Using an optimal estimation algorithm to describe the mass-dimensional properties of ice clouds

Update Item Information
Publication Type dissertation
School or College College of Mines & Earth Sciences
Department Atmospheric Sciences
Author Mascio, Jeana Rose
Title Using an optimal estimation algorithm to describe the mass-dimensional properties of ice clouds
Date 2018
Description Interpretations of remote sensing measurements collected in sample volumes containing ice-phase hydrometeors are very sensitive to assumptions regarding the distributions of mass with ice crystal dimension, otherwise known as mass-dimensional (m-D) relationships. The uncertainties from these assumptions extend to backscattered cross-sections and radar forward modeled reflectivity factors. These uncertainties and m-D variability were derived using an optimal estimation (OE) algorithm applied to reflectivity factors measured by CloudSat and combined with particle size distributions (PSDs) collected by coincident in-situ aircraft during SPartICus. This OE algorithm minimized the difference between observed radar reflectivity and PSD calculated reflectivity, to output optimal m-D relationships per PSD. I found that ice crystal populations tend to be distributed over a continuum-defying simple categorization. Also, the quantified uncertainties in backscatter cross-section and reflectivity factors can be appropriately applied to remote sensing algorithms. Further investigation of the ice particle m-D relationship was studied with in-situ measurements collected during TC4. Two OE algorithms were used -- one algorithm minimized radar reflectivity (MZ), the other minimized observed ice water content (IWC) and PSD calculated IWC (XIWC). The XIWC results show that both parameters in the m-D relationship increase with temperature. With the prefactor varying by a factor of 5 and the exponent varying by some 16% over a typical range of ice cloud temperatures, forward modeling errors in radar reflectivity could be in excess of 5 dB, further suggesting that retrievals of precipitation rates from radar measurements in ice clouds be in error by factors easily exceeding 3. The MZ algorithm, adjusted for slant radar incidence, was applied to in-situ and radar data collected in mountainous terrain during StormVEx. The outputs of the MZ algorithm here were analyzed along with the enhancement of backscatter (EB) cross-section in the zenith and slant 45º depolarization ratio (DR). Statistics of the results show that forward model errors can create reflectivity differences around 7 dB compared to using fixed m-D relationships, resulting in snowfall rate differences of 1.7 mm per hour. An inverse (direct) relationship between the m-D prefactor and slant 45º DR (zenith EB) can help improve radar-based retrievals by reducing forward model errors.
Type Text
Publisher University of Utah
Subject Atmospheric sciences; Remote sensing
Dissertation Name Doctor of Philosophy
Language eng
Rights Management (c) Jeana Rose Mascio
Format Medium application/pdf
ARK ark:/87278/s6m94qr8
Setname ir_etd
ID 1424035
Reference URL https://collections.lib.utah.edu/ark:/87278/s6m94qr8
Back to Search Results