Feasibility study of multi-pixel retrieval of optical thickness and droplet effective radius of inhomogeneous clouds using deep learning

Okamura, Rintaro; Iwabuchi, Hironobu; Schmidt, S.

doi:10.5194/amt-10-4747-2017

Cited by 28 publications

(29 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Indeed, earth-observing satellites cannot fully account for cloud inhomogeneity and this points to the necessity of a more extensive implementation of 3-D radiative transfer in the atmosphere with inhomogeneous clouds (Okamura et al, 2017;Iwabuchi, 2006). Even whether such issues are masked by the broad time resolution used in validation exercises, as we will see, REs can introduce bias into the evaluation of satellite estimates.…”

mentioning

confidence: 96%

Evaluation of Himawari-8 surface downwelling solar radiation by ground-based measurements

et al. 2018

View full text Add to dashboard Cite

Abstract. Observations from the new Japanese geostationary satellite Himawari-8 permit quasi-real-time estimation of global shortwave radiation at an unprecedented temporal resolution. However, accurate comparisons with groundtruthing observations are essential to assess their uncertainty. In this study, we evaluated the Himawari-8 global radiation product AMATERASS using observations recorded at four SKYNET stations in Japan and, for certain analyses, from the surface network of the Japanese Meteorological Agency in 2016. We found that the spatiotemporal variability of the satellite estimates was smaller than that of the ground observations; variability decreased with increases in the time step and spatial domain. Cloud variability was the main source of uncertainty in the satellite radiation estimates, followed by direct effects caused by aerosols and bright albedo. Under all-sky conditions, good agreement was found between satellite and ground-based data, with a mean bias in the range of 20-30 W m −2 (i.e., AMATERASS overestimated ground observations) and a root mean square error (RMSE) of approximately 70-80 W m −2 . However, results depended on the time step used in the validation exercise, on the spatial domain, and on the different climatological regions. In particular, the validation performed at 2.5 min showed largest deviations and RMSE values ranging from about 110 W m −2 for the mainland to a maximum of 150 W m −2 in the subtropical region. We also detected a limited overestimation in the number of clear-sky episodes, particularly at the pixel level. Overall, satellite-based estimates were higher under overcast conditions, whereas frequent episodes of cloud-induced enhanced surface radiation (i.e., measured radiation was greater than expected clear-sky radiation) tended to reduce this difference. Finally, the total mean bias was approximately 10-15 W m −2 under clear-sky conditions, mainly because of overall instantaneous direct aerosol forcing efficiency in the range of 120-150 W m −2 per unit of aerosol optical depth (AOD). A seasonal anticorrelation between AOD and global radiation differences was evident at all stations and was also observed within the diurnal cycle.

show abstract

mentioning

confidence: 96%

Evaluation of Himawari-8 surface downwelling solar radiation by ground-based measurements

et al. 2018

View full text Add to dashboard Cite

show abstract

“…Already Varnai and Marshak (2003) discussed and developed a method to determine how the surrounding of a cloud pixel influences the pixel brightness. In a recent study, Okamura et al (2017) also used surrounding pixels to train a neural network to retrieve cloud optical properties more reliably. Our study will try to find a link between the pixel surrounding and the radiance ambiguity discussed in the preceding section.…”

Section: Additional Information From Surrounding Pixelsmentioning

confidence: 99%

Remote sensing of cloud droplet radius profiles using solar reflectance from cloud sides – Part 1: Retrieval development and characterization

Ewald

Zinner²,

Kölling³

et al. 2019

Atmos. Meas. Tech.

View full text Add to dashboard Cite

Abstract. Convective clouds play an essential role for Earth's climate as well as for regional weather events since they have a large influence on the radiation budget and the water cycle. In particular, cloud albedo and the formation of precipitation are influenced by aerosol particles within clouds. In order to improve the understanding of processes from aerosol activation, from cloud droplet growth to changes in cloud radiative properties, remote sensing techniques become more and more important. While passive retrievals for spaceborne observations have become sophisticated and commonplace for inferring cloud optical thickness and droplet size from cloud tops, profiles of droplet size have remained largely uncharted territory for passive remote sensing. In principle they could be derived from observations of cloud sides, but faced with the small-scale heterogeneity of cloud sides, “classical” passive remote sensing techniques are rendered inappropriate. In this work the feasibility is demonstrated to gain new insights into the vertical evolution of cloud droplet effective radius by using reflected solar radiation from cloud sides. Central aspect of this work on its path to a working cloud side retrieval is the analysis of the impact unknown cloud surface geometry has on effective radius retrievals. This study examines the sensitivity of reflected solar radiation to cloud droplet size, using extensive 3-D radiative transfer calculations on the basis of realistic droplet size resolving cloud simulations. Furthermore, it explores a further technique to resolve ambiguities caused by illumination and cloud geometry by considering the surroundings of each pixel. Based on these findings, a statistical approach is used to provide an effective radius retrieval. This statistical effective radius retrieval is focused on the liquid part of convective water clouds, e.g., cumulus mediocris, cumulus congestus, and trade-wind cumulus, which exhibit well-developed cloud sides. Finally, the developed retrieval is tested using known and unknown cloud side scenes to analyze its performance.

show abstract

“…To reduce nonlinearity effects of cloud parameters on the retrievals, COT and column CDER are tabulated by converting to the following parameters s and q according to Okamura et al () as

s = \frac{((), 1 - \overset{g}{true}) τ}{1 + ((), 1 - \overset{g}{true}) τ},

q = \sqrt{R_{normale}},

where

\overset{g}{true}

is a representative asymmetry factor at the wavelength of 550 nm, which is assumed to be 0.86. The simulated measurements and their differentials are calculated from the LUT with a combination of linear interpolation and Akima interpolation (Akima, ).…”

Section: Algorithmmentioning

confidence: 99%

“…To reduce nonlinearity effects of cloud parameters on the retrievals, COT and column CDER are tabulated by converting to the following parameters s and q according to Okamura et al (2017) as…”

Section: Multispectral Solar Reflectance and Radiancementioning

confidence: 99%

An Efficient Method for Microphysical Property Retrievals in Vertically Inhomogeneous Marine Water Clouds Using MODIS‐CloudSat Measurements

Saito

Yang

et al. 2019

JGR Atmospheres

View full text Add to dashboard Cite

An efficient method is developed to infer cloud optical thickness (COT) and cloud droplet effective radius (CDER) of marine water clouds from Moderate Resolution Imaging Spectroradiometer (MODIS) and CloudSat measurements, incorporating droplet size vertical inhomogeneity. Empirical orthogonal function (EOF) analysis is employed to reduce the degrees of freedom of the droplet size profile. The first two EOFs can explain 94% of the variability in the droplet size profile. Compared to the existing bispectral CDER retrieval from MODIS assuming plane parallel vertically homogeneous clouds, the new retrieval produces smaller CDER values in clouds in which the adiabatic growth process is dominant and larger CDER values in clouds in which the collision‐coalescence process is dominant. To evaluate the performance of the retrieval algorithm, we compare retrieved COT and CDER in this study with their MODIS and CloudSat counterparts on a pixel‐by‐pixel basis. CDER retrieval based on the vertically homogeneous assumption may be underestimated by 30% due to droplet size vertical inhomogeneity when COT is large and the collision‐coalescence process is dominant in the cloud. Retrieved CDER in conjunction with the two scores for EOFs can reconstruct the vertical profile of CDER, which is useful for cloud microphysical process studies. Furthermore, potential expansion of this algorithm to MODIS pixels without CloudSat collocations is discussed.

show abstract

Feasibility study of multi-pixel retrieval of optical thickness and droplet effective radius of inhomogeneous clouds using deep learning

Cited by 28 publications

References 34 publications

Evaluation of Himawari-8 surface downwelling solar radiation by ground-based measurements

Evaluation of Himawari-8 surface downwelling solar radiation by ground-based measurements

Remote sensing of cloud droplet radius profiles using solar reflectance from cloud sides – Part 1: Retrieval development and characterization

An Efficient Method for Microphysical Property Retrievals in Vertically Inhomogeneous Marine Water Clouds Using MODIS‐CloudSat Measurements

Contact Info

Product

Resources

About