Studying Anomalous Discrepancies between MODIS and CALIOP Cloud Observations CyberTraining: Big Data + High-Performance Computing + Atmospheric Sciences

Thumbnail Image

Files

CT2020Team2.pdf (681 KB)

Links to Files

http://hpcf-files.umbc.edu/research/papers/CT2020Team2.pdf

Permanent Link

http://hdl.handle.net/11603/21269

Collections

UMBC Information Systems Department
UMBC Faculty Collection
UMBC Physics Department
UMBC Student Collection

Author/Creator

Author/Creator ORCID

Date

2020

Type of Work

technical reports
Text

Department

Program

Citation of Original Publication

Abraham, Christine; Norman, Olivia; Shepherd, Erick; Zheng, Jianyu; Zhang, Zhibo; Studying Anomalous Discrepancies between MODIS and CALIOP Cloud Observations CyberTraining: Big Data + High-Performance Computing + Atmospheric Sciences; http://hpcf-files.umbc.edu/research/papers/CT2020Team2.pdf

Rights

This item is likely protected under Title 17 of the U.S. Copyright Law. Unless on a Creative Commons license, for uses protected by Copyright Law, contact the copyright holder or the author.

Subjects

UMBC High Performance Computing Facility (HPCF)

Abstract

When examining collocated data from the A-Train satellite constellation, there are a notable number of clouds that CALIOP identifies as transparent but MODIS paradoxically reports as having a high cloud optical thickness (COT), therefore implying that the cloud is opaque. We refer to these as ”anomalous transparent clouds”. Our team is investigating two hypotheses in an effort to explain the occurrence of these anomalies. The first hypothesis is that the anomalies could be MODIS COT retrieval errors due to the misclassification of high albedo surfaces, such as snow and sea ice, as clouds. The other hypothesis is that the anomalies could be clouds which are misidentified as having a high COT due to 3D radiative effects. The former hypothesis was tested by collocating the single-layer cloud anomalies that were over water with NSIDC AMSRE sea ice observations using a k-nearest neighbors (k-NN) algorithm. We determined that around 50% of such anomalies occur over areas with high sea ice concentrations (95-100%). Further research is required to account for the other half of the anomalies that showed no correlation with high albedo surfaces. We have taken preliminary steps toward exploring whether the cloud 3D radiative effects hypothesis might explain the remaining anomalies.