Title Page
Abstract
Contents
Chapter 1. Introduction 17
1.1. Background and motivation 17
1.2. Scientific Questions 21
(1) What are the characteristics of Arctic mixed-phase clouds, and how do cloud microphysical processes affect the persistence of mixed-phase clouds? 21
(2) What is the performance of current model simulations on the occurrence of Arctic mixed-phase clouds, and to which cloud microphysical processes is the PWRF model simulation most sensitive in estimating the amount of mixed-phase clouds? 22
1.3. Objectives of this study 24
Chapter 2. Data and model description 25
2.1. ACLOUD campaign 25
2.2. Cloud radar data 27
2.3. Surface radiation data 29
2.4. PWRF model configuration 30
Chapter 3. Arctic cloud properties at Ny-Ålesund, Svalbard 35
3.1. Definition of diagnostics 35
3.2. Classification of hydrometeors and clouds properties 40
3.3. Statistics of Arctic mixed-phase clouds and their radiative effect 53
Chapter 4. Arctic mixed-phase clouds: comparison between observation and model 61
4.1. Meteorological contexts during the ACLOUD campaign 65
4.2. Cloud microphysical properties: Observation vs. model 70
4.3. Theoretical analysis of scheme algorithm 93
4.4. Radiative forcing at the surface 103
Chapter 5. Summary and future direction 107
References 112
국문 초록 137
Table 1. Previous field campaigns to study the Arctic clouds. 64
Table 2. Bias and RMSE in the liquid water content and ice water content for each liquid, ice, and mixed-phase layers between PWRF cloud microphysics schemes and ACLOUD observations. 85
Table 3. Characteristics of ice formation and ice-nucleating particle parameterization for the microphysics schemes. 99
Figure 1. A conceptual model illustrating the primary processes and basic structure of Arctic mixed-phase clouds. 20
Figure 2. Probability density of ice water fraction observed during the study period at Ny-Ålesund. 37
Figure 3. Probability density of (a) liquid, (b) ice, and (c) mixed-phase cloud using Cloudnet algorithm (contour) and ice water fraction... 39
Figure 4. Cloudnet (a) data availability and (b) frequency of cloud occurrence at Ny-Ålesund station from February 2017 to February 2023. The absence of data for the month is shown in (a) black and (b) white contour. 42
Figure 5. Example of cloud classification (Multilayer, 24 Oct 2017; liquid, 20 July 2017; ice, 20 Feb 2017; mixed-phase layer at top, 24 Aug 2017;... 43
Figure 6. Monthly frequency of occurrence of different types of multilayer clouds, single-layer clouds (liquid, ice, and mixed-phase), and... 45
Figure 7. Profiles of cloud fraction for the data period February 2017 to February 2023 and each season: Spring (yellow), Summer (red), Autumn... 46
Figure 8. Statistical overview of cloud occurrence with respect to (a) cloud top temperature, (b) geometric thickness, (c) cloud top height, and... 48
Figure 9. Monthly cross-sections of (a) liquid water content and (b) ice water content from Cloudnet observations. 50
Figure 10. Histogram of the frequency of occurrence of (a) liquid water content and (b) ice water content for all (black), liquid (blue), and mixed-... 51
Figure 11. Relative frequency of cloud top temperature for each all (shaded), liquid (blue), ice (grey), and mixed-phase (yellow) cloud... 52
Figure 12. Histogram of the occurrence frequency of (a) longwave upward, (b) longwave downward, (c) shortwave upward, and (d)... 55
Figure 13. Net shortwave radiation at (a) the surface, (b) the top of the atmosphere, net longwave radiation at (c) the surface, and (d) the top of... 57
Figure 14. Comparison of cloud top temperature for net longwave radiation at the (a) top of the atmosphere and the (b) surface using... 60
Figure 15. (a) Sea level pressure (in hPa; line) with 2-hPa contour interval and 2-m temperature (in °C; shading) using ERA5 reanalysis, (b) MODIS cloud fraction (shading) using MODIS-Aqua cloud mask, and (c) daily air mass backward trajectories... 67
Figure 16. Vertical distributions of (a) liquid water content, (b) ice water content, and (c) temperature for liquid (blue), ice (grey), and mixed-... 69
Figure 17. Spatial distribution of observations (a,f, and k) and the difference between model simulations of (b-e) temperature,(g-j) liquid water content, and (i-o) ice water content. The dashed line indicates the 15% ice cover averaged over the study period. 72
Figure 18. Cloud classification (contour), liquid water content (black), and ice water content (red) from 12:30 to 17:30 on 14 June 2017. 73
Figure 19. Time series of cloud classification, rainfall rate, backscatter coefficient, liquid water path, and radar reflectivity factor for the period on 14 June 2017. 75
Figure 20. Cloud fraction on 14 June 2017 for each cloud microphysics scheme. 77
Figure 21. Ice water content on 14 June 2017 for observation and four microphysics schemes. 79
Figure 22. Vertical distribution of (a) LWC and (b) IWC for Cloudnet observation and model simulations (Morrison scheme, red; WDM6... 81
Figure 23. Rainfall rate for each scheme simulation during the period from 12:30 to 17:30 on 14 June 2017. 82
Figure 24. PWRF-model-simulated and the ACLOUD-observed profiles of (a) liquid water content, (b) ice water content, and (c) temperature. 83
Figure 25. Spatial distributions of the year averaged (March 2008 to February 2009) cloud fraction using the MORR scheme and the associated difference from four selected schemes, given as the Morrison (minus) WDM6, MILB, THOM, and NSSL. 87
Figure 26. Vertical distribution of cloud fraction for Morrison and difference of four selected schemes, from March 2008 to February 2009 in the Arctic region (70-83°N). 89
Figure 27. Area mean cloud-water and cloud-ice mixing ratio from model simulations during October 2008 over the Arctic region (70-83°N). 90
Figure 28. CALIPSO and CloudSat-based cloud fraction on the merged cloud mask product on a 2°x2° grid from March 2008 to February 2009. 91
Figure 29. Performance scores of (a) frequency bias and (b) log-odds ratio on the PWRF cloud occurrence prediction derived from cloud... 95
Figure 30. Vertical profiles of cloud liquid, rain, ice, snow, and graupel mixing ratio (g kg-1) of hydrometeors simulated by four schemes: (a)... 97
Figure 31. Vertical profile of observed and model-simulated temperature during the ACLOUD campaign. Polar 5 flight (black),... 102
Figure 32. Box plots of (a) upwelling longwave flux, (b) downwelling flux, (c) upwelling shortwave flux, and (d) downwelling flux at the surface... 106