Frequency and causes of failed MODIS cloud property retrievals for liquid phase clouds over global oceans | JoVE Visualize

Frequency and causes of failed MODIS cloud property retrievals for liquid phase clouds over global oceans

Hyoun-Myoung Cho¹, Zhibo Zhang², Kerry Meyer³

¹Joint Center of Earth Systems Technology Baltimore Maryland USA.

Journal of Geophysical Research. Atmospheres : JGR|September 23, 2016

View abstract on PubMed

Summary

Failed Moderate Resolution Imaging Spectroradiometer (MODIS) retrievals for marine liquid phase clouds occur frequently, often due to large droplet sizes or challenging viewing conditions. Precipitating clouds also contribute to retrieval failures, with the exact cause of extremely large droplet sizes remaining unclear.

Area of Science:

Atmospheric Science
Remote Sensing
Cloud Physics

Background:

Moderate Resolution Imaging Spectroradiometer (MODIS) uses look-up tables (LUTs) to retrieve cloud properties like effective radius (r_c) and optical thickness (τ).
Retrieval failures occur when observed cloud reflectances fall outside the precomputed LUT, indicating no valid combination of τ and r_c explains the observations.

Purpose of the Study:

To analyze the frequency and causes of failed MODIS retrievals for marine liquid phase (MLP) clouds.
To investigate the influence of cloud properties, viewing geometries, and ancillary data issues on retrieval failures.

Main Methods:

Analysis of one year of Aqua MODIS Collection 6 data.
Collocation with Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIOP) and CloudSat observations.
Statistical analysis of retrieval failure rates across different cloud regimes and conditions.

Main Results:

Overall MODIS retrieval failure rates are 16% (0.86/2.1 µm channels) and 10% (0.86/3.7 µm channels).
Failure rates are higher in broken trade wind cumulus regimes compared to stratocumulus.
The primary failure mode is 'r_c too large' (60-85%), followed by 'r_c too small' or τ retrieval issues.
Enhanced failures correlate with partial cloudiness, subpixel inhomogeneity, sunglint, large viewing/solar zenith angles, cloudbow/glory angles, and issues with cloud masking, overlapping, or phase retrieval.

Conclusions:

Over 84% of failed retrievals along the CALIPSO track are linked to identified potential causes.
Remaining failures, particularly those observed by CloudSat, are often associated with precipitating clouds.
The cause of extremely large r_c in some MLP clouds (microphysics vs. artifacts) requires further investigation.