Our understanding of cloud processes remains a major source of uncertainty for precipitation forecasts and the radiative feedback of clouds. To address this, the Cloudlab project treats a natural cloud as a controllable laboratory. We use a UAV to introduce seed particles into supercooled liquid clouds which triggers ice formation. We then observe this process with holographic imagers and cloud radars. Across 78 experiments, we found that ice crystal growth is significantly slower in these natural clouds than in laboratory settings. This slowdown is caused by spatial inhomogeneities or pockets of only ice or only liquid that disrupt the efficient transfer of water vapor. Current high-resolution models simulate this process even less efficiently. They retain too much liquid water and likely miscalculate the onset of precipitation. These findings demonstrate how this unique field data set can be used to improve model parameterizations.
Identyfikator: 926 8157 8294 Kod: 724553