Clouds over the Highest Mountains Dance a Little Differently, by Rebecca Dzombak (Eos, 16 October 2026), summarizes a study by Zhang et al. in AGU Advances on how mountain clouds, especially supercooled ones, are responding to climate warming.

Most cloud- climate research focuses on oceans. This work instead examined long- term (2003- 2022) satellite observations and climate reanalysis data for cloud fraction, cloud- top temperature, and cloud- top height over major mountain ranges (Tibetan Plateau, European Alps, Andes, Appalachians, Rockies) and their adjacent lowlands.

Key findings

• Over most mountainous regions, warm- liquid and ice cloud phases increased, while the supercooled phase (liquid water existing below freezing) decreased. Vertical cloud development was generally suppressed.
• Ice cloud tops over mountains rose and cooled 2- 4 times faster than over nearby lowlands.
• Supercooled clouds behaved differently by elevation:
  • Over lowlands and lower- elevation mountains, supercooled cloud tops warmed and descended (indicating suppressed vertical growth).
  • Over the highest regions, the Tibetan Plateau and the Andes, supercooled clouds cooled further and ascended.

Possible reasons for the high-mountain difference

• Tibetan Plateau: Stronger warming at lower altitudes may destabilize the vertical thermal profile, allowing clouds to develop higher into colder air.
• Andes: Higher atmospheric moisture content appears linked to the colder, higher clouds.⁠

Overall, high- elevation clouds show both a distinct and often stronger response to warming than clouds elsewhere. The authors note that further work is needed on how these icy, high- elevation cloud changes affect local radiative energy budgets (how much heat is trapped or reflected) and regional hydrology, especially critical for the Tibetan Plateau, which supplies water to hundreds of millions of people amid shrinking snowpack and shifting precipitation.

The original research paper is: Zhang et al., “Amplified Cloud-Phase Responses Over Global Mountain Regions Under Climate Warming,” AGU Advances (2026), DOI: 10.1029/2026AV002380.

Supercooled CF decrease means that the cloud fraction (CF) of supercooled clouds is getting smaller over time.

Quick definitions

• CF (Cloud Fraction): The portion of the sky (or of a given area) covered by clouds of a particular type. It is usually expressed as a percentage or a fraction (e.g., 0.25 = 25% cloud cover).
• Supercooled clouds: Clouds that contain liquid water droplets even though the temperature is below 0 °C (32 °F). These are a transitional “mixed phase” state between ordinary warm liquid clouds and ice clouds.

What the decrease means in the study

In the Zhang et al. (2026) paper, researchers observed that, under climate warming:

• The fraction of the sky covered by supercooled clouds is shrinking.
• At the same time, the fractions of warm- liquid clouds and ice clouds are generally increasing.

In plain terms: fewer clouds are staying in the supercooled (liquid- below- freezing) state. Many of them are either warming into ordinary liquid clouds or converting into ice clouds.

This change is stronger over most mountain regions than over lowlands or oceans, with the notable exception of the highest mountains (Tibetan Plateau and Andes), where the remaining supercooled clouds behave differently (they rise and cool rather than warm and descend).

_____________________________________________________________________________________

Amplified Cloud-Phase Responses Over Global Mountain Regions Under Climate Warming

This study examines how cloud phase (warm liquid, supercooled liquid, and ice) and vertical structure respond to climate warming over major mountain ranges worldwide, compared with adjacent lowlands. It uses 20 years (2003- 2022) of MODIS Aqua satellite observations of cloud fraction (CF), cloud- top temperature (CTT), and cloud- top height (CTH), combined with ERA5 reanalysis data.

Mountain regions have received less attention than oceans in cloud- climate research, despite their importance for regional hydrology, radiation, and water resources (especially the Tibetan Plateau).

Main Findings

• Baseline differences: Mountains show systematically higher fractions of supercooled clouds; warm clouds are more common over lowlands (clear elevation dependence).
• Amplified phase responses over mountains:
  • Ice CF and warm -liquid CF increase.
  • Supercooled CF decreases.
  • These changes occur at rates generally exceeding those over global land and ocean.
• Ice clouds: Tops rise and cool at rates 2- 4 times faster over mountains than over lowlands or oceans.
• Supercooled clouds (the distinctive result):
  • Over most lowlands and lower- elevation mountains → cloud tops warm and descend (suppressed vertical development).
  • Over the highest regions (Tibetan Plateau and Andes) → cloud tops show pronounced cooling and ascent, consistent with more active supercooled- to- ice phase transitions.⁠

Specific cooling rates for supercooled cloud tops include approximately −2.7 °C per decade over the Tibetan Plateau and −1.1 °C per decade over the Andes.

Environmental Drivers

• Global lowlands: Decreases in supercooled clouds linked to enhanced thermal stability.
• Tibetan Plateau: Stronger low- level warming than aloft → enhanced thermal destabilization → allows clouds to develop higher into colder air.
• Andes: Moister mid- tropospheric background environment associated with colder, higher supercooled clouds.⁠

These patterns contrast with the broader tendency toward atmospheric stabilization under global warming.

Implications

High- elevation mountain clouds show a distinct and often amplified response to warming. Changes in cloud phase and CTT over these regions may represent an important but still under- explored pathway affecting regional radiative responses (how much energy is trapped or reflected). Further research is needed on the consequences for local hydrology and energy budgets.

The paper’s summary section emphasizes that these mountain- specific adjustments differ from typical lowland and oceanic responses and highlight the value of elevation- resolved cloud observations for understanding climate change impacts in high- mountain environments.

Journal: AGU Advances, (2026, Open Access)

DOI: 10.1029/2026av002380

Provided: Eos

Authors: Ying Zhang, Minghuai Wang, Yannian Zhu, Daniel Rosenfeld, Jihu Liu, Xuexu Wu

First published: 19 September 2026

Abstract

Changes in cloud phase can substantially affect cloud radiative feedback and precipitation processes, yet their responses to climate warming remain poorly understood, particularly over mountain regions. Using 20 years of satellite observations together with ERA5 reanalysis data, this study investigates changes in cloud fraction (CF), cloud top temperature, and cloud top height across major mountain regions worldwide and their adjacent lowlands. We find that cloud-phase responses to warming are generally amplified over mountains: ice and warm CF increase, while supercooled CF decreases, at rates exceeding those over global land and ocean. These changes are accompanied by ice cloud tops rising and cooling at rates two to four times faster than over lowlands. Over most mountain and lowland regions, supercooled cloud tops warm and descend, indicating suppressed vertical development. In contrast, the Tibetan Plateau and Andes exhibit pronounced cooling and ascent of supercooled cloud tops, suggesting active supercooled-to-ice phase transitions. Over the Tibetan Plateau, this behavior is associated with enhanced thermal destabilization linked to stronger low-level warming than warming aloft. Over the Andes, a moister mid-tropospheric background environment is consistent with colder and higher supercooled clouds. This stands in contrast to the broader tendency toward atmospheric stabilization under global warming, alongside generally limited changes in relative humidity. These findings highlight distinct and amplified cloud-phase and cloud top responses over high-elevation regions and suggest that mountain cloud changes may represent an important but still insufficiently explored pathway affecting regional radiative response.