Retrieval of Cloud Optical Thickness from Sky-View Camera Images using a Deep Convolutional Neural Network based on Three-Dimensional Radiative Transfer

Observation of the spatial distribution of cloud optical thickness (COT) is useful for the prediction and diagnosis of photovoltaic power generation. However, there is not a one-to-one relationship between transmitted radiance and COT (so-called COT ambiguity), and it is difficult to estimate COT because of three-dimensional (3D) radiative transfer effects. We propose a method to train a convolutional neural network (CNN) based on a 3D radiative transfer model, which enables the quick estimation of the slant-column COT (SCOT) distribution from the image of a ground-mounted radiometrically calibrated digital camera. The CNN retrieves the SCOT spatial distribution using spectral features and spatial contexts. An evaluation of the method using synthetic data shows a high accuracy with a mean absolute percentage error of 18% in the SCOT range of 1–100, greatly reducing the influence of the 3D radiative effect. As an initial analysis result, COT is estimated from a sky image taken by a digital camera, and a high correlation is shown with the effective COT estimated using a pyranometer. The discrepancy between the two is reasonable, considering the difference in the size of the field of view, the space–time averaging method, and the 3D radiative effect.

[1]  María P. Utrillas,et al.  Effective cloud optical depth and enhancement effects for broken liquid water clouds in Valencia (Spain) , 2017 .

[2]  Steven Platnick,et al.  Impact of three‐dimensional radiative effects on satellite retrievals of cloud droplet sizes , 2006 .

[3]  Jan Kleissl,et al.  Coupling sky images with radiative transfer models: a new method to estimatecloud optical depth , 2016 .

[4]  Stephen E. Schwartz,et al.  High‐resolution photography of clouds from the surface: Retrieval of optical depth of thin clouds down to centimeter scales , 2017 .

[5]  Thierry Faure,et al.  Neural network retrieval of cloud parameters of inhomogeneous and fractional clouds , 2001 .

[6]  Anthony B. Davis,et al.  Airborne Three-Dimensional Cloud Tomography , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[7]  J. Kleissl,et al.  Intra-hour forecasting with a total sky imager at the UC San Diego solar energy testbed , 2011 .

[8]  Roger Davies,et al.  Effect of cloud inhomogeneities on the solar zenith angle dependence of nadir reflectance , 1997 .

[9]  Yousuke Sato,et al.  Influence of grid aspect ratio on planetary boundary layer turbulence in large-eddy simulations , 2015 .

[10]  Roger Davies,et al.  Effects of Cloud Heterogeneities on Shortwave Radiation: Comparison of Cloud-Top Variability and Internal Heterogeneity , 1999 .

[11]  Hironobu Iwabuchi,et al.  Efficient Monte Carlo Methods for Radiative Transfer Modeling , 2006 .

[12]  Graham Feingold,et al.  A novel ensemble method for retrieving properties of warm cloud in 3‐D using ground‐based scanning radar and zenith radiances , 2014 .

[13]  Jerome Riedi,et al.  Case study of inhomogeneous cloud parameter retrieval from MODIS data , 2005 .

[14]  Céline Cornet,et al.  Neural network retrieval of cloud parameters of inhomogeneous clouds from multispectral and multiscale radiance data: Feasibility study , 2004 .

[15]  Hideaki Takenaka,et al.  Evaluation of Himawari-8 surface downwelling solar radiation by ground-based measurements , 2018 .

[16]  Seiji Kato,et al.  Estimate of satellite‐derived cloud optical thickness and effective radius errors and their effect on computed domain‐averaged irradiances , 2006 .

[17]  Jian Sun,et al.  Identity Mappings in Deep Residual Networks , 2016, ECCV.

[18]  Yoav Y. Schechner,et al.  In-situ multi-view multi-scattering stochastic tomography , 2016, 2016 IEEE International Conference on Computational Photography (ICCP).

[19]  Luca Bugliaro,et al.  Ground-based imaging remote sensing of ice clouds: uncertainties caused by sensor, method and atmosphere , 2016 .

[20]  Alexander Marshak,et al.  Observations of Three-Dimensional Radiative Effects that Influence MODIS Cloud Optical Thickness Retrievals , 2002 .

[21]  Alexander Marshak,et al.  The Potential for Improved Boundary Layer Cloud Optical Depth Retrievals from the Multiple Directions of MISR , 2008 .

[22]  Robert F. Cahalan,et al.  The albedo of fractal stratocumulus clouds , 1994 .

[23]  Gottfried Hänel,et al.  The Properties of Atmospheric Aerosol Particles as Functions of the Relative Humidity at Thermodynamic Equilibrium with the Surrounding Moist Air , 1976 .

[24]  Rintaro Okamura,et al.  Feasibility study of multi-pixel retrieval of optical thickness and droplet effective radius of inhomogeneous clouds using deep learning , 2017 .

[25]  Thierry Faure,et al.  Neural network retrieval of cloud parameters from high-resolution multispectral radiometric data , 2002 .

[26]  Yuri Knyazikhin,et al.  Cloud optical depth retrievals from the Aerosol Robotic Network (AERONET) cloud mode observations , 2010 .

[27]  María P. Utrillas,et al.  Effective cloud optical depth for overcast conditions determined with a UV radiometers , 2014 .

[28]  Tamio Takamura,et al.  An Intensive Campaign-Based Intercomparison of Cloud Optical Depth from Ground and Satellite Instruments under Overcast Conditions , 2019, SOLA.

[29]  Jan Kleissl,et al.  Cloud tomography applied to sky images: A virtual testbed , 2018, Solar Energy.

[30]  Frank J. Iannarilli,et al.  Application of oxygen A-band equivalent width to disambiguate downwellingradiances for cloud optical depth measurement , 2016 .

[31]  Yousuke Sato,et al.  Impacts of cloud microphysics on trade wind cumulus: which cloud microphysics processes contribute to the diversity in a large eddy simulation? , 2015, Progress in Earth and Planetary Science.

[32]  Hironobu Iwabuchi,et al.  A multi-spectral non-local method for retrieval of boundary layer cloud properties from optical remote sensing data , 2003 .

[33]  Isao Murata,et al.  Estimation of spectral distribution of sky radiance using a commercial digital camera. , 2016, Applied optics.

[34]  Rintaro Okamura,et al.  Multispectral Monte Carlo radiative transfer simulation by the maximum cross-section method , 2017 .

[35]  Yousuke Sato,et al.  Potential of Retrieving Shallow-Cloud Life Cycle from Future Generation Satellite Observations through Cloud Evolution Diagrams: A Suggestion from a Large Eddy Simulation , 2014 .

[36]  Peter Pilewskie,et al.  A spectral method for retrieving cloud optical thickness and effective radius from surface-based transmittance measurements , 2011 .

[37]  Akihide Kamei,et al.  Cloud optical thickness and effective particle radius derived from transmitted solar radiation measurements: Comparison with cloud radar observations , 2006 .

[38]  Otto P. Hasekamp,et al.  A demonstration of adjoint methods for multi-dimensional remote sensing of the atmosphere and surface , 2018 .