Obtaining Cloud Base Height and Phase From Thermal Infrared Radiometry Using a Deep Learning Algorithm

doi:10.1109/TGRS.2023.3317532

Obtaining Cloud Base Height and Phase From Thermal Infrared Radiometry Using a Deep Learning Algorithm

2023-01-01

8

PRE

AI

DOI：10.1109/TGRS.2023.3317532

Original

Original request for help

Save

Abstract

En 中文

In this study, a thermal infrared (TIR)-based convolutional neural network (TIR-CNN), originally designed for retrieving cloud optical properties from TIR radiometry, is further developed to obtain global cloud base height (CBH) and cloud thermodynamic phase during both daytime and nighttime. It employs TIR radiances, cloud optical properties retrieved by TIR-CNN, altitude, land cover, and lifting condensation level (LCL) as inputs to estimate global CBH and cloud phase for both single-layer and multilayer clouds. This new model is trained, validated, and tested using radar-lidar products from CloudSat/ Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO). It provides global CBH with root-mean-square errors of 1.19 and 1.91 km for single-layer and multilayer clouds, respectively. A cloud layer classifier is trained to provide information on the quality of retrieved CBH, with total accuracies of 82% and 85% for single-layer and multilayer clouds, respectively. In addition, the new model has remarkable accuracy in identifying the cloud phase within each pixel's vertical column, particularly in differentiating mixed-phase clouds with an ice cloud top.

Keywords:

Clouds

Cloud computing

Integrated optics

Ocean temperature

Atmospheric measurements

Phase measurement

Optical variables measurement

Cloud base height

cloud layer flag

cloud thermodynamic phase

convolutional neural network (CNN)

diurnal cycle

Obtaining Cloud Base Height and Phase From Thermal Infrared Radiometry Using a Deep Learning Algorithm

Abstract

Journal

Organization