Search for: All records

Abstract Arctic single‐layer mixed‐phase clouds were studied using a one‐dimensional model that incorporated the adaptive habit growth model for ice microphysics. The base case was from the Indirect and Semidirect Aerosol Campaign, and it was perturbed over a range of cloud‐average temperatures, maximum (per model run) ice nuclei (IN) concentrations, and large‐scale subsidence velocities. For each parameter combination, the model was iterated out to 48 hr, and the time, called the glaciation time, to complete disappearance of liquid recorded if this occurred within the 48 hr. Dependence of glaciation times on cloud‐average temperatures from −30°C to −5°C, maximum IN concentrations from 0.10 to 30 L⁻¹, and strong–no subsidence, with both isometric and habit‐dependent ice crystal growth, were investigated. For isometric crystal growth, the relationship between the critical maximum IN concentration (IN_crit), the maximum (per model run) IN concentration above which a mixed‐phase cloud glaciated within a fixed model runtime, and cloud‐average temperature was monotonic. IN_critdecreased with decreasing cloud‐average temperature. Strengthening of subsidence led to a further decrease in IN_critfor every cloud‐average temperature. For habit‐dependent ice crystal growth, the relationship between IN_critand cloud‐average temperature was nonmonotonic. Ice crystals develop dendritic and columnar habits near −15°C and −7°C, respectively, and at these two temperatures, ice crystals grew and depleted supercooled liquid water faster than the case when ice crystals grew isometrically. This led to deep local minima in IN_critaround these two temperatures in the model runs. Habit‐dependent ice crystal growth, coupled with changes in cloud‐average temperature, IN_crit, and subsidence strength, led to significant changes in Arctic single‐layer mixed‐phase cloud lifetimes. 

Abstract Convective initiation (CI) nowcasting remains a challenging problem for both numerical weather prediction models and existing nowcasting algorithms. In this study, an object-based probabilistic deep learning model is developed to predict CI based on multichannel infraredGOES-16satellite observations. The data come from patches surrounding potential CI events identified in Multi-Radar Multi-Sensor Doppler weather radar products over the Great Plains region from June and July 2020 and June 2021. An objective radar-based approach is used to identify these events. The deep learning model significantly outperforms the classical logistic model at lead times up to 1 h, especially on the false alarm ratio. Through case studies, the deep learning model exhibits dependence on the characteristics of clouds and moisture at multiple altitudes. Model explanation further reveals that the contribution of features to model predictions is significantly dependent on the baseline, a reference point against which the prediction is compared. Under a moist baseline, moisture gradients in the lower and middle troposphere contribute most to correct CI forecasts. In contrast, under a clear-sky baseline, correct CI forecasts are dominated by cloud-top features, including cloud-top glaciation, height, and cloud coverage. Our study demonstrates the advantage of using different baselines in further understanding model behavior and gaining scientific insights.