北極層狀雲逆相位之形成機制

Abstract

北極層狀雲通常有逆相位的結構（即雲頂為液態、下方為混合態或冰態的結構），並且可以持續很長的時間。先前的研究認為逆相位結構是由冰晶的重力沉降和持續在高空生成液態雲的結果，而浸入核化被認為是北極層狀雲中最主要的冰晶核化過程。然而，冰核的作用以及其和不同環境因素之間的非線性過程對此逆相位現象的的影響尚未得到充分的了解。冰晶核化過程受到溫度、飽和度、冰核的種類和數量濃度影響。在雲的上層中，有任何不利於這些因子的條件都會使冰晶核化速率減小，從而能夠保持逆相位的結構。本研究利用WRF模式搭配詳盡的冰晶核化方案，試圖找出北極層狀雲中逆相位結構的微物理機制。該冰晶核化方案可以預報不同種類冰核在三個相態間的轉換，並考慮了不同的冰晶核化途徑以及隨著降水粒子的沉降。本研究模擬的個案在2008年3月4日到3月5日期間於阿拉斯加州巴羅鎮的ARM計畫中所觀測到。沙塵和煤灰為兩個主要測試的冰核種類，並利用MERRA-2中的氣溶膠再分析資料來提供模式的初始和邊界條件。模式結果顯示此個案的雲主要是因為鋒面上升運動而產生。沙塵和煤灰都成功模擬出逆相位結構，但在煤灰的實驗中，沒有任何水雲到達巴羅，顯示在這個個案中沙塵可能是主要的冰核。在雲中，沙塵主要透過異質凝華核化產生冰晶，而非如前人研究中所認為的浸入核化。但是因為冰晶核化速率很低，且重力沉降移除了沙塵，使生成的冰晶數量相對較少，而導致華格納-白吉隆-芬代生過程十分緩慢，從而可以讓雲水藉由舉升絕熱冷卻持續形成。此外，一旦沙塵進入雲滴中，冰核就只能透過浸入核化產生冰晶，但在這模擬的個案條件中，此核化過程效率較低。在本研究中，北極層狀雲的逆相位結構機制是由於冰核數量有限，且部分陷入較不易核化的水雲中，以及通過重力沉降移除的沙塵所導致的。

Arctic stratiform clouds (ASC) often exhibit a phase inversion structure (i.e., liquid top and mixed- or ice-phase below), and can persist for quite a long time. Previous studies indicated that the phase inversion structure is the result of persistent liquid cloud generation aloft, and gravitational sedimentation of ice precipitation that formed dominantly by immersion freezing; however, the role of ice nuclei (IN) and nonlinear effect of different environmental factors on phase inversion was not fully addressed before. Ice nucleation processes are affected by temperature, saturation, and the species and number concentration of IN. Unfavorable conditions in any of these factors at the upper part of the cloud can minimize the ice nucleation rate there and thus maintain the phase inversion. This study aims to find out the microphysics mechanism for ASC phase inversion by using the Weather Research and Forecasting (WRF) model coupled with a detailed ice nucleation scheme. This ice nucleation scheme considers prognostic IN species, their conversion between different phases, and different routes of ice nucleation and the sedimentation with precipitation particles. The 2008 Mar 04-05 case observed with an Atmospheric Radiation Measurement (ARM) facility at Barrow, Alaska, is simulated. Dust and soot are considered as the two main IN, using the Modern-Era Retrospective analysis for Research and Applications, Version 2 (MERRA-2) aerosol reanalysis data for providing initial and boundary conditions. The simulation result reveals that clouds formed because of the upward motion associated with a frontal system. Simulations with either dust or soot particles can both reproduce the phase inversion structure, but no liquid cloud can arrive Barrow in the soot run, inferring that dust may be the dominant IN in this case. In the cloud, ice particles nucleated from dust mainly through deposition nucleation rather than immersion as suggested in earlier studies. However, due to low ice nucleation rate and gravitational removal, the amount of ice particles that produced is relatively low, which yielded a slow Wegener–Bergeron–Findeisen conversion to allow the sustentation of cloud water during adiabatic cooling; this results in a persistent liquid cloud aloft. Furthermore, once dust particles get into the cloud drops, ice nucleation can only proceed via immersion freezing, which is rather inefficient comparing to deposition nucleation under the simulated conditions. The limited amount of dust together with the trap of dust in an inefficient nucleation phase and the nucleation scavenging of dust by falling ice precipitation serve as the main mechanism of ASC phase inversion in this study.