Logo Logo
Help
Contact
Switch language to German
Information content of halo displays for remote sensing of ice crystal properties
Information content of halo displays for remote sensing of ice crystal properties
Cirrus clouds consist of small non-spherical ice crystals, which scatter solar radiation and absorb and emit thermal infrared radiation. Depending on which of the effects dominates, these clouds have either a cooling or a warming effect on the earth’s atmosphere and surface. Ice crystals can grow to different sizes and shapes depending on the temperature and relative humidity of their environment. Potential shapes range from hexagonal columns and plates to complex ice crystal aggregates and dendrites. Inadequate knowledge about ice crystal shape results in large uncertainties in retrievals of ice crystal optical and microphysical properties. Any information about ice crystal shape is therefore valuable to improve ice cloud remote sensing and to better quantify effects on the radiation budget and thus on climate. This thesis aims at investigating the information content of halo displays regarding ice crystal properties, which comprise ice crystal size, shape, surface roughness, and orientation in cirrus clouds. Halo displays form by refraction and reflection of light by ice crystals with different shapes and orientations and could therefore provide important information about these properties. The frequently observed 22° halo and the rare 46° halo are both formed by randomly oriented hexagonal crystals. Upper and lower tangent arcs are caused by oriented ice crystal columns and sundogs emerge from light refracted by oriented ice crystal plates. The optical displays can be simulated quantitatively using radiative transport models, which are well established for the simulation of randomly oriented crystals. To allow simulation of oriented crystals, the raytracing algorithm CrystalTrace was developed and implemented into the Monte Carlo model MYSTIC. Within the scope of this work the weather-proof camera system HaloCam was designed and installed on the rooftop platform of the Meteorological Institute Munich to allow for continuous observation of halo displays. Together with the newly developed automated halo detection algorithm HaloForest, HaloCam provides a consistent dataset of halo observations. Between January 2014 and June 2016 HaloForest detected 22° halos 2% of the time. Using co-located ceilometer data, it was estimated that about 25% of the cirrus clouds produced a visible 22° halo. The RICO retrieval was developed to estimate ice crystal size, shape and the fraction of smooth and rough ice crystals from calibrated radiance measurements of 22° halos in combination with radiative transfer simulations. Analyzing HaloCam data between September 2015 and November 2016 with RICO revealed effective ice crystal radii of about 20 μm on average, with more than 90% of the radii being smaller than 40 μm. Within the measurement uncertainty, the 22° halo observations can be reproduced with plate-like, hollow, and columnar crystals with a smooth crystal fraction of about 80%, 60%, and 30%, respectively. The optical properties of ice crystal plates were found to best match the HaloCam observations in the region of the 22° halo, followed by solid columns and aggregates of columns. The scattering properties of the ice crystal plates used in this study, produce a pronounced 46° halo in addition to the 22° halo for effective radii up to about 50 μm. Since the evaluated HaloCam observations showed only 22° halos without visible 46° halo, plates must be excluded as representative ice crystal shape. Therefore, solid columns and aggregates of columns are the best matching habits for this dataset. During the ML-CIRRUS campaign a unique dataset was collected combining HaloCam observations with in-situ measurements of the research aircraft HALO. This dataset contains observations of complex halo displays formed by oriented ice crystals. Using CrystalTrace, a method was presented to retrieve the fraction of oriented ice crystals from observations of sundogs and upper tangent arcs. This study demonstrates that halo displays contain valuable information about ice crystal properties such as size, shape, roughness, and orientation and performs for the first time a systematic investigation. Operating HaloCam in combination with HaloForest provides a consistent dataset of long-term halo observations. In synergy with the RICO retrieval and CrystalTrace, these observations contribute to an improved understanding of ice crystal properties.
Halo displays, cirrus clouds, remote sensing, ice crystal properties
Forster, Linda
2017
English
Universitätsbibliothek der Ludwig-Maximilians-Universität München
Forster, Linda (2017): Information content of halo displays for remote sensing of ice crystal properties. Dissertation, LMU München: Faculty of Physics
[img]
Preview
PDF
Forster_Linda.pdf

41MB

Abstract

Cirrus clouds consist of small non-spherical ice crystals, which scatter solar radiation and absorb and emit thermal infrared radiation. Depending on which of the effects dominates, these clouds have either a cooling or a warming effect on the earth’s atmosphere and surface. Ice crystals can grow to different sizes and shapes depending on the temperature and relative humidity of their environment. Potential shapes range from hexagonal columns and plates to complex ice crystal aggregates and dendrites. Inadequate knowledge about ice crystal shape results in large uncertainties in retrievals of ice crystal optical and microphysical properties. Any information about ice crystal shape is therefore valuable to improve ice cloud remote sensing and to better quantify effects on the radiation budget and thus on climate. This thesis aims at investigating the information content of halo displays regarding ice crystal properties, which comprise ice crystal size, shape, surface roughness, and orientation in cirrus clouds. Halo displays form by refraction and reflection of light by ice crystals with different shapes and orientations and could therefore provide important information about these properties. The frequently observed 22° halo and the rare 46° halo are both formed by randomly oriented hexagonal crystals. Upper and lower tangent arcs are caused by oriented ice crystal columns and sundogs emerge from light refracted by oriented ice crystal plates. The optical displays can be simulated quantitatively using radiative transport models, which are well established for the simulation of randomly oriented crystals. To allow simulation of oriented crystals, the raytracing algorithm CrystalTrace was developed and implemented into the Monte Carlo model MYSTIC. Within the scope of this work the weather-proof camera system HaloCam was designed and installed on the rooftop platform of the Meteorological Institute Munich to allow for continuous observation of halo displays. Together with the newly developed automated halo detection algorithm HaloForest, HaloCam provides a consistent dataset of halo observations. Between January 2014 and June 2016 HaloForest detected 22° halos 2% of the time. Using co-located ceilometer data, it was estimated that about 25% of the cirrus clouds produced a visible 22° halo. The RICO retrieval was developed to estimate ice crystal size, shape and the fraction of smooth and rough ice crystals from calibrated radiance measurements of 22° halos in combination with radiative transfer simulations. Analyzing HaloCam data between September 2015 and November 2016 with RICO revealed effective ice crystal radii of about 20 μm on average, with more than 90% of the radii being smaller than 40 μm. Within the measurement uncertainty, the 22° halo observations can be reproduced with plate-like, hollow, and columnar crystals with a smooth crystal fraction of about 80%, 60%, and 30%, respectively. The optical properties of ice crystal plates were found to best match the HaloCam observations in the region of the 22° halo, followed by solid columns and aggregates of columns. The scattering properties of the ice crystal plates used in this study, produce a pronounced 46° halo in addition to the 22° halo for effective radii up to about 50 μm. Since the evaluated HaloCam observations showed only 22° halos without visible 46° halo, plates must be excluded as representative ice crystal shape. Therefore, solid columns and aggregates of columns are the best matching habits for this dataset. During the ML-CIRRUS campaign a unique dataset was collected combining HaloCam observations with in-situ measurements of the research aircraft HALO. This dataset contains observations of complex halo displays formed by oriented ice crystals. Using CrystalTrace, a method was presented to retrieve the fraction of oriented ice crystals from observations of sundogs and upper tangent arcs. This study demonstrates that halo displays contain valuable information about ice crystal properties such as size, shape, roughness, and orientation and performs for the first time a systematic investigation. Operating HaloCam in combination with HaloForest provides a consistent dataset of long-term halo observations. In synergy with the RICO retrieval and CrystalTrace, these observations contribute to an improved understanding of ice crystal properties.