to two times. Furthermore, it is shown that MLT predictability is strongly influenced by the vertical propagation of waves from the lower atmosphere. This research highlights the importance of high‐resolution models to understand and predict the atmosphere at higher altitudes, in the MLT. Key Points: Predictability timescales vary by atmospheric layer, with longest predictability in the stratosphere and shortest in the MLTMLT predictability is influenced by vertical wave propagation, reducing small‐scale predictability and extending that of planetary scalesPredictability horizons revealed by gravity‐wave‐permitting models are overstimated by a factor of 2 in coarse resolution models [ABSTRACT FROM AUTHOR]

Copyright of Journal of Geophysical Research. Atmospheres is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)