Formation of a tropopause cirrus layer observed over Florida during CRYSTAL-FACE
Formation of a tropopause cirrus layer observed over Florida during CRYSTAL-FACE
复制标题
CRYSTAL-FACE 期间在佛罗里达州上空观测到对流层顶卷云层的形成
DOI：
--
复制
发表时间：
2005
期刊：
影响因子：
--
通讯作者：
M. McGill
中科院分区：
文献类型：
--
作者：
E. Jensen;L. Pfister;T. Bui;A. Weinheimer;E. Weinstock;Jessica B. Smith;J. Pittman;D. Baumgardner;P. Lawson;M. McGill
[1]2002年7月13日，佛罗里达地区上空出现了一个大范围的、肉眼不可见的对流层顶卷云层。用美国航天局热带铁砧和卷云层卷云区域研究-佛罗里达地区卷云实验（CRYSTAL-FACE）仪器对这种云进行了非常详细的观测，包括用WB-57飞机进行的现场测量。本文以7月13日云为例，分析了对流层顶卷云形成和演化的物理过程。微物理测量表明，云层中冰晶直径约为7 ~ 50 μm，峰值数模约为10-25 μm。在云中的原位水蒸气和温度测量表明，相对于整个冰过饱和，冰饱和比高达1.8。即使当冰的表面积密度高达约500 μm2 cm−3时，也观察到20-30%的冰过饱和。轨迹分析表明，这一天在对流层顶附近采样的空气一般来自北方，并且在前几天有相当大的冷却。对WB-57飞行轨迹上沿着空气包轨迹的红外卫星图像的检查表明，对流层顶云层的形成，一般来说，不仅仅是最近生成的铁砧卷云留下的冰。使用沿着轨迹路径的时间-高度温度幕模拟云的形成表明，当空气平流进入南佛罗里达地区并冷却到异常低的温度时，云可能在对流层顶附近原位形成。如果我们假设一个高的阈值，通过均匀冻结的含水硫酸盐气溶胶的冰成核，该模型再现了观测到的云结构，冰晶尺寸分布，和冰过饱和度的统计。包括观测到的重力波温度扰动的模拟是必不可少的再现观测到的云的属性。没有波，晶体数密度太低，晶体尺寸太大，晶体脱落太快，在模拟结束时只留下很少的云。在云模拟中，高过饱和度和高表面积的重合可以通过最近的成核或晶体沉积到过饱和层中来产生。如果我们假设在T < 200 K时卷云内的稳态相对湿度提高了约30%，则模式结果与观测到的过饱和度之间的一致性有所改善。WB-57的观测结果和模式结果表明，对流层顶附近的云层不可逆地脱水。
[1] On 13 July 2002 a widespread, subvisible tropopause cirrus layer occurred over the Florida region. This cloud was observed in great detail with the NASA Cirrus Regional Study of Tropical Anvils and Cirrus Layers–Florida Area Cirrus Experiment (CRYSTAL-FACE) instrumentation, including in situ measurements with the WB-57 aircraft. In this paper, we use the 13 July cloud as a case study to evaluate the physical processes controlling the formation and evolution of tropopause cirrus layers. Microphysics measurements indicate that ice crystal diameters in the cloud layer ranged from about 7 to 50 μm, and the peak number mode was about 10–25 μm. In situ water vapor and temperature measurements in the cloud indicated supersaturation with respect to ice throughout, with ice saturation ratios as large as 1.8. Even when the ice surface area density was as high as about 500 μm2 cm−3, ice supersaturations of 20–30% were observed. Trajectory analysis shows that the air sampled near the tropopause on this day generally came from the north and cooled considerably during the previous few days. Examination of infrared satellite imagery along air parcel back trajectories from the WB-57 flight track indicates that the tropopause cloud layer formation was, in general, not simply left over ice from recently generated anvil cirrus. Simulations of cloud formation using time-height curtains of temperature along the trajectory paths show that the cloud could have formed in situ near the tropopause as the air was advected into the south Florida region and cooled to unusually low temperatures. If we assume a high threshold for ice nucleation via homogeneous freezing of aqueous sulfate aerosols, the model reproduces the observed cloud structure, ice crystal size distributions, and ice supersaturation statistics. Inclusion of observed gravity wave temperature perturbations in the simulations is essential to reproduce the observed cloud properties. Without waves, crystal number densities are too low, crystal sizes are too large, and the crystals fall out too fast, leaving very little cloud persisting at the end of the simulations. In the cloud simulations, coincidence of high supersaturations and high surface areas can be produced by either recent nucleation or sedimentation of crystals into supersaturated layers. The agreement between model results and observed supersaturations is improved somewhat if we assume that the steady state relative humidity within cirrus at T < 200 K is enhanced by about 30%. The WB-57 measurements and the model results suggest that the cloud layer irreversibly dehydrated air near the tropopause.