Predicting Top-of-Atmosphere Thermal Radiance Using MERRA-2 Atmospheric Data with Deep Learning

Abstract

Abstract:Image data from space-borne thermal infrared (IR) sensors are used for a variety of applications, however they are often limited by their temporal resolution (i.e., repeat coverage). To potentially increase the temporal availability of thermal image data, a study was performed to determine the extent to which thermal image data can be simulated from available atmospheric and surface data. The work conducted here explored the use of Modern-Era Retrospective analysis for Research and Applications, Version 2 (MERRA-2) developed by The National Aeronautics and Space Administration (NASA) to predict top-of-atmosphere (TOA) thermal IR radiance globally at time scales finer than available satellite data. For this case study, TOA radiance data was derived for band 31 (10.97 µm) of the Moderate-Resolution Imaging Spectroradiometer (MODIS) sensor. Two approaches have been followed, namely an atmospheric radiative transfer forward modeling approach and a supervised learning approach. The first approach uses forward modeling to predict TOA radiance from the available surface and atmospheric data. The second approach applied four different supervised learning algorithms to the atmospheric data. The algorithms included a linear least squares regression model, a non-linear support vector regression (SVR) model, a multi-layer perceptron (MLP), and a convolutional neural network (CNN). This research found that the multi-layer perceptron model produced the lowest overall error rates with an root mean square error (RMSE) of 1.36 W/m 2 ·sr·µm when compared to actual Terra/MODIS band 31 image data. These studies found that for radiances above 6 W/m 2 ·sr·µm, the forward modeling approach could predict TOA radiance to within 12 percent, and the best supervised learning approach can predict TOA to within 11 percent.

3D radiative heating of tropical upper tropospheric cloud systems derived from synergistic A-Train observations and machine learning

“…The radiative fluxes and heating rates of 2B-FLXHR-LIDAR (version R04; Henderson et al, 2013;L'Ecuyer et al, 2008) were derived by applying the BUGSrad broadband radiative transfer model (Ritter and Geleyn, 1992) to the scenes observed by CALIPSO-CloudSat, using as inputs the vertical location of the cloud layers (2B-GEOPROF-LIDAR; Mace et al, 2009), the cloud water/ice content and effective particle sizes retrieved from radar only (2B-CWC-RO; Austin et al, 2009), distinction between cloud and rain water contents from 2C-PRECIP-COLUMN (Haynes et al, 2009) and collocated atmospheric and surface auxiliary data from ECMWF. For the clouds and aerosols which are undetected by CloudSat, the MODIS-based cloud optical depth (2B-TAU) and CALIPSO version-3 products (Winker et al, 2010) are used to calculate the corresponding radiative properties.…”

Predictive Resource Management in Energy-constrained Embedded Systems

“…Support Vector Regression optimization problem proposed by Kleynhans et Al. [7] weather forecast service. Moreover, a further aspect to take into account for filling the dataset for the model construction, is that the weather sensors provide a new set of data samples every second.…”

AeroRIT: A New Scene for Hyperspectral Image Analysis

“…Hughes et al used a siamese CNN architecture to match high resolution optical images with their corresponding synthetic aperture radar images [4]. Kleynhans et al compared the performance of predicting top-of-atmosphere thermal radiance by using forward modeling with radiosonde data and radiative transfer modeling (MODTRAN [8]) against a multi-layer perception (MLP) and CNN and observed better performance from MLP and CNN in all experimental cases [5]. Kemker et al used multi-scale independent component analysis and stacked convolutional autoencoders as self-supervised learning tasks before performing semantic segmentation on multispectral and hyperspectral imagery [6], [7].…”