Ice crystal number concentration estimates from lidar–radar satellite remote sensing – Part 1: Method and evaluation

Odran Sourdeval, Edward Gryspeerdt, Martina Krämer, Tom Goren, Julien Delanoe͏̈, Armin Afchine, Friederike Hemmer, Johannes QuaasAtmospheric 2018 · DOI

TL;DR

This paper introduces a new method, DARDAR-Nice, to estimate ice crystal number concentration (Ni) from combined satellite lidar and radar data. This is a crucial step towards better understanding ice clouds' role in climate, as global observational constraints for Ni have been lacking. The study validates this method against in-situ measurements and presents initial global climatologies, showing promising results despite some limitations.

What Problem Does This Paper Address?

Ice clouds play a significant role in Earth's climate, influencing radiation balance and weather patterns. However, accurately quantifying their properties, particularly the ice crystal number concentration (Ni), has been challenging. While advancements have been made for liquid clouds, global, satellite-based estimates of Ni are scarce. This lack of data hinders our ability to understand aerosol-cloud interactions and to validate climate and weather prediction models. Existing satellite methods primarily retrieve other cloud properties like ice water content (IWC) or effective radius, but not Ni directly. This paper aims to fill this gap by developing and evaluating a method to derive Ni from combined lidar and radar measurements.

Methods

The core of this study is the development of the DARDAR-Nice dataset, which leverages the existing DARDAR (DARadar-Lidar) product. DARDAR combines CloudSat radar and CALIOP lidar data to retrieve ice cloud properties by constraining moments of a parameterized particle size distribution (PSD). This study adapts DARDAR's framework to specifically estimate Ni. The method relies on the D05 PSD parameterization, which assumes a "universal" PSD shape and uses scaling parameters to fit observed data. By relating lidar (extinction) and radar (reflectivity) measurements to specific moments of the PSD, the authors can infer the necessary parameters to calculate Ni, which is the zeroth moment of the PSD. The validation process involves comparing these satellite-derived Ni estimates with in-situ measurements from various airborne campaigns and analyzing case studies and global climatologies.

Datasets

  • DARDAR Product (v2.1.1): The primary dataset used for generating satellite Ni estimates. This product combines CloudSat radar and CALIOP lidar measurements.
  • In-situ Measurements: Data from five airborne campaigns (COALESC 2011, ML-CIRRUS 2014, ACRIDICON-CHUVA 2014, SPARTICUS 2010, and ATTREX 2014) were used for validation. These campaigns provided particle size distribution (PSD) measurements.
  • HYSPLIT Trajectories and NARR Reanalyses: Used for case study analysis to understand air parcel movement and meteorological conditions.
  • MODIS/Aqua 11 µm Brightness Temperature: Used in the case study for context on cloud structures.

Key Insights and Ideas

  • Lidar-Radar Synergy is Key: Combining lidar and radar data is crucial for constraining Ni. Lidar is sensitive to smaller particles (important for Ni), while radar is sensitive to larger particles. Together, they provide a more comprehensive view of the PSD.
  • PSD Shape Assumptions Matter: The D05 PSD parameterization, while useful, assumes a monomodal shape. This can lead to overestimations of Ni, especially in colder temperatures or when PSDs are bimodal (common in certain cloud regimes), as the model struggles to accurately extrapolate concentrations of small particles.
  • Updrafts Drive High Ni: The study highlights a strong correlation between high Ni values and strong updrafts, particularly in convective and orographic regions. This suggests that atmospheric dynamics play a significant role in determining ice crystal concentrations.
  • Temperature Dependence is Pronounced: Ni generally increases with decreasing temperature, especially below -40°C, consistent with the increasing importance of homogeneous freezing at colder temperatures.

Key Results

  • DARDAR-Nice Validation: The DARDAR-Nice Ni estimates show good agreement with in-situ measurements, particularly for larger ice crystals (D > 100 µm).
  • Overestimation of Small Particles: For smaller ice crystals (D < 25 µm), DARDAR-Nice (and the underlying D05 parameterization) tends to overestimate Ni, especially at temperatures above -50°C, due to the monomodal PSD assumption. This overestimation can be by a factor of 2-3.
  • Case Study Insights: A case study of a frontal ice cloud revealed distinct vertical Ni patterns, with higher concentrations associated with strong updrafts and deeper cloud layers for larger crystals.
  • Climatological Patterns: Global analysis shows a strong temperature dependence of Ni and significant regional variations, with higher Ni in convective and orographic regions. Seasonal variations are also observed, with stronger Ni in tropical summer and midlatitude winter storm tracks.

Limitations and Future Work

  • PSD Parameterization: The primary limitation acknowledged is the reliance on the D05 PSD parameterization, which assumes a monomodal shape. This can lead to biases, particularly in the concentration of small ice crystals. Future work should focus on PSD parameterizations that better represent bimodality and are more sensitive to retrieving Ni.
  • Uncertainties in Small Particle Measurements: In-situ measurements of small ice crystals (5-25 µm) can have significant uncertainties due to instrument limitations and potential ice shattering. This affects the validation of Ni estimates for these sizes.
  • Lidar-Only and Radar-Only Conditions: Retrievals in lidar-only or radar-only conditions are less constrained, leading to potentially higher uncertainties in Ni estimates.
  • Future Work: The authors emphasize the need for further evaluation of DARDAR-Nice, including more in-situ comparisons and comparisons with modeling studies. Part 2 of this series will delve deeper into the controls on Ni using this new dataset. Improving the PSD parameterization is a key area for future development.

Summary updated

Abstract

The number concentration of cloud particles is a key quantity for understanding aerosol–cloud interactions and describing clouds in climate and numerical weather prediction models. In contrast with recent advances for liquid clouds, few observational constraints exist regarding the ice crystal number concentration (Ni). This study investigates how combined lidar–radar measurements can be used to provide satellite estimates of Ni, using a methodology that constrains moments of a parameterized particle size distribution (PSD). The operational liDAR–raDAR (DARDAR) product serves as an existing base for this method, which focuses on ice clouds with temperatures Tc&lt;-30 ∘C. Theoretical considerations demonstrate the capability for accurate retrievals of Ni, apart from a possible bias in the concentration in small crystals when Tc≳−50 ∘C, due to the assumption of a monomodal PSD shape in the current method. This is verified via a comparison of satellite estimates to coincident in situ measurements, which additionally demonstrates the sufficient sensitivity of lidar–radar observations to Ni. Following these results, satellite estimates of Ni are evaluated in the context of a case study and a preliminary climatological analysis based on 10 years of global data. Despite a lack of other large-scale references, this evaluation shows a reasonable physical consistency in Ni spatial distribution patterns. Notably, increases in Ni are found towards cold temperatures and, more significantly, in the presence of strong updrafts, such as those related to convective or orographic uplifts. Further evaluation and improvement of this method are necessary, although these results already constitute a first encouraging step towards large-scale observational constraints for Ni. Part 2 of this series uses this new dataset to examine the controls on Ni.
BibTeX
@inproceedings{sourdeval2018ice,
  author = {Odran Sourdeval and Edward Gryspeerdt and Martina Krämer and Tom Goren and Julien Delanoe͏̈ and Armin Afchine and Friederike Hemmer and Johannes Quaas},
  title = {{Ice crystal number concentration estimates from lidar–radar satellite remote sensing – Part 1: Method and evaluation}},
  booktitle = {Atmospheric},
  year = {2018},
  doi = {10.5194/acp-18-14327-2018},
}
Figure