Discrete Ordinate Radiative Transfer Model With the Neural Network Based Eigenvalue Solver: proof Of Concept

2021 ◽  
pp. 56-62
Author(s):  
Dmitry S. Efremenko

Artificial neural networks are attracting increasing attention in various applications. They can be used as ‘universal approximations’, which substitute computationally expensive algorithms by relatively simple sequences of functions, which simulate a reaction of a set of neurons to the incoming signal. In particular, neural networks have proved to be efficient for parameterization of the computationally expensive radiative transfer models (RTMs) in atmospheric remote sensing. Although a direct substitution of RTMs by neural networks can lead to the multiple performance enhancements, such an approach has certain drawbacks, such as loss of generality, robustness issues, etc. In this regard, the neural network is usually trained for a specific application, predefined atmospheric scenarios and a given spectrometer. In this paper a new concept of neural-network based RTMs is examined, in which the neural network substitutes not the whole RTM but rather a part of it (the eigenvalue solver), thereby reducing the computational time while maintaining its generality. The explicit dependencies on geometry of observation and optical thickness of the medium are excluded from training. It is shown that although the speedup factor due to this approach is modest (around 3 times against 103 speed up factor of other approaches reported in recent papers), the resulting neural network is flexible and easy to train. It can be used for arbitrary number of atmospheric layers. Moreover, this approach can be used in conjunction with any RTMs based on the discrete ordinate method. The neural network is applied for simulations of the radiances at the top of the atmosphere in the Huggins band.

2007 ◽  
Vol 64 (11) ◽  
pp. 3854-3864 ◽  
Author(s):  
K. Franklin Evans

Abstract The spherical harmonics discrete ordinate method for plane-parallel data assimilation (SHDOMPPDA) model is an unpolarized plane-parallel radiative transfer forward model, with corresponding tangent linear and adjoint models, suitable for use in assimilating cloudy sky visible and infrared radiances. It is derived from the spherical harmonics discrete ordinate method plane-parallel (SHDOMPP, also described in this article) version of the spherical harmonics discrete ordinate method (SHDOM) model for three-dimensional atmospheric radiative transfer. The inputs to the SHDOMPPDA forward model are profiles of pressure, temperature, water vapor, and mass mixing ratio and number concentration for a number of hydrometeor species. Hydrometeor optical properties, including detailed phase functions, are determined from lookup tables as a function of mass mean radius. The SHDOMPP and SHDOMPPDA algorithms and construction of the tangent-linear and adjoint models are described. The SHDOMPPDA forward model is validated against the Discrete Ordinate Radiative Transfer Model (DISORT) by comparing upwelling radiances in multiple directions from 100 cloud model columns at visible and midinfrared wavelengths. For this test in optically thick clouds the computational time for SHDOMPPDA is comparable to DISORT for visible reflection, and roughly 5 times faster for thermal emission. The tangent linear and adjoint models are validated by comparison to finite differencing of the forward model.


2021 ◽  
Vol 13 (2) ◽  
pp. 270
Author(s):  
Adrian Doicu ◽  
Dmitry S. Efremenko ◽  
Thomas Trautmann

An algorithm for the retrieval of total column amount of trace gases in a multi-dimensional atmosphere is designed. The algorithm uses (i) certain differential radiance models with internal and external closures as inversion models, (ii) the iteratively regularized Gauss–Newton method as a regularization tool, and (iii) the spherical harmonics discrete ordinate method (SHDOM) as linearized radiative transfer model. For efficiency reasons, SHDOM is equipped with a spectral acceleration approach that combines the correlated k-distribution method with the principal component analysis. The algorithm is used to retrieve the total column amount of nitrogen for two- and three-dimensional cloudy scenes. Although for three-dimensional geometries, the computational time is high, the main concepts of the algorithm are correct and the retrieval results are accurate.


Sensors ◽  
2019 ◽  
Vol 19 (7) ◽  
pp. 1532 ◽  
Author(s):  
Guido Masiello ◽  
Carmine Serio ◽  
Sara Venafra ◽  
Laurent Poutier ◽  
Frank-M. Göttsche

Timely processing of observations from multi-spectral imagers, such as SEVIRI (Spinning Enhanced Visible and Infrared Imager), largely depends on fast radiative transfer calculations. This paper mostly concerns the development and implementation of a new forward model for SEVIRI to be applied to real time processing of infrared radiances. The new radiative transfer model improves computational time by a factor of ≈7 compared to the previous versions and makes it possible to process SEVIRI data at nearly real time. The new forward model has been applied for the retrieval of surface parameters. Although the scheme can be applied for the simultaneous retrieval of temperature and emissivity, the paper mostly focuses on emissivity. The inverse scheme relies on a Kalman filter approach, which allows us to exploit a sequential processing of SEVIRI observations. Based on the new forward model, the paper also presents a validation retrieval performed with in situ observations acquired during a field experiment carried out in 2017 at Gobabeb (Namib desert) validation station. Furthermore, a comparison with IASI (Infrared Atmospheric Sounder Interferometer) emissivity retrievals has been performed as well. It has been found that the retrieved emissivities are in good agreement with each other and with in situ observations, i.e., average differences are generally well below 0.01.


2018 ◽  
Vol 75 (7) ◽  
pp. 2217-2233 ◽  
Author(s):  
Guanglin Tang ◽  
Ping Yang ◽  
George W. Kattawar ◽  
Xianglei Huang ◽  
Eli J. Mlawer ◽  
...  

Abstract Cloud longwave scattering is generally neglected in general circulation models (GCMs), but it plays a significant and highly uncertain role in the atmospheric energy budget as demonstrated in recent studies. To reduce the errors caused by neglecting cloud longwave scattering, two new radiance adjustment methods are developed that retain the computational efficiency of broadband radiative transfer simulations. In particular, two existing scaling methods and the two new adjustment methods are implemented in the Rapid Radiative Transfer Model (RRTM). The results are then compared with those based on the Discrete Ordinate Radiative Transfer model (DISORT) that explicitly accounts for multiple scattering by clouds. The two scaling methods are shown to improve the accuracy of radiative transfer simulations for optically thin clouds but not effectively for optically thick clouds. However, the adjustment methods reduce computational errors over a wide range, from optically thin to thick clouds. With the adjustment methods, the errors resulting from neglecting cloud longwave scattering are reduced to less than 2 W m−2 for the upward irradiance at the top of the atmosphere and less than 0.5 W m−2 for the surface downward irradiance. The adjustment schemes prove to be more accurate and efficient than a four-stream approximation that explicitly accounts for multiple scattering. The neglect of cloud longwave scattering results in an underestimate of the surface downward irradiance (cooling effect), but the errors are almost eliminated by the adjustment methods (warming effect).


2012 ◽  
Vol 500 ◽  
pp. 390-396 ◽  
Author(s):  
Sheng Lan Zhang ◽  
Li Sheng Xu ◽  
Ji Lie Ding ◽  
Hai Lei Liu ◽  
Xiao Bo Deng

A neural network (NN) based algorithm for retrieval of precipitable water vapor (PWV) from the Atmospheric Infrared Sounder (AIRS) observations is proposed. An exact radial basis function (RBF) network is selected, in which the at-sensor brightness temperatures are the input variables, and PWV is the output variable. The training data sets for the RBF network are mainly simulated from the fast radiative transfer model (Community Radiative Transfer Model, CRTM) and the latest global assimilation data. The algorithm is validated by retrieving the PWV over west area in China using AIRS data. Compared with the AIRS PWV products, the RMSE of the PWV retrieved by our algorithm is 0.67 g/cm2, and a comparison between the retrieved PWV and radiosonde data is carried out. The result suggests that the RBF neural network based algorithm is applicable and feasible in actual conditions. Furthermore, spatial resolution of water vapor derived by RBF neural network is superior as compared to that of AIRS-L 2 standard product. Finally a PCA scheme is used for the preliminary investigation of the compression of AIRS high dimension observations.


2019 ◽  
Vol 12 (12) ◽  
pp. 6619-6634 ◽  
Author(s):  
Swadhin Nanda ◽  
Martin de Graaf ◽  
J. Pepijn Veefkind ◽  
Mark ter Linden ◽  
Maarten Sneep ◽  
...  

Abstract. To retrieve aerosol properties from satellite measurements of the oxygen A-band in the near-infrared, a line-by-line radiative transfer model implementation requires a large number of calculations. These calculations severely restrict a retrieval algorithm's operational capability as it can take several minutes to retrieve the aerosol layer height for a single ground pixel. This paper proposes a forward modelling approach using artificial neural networks to speed up the retrieval algorithm. The forward model outputs are trained into a set of neural network models to completely replace line-by-line calculations in the operational processor. Results comparing the forward model to the neural network alternative show an encouraging outcome with good agreement between the two when they are applied to retrieval scenarios using both synthetic and real measured spectra from TROPOMI (TROPOspheric Monitoring Instrument) on board the European Space Agency (ESA) Sentinel-5 Precursor mission. With an enhancement of the computational speed by 3 orders of magnitude, TROPOMI's operational aerosol layer height processor is now able to retrieve aerosol layer heights well within operational capacity.


Author(s):  
Guido Masiello ◽  
Carmine Serio ◽  
Sara Venafra ◽  
Laurent Poutier ◽  
Frank-M. Göttsche

Timely processing of observations from hyper-spectral imagers, such as SEVIRI (Spinning Enhanced Visible and Infrared Imager), largely depends on fast radiative transfer calculations. This paper mostly concerns the development and implementation of a new forward model for SEVIRI to be applied to real time processing of infrared radiances for the physical retrieval of surface temperature and emissivity. The new radiative transfer model improves computational time by a factor of ≈ 7 compared to the previous versions and makes it possible to process SEVIRI data at nearly real time. The new forward model has been applied for the simultaneous retrieval of surface temperature and emissivity in three infrared channels (8.7, 10.8, 12 μm). The inverse scheme relies on a Kalman filter approach, which allows us to exploit a sequential processing of SEVIRI observations. Based on the new forward model, the paper also presents a validation retrieval performed with in situ observations acquired during a field experiment carried out in 2017 at Gobabeb (Namib desert) validation station. Furthermore, a comparison with IASI (Infrared Atmospheric Sounder Interferometer) emissivity retrievals has been performed as well. It has been found that the retrieved emissivities are in good agreement with each other and with in situ observations, i.e. average differences are generally well below 0.01.


2011 ◽  
Vol 11 (20) ◽  
pp. 10471-10485 ◽  
Author(s):  
A. Kylling ◽  
B. Mayer ◽  
M. Blumthaler

Abstract. Rotational Raman scattering in the Earth's atmosphere explains the filling-in of Fraunhofer lines in the solar spectrum. A new model including first-order rotational Raman scattering has been developed, based on a reimplementation of the versatile discrete ordinate radiative transfer (DISORT) solver in the C computer language. The solver is fully integrated in the freely available libRadtran radiative transfer package. A detailed description is given of the model including the spectral resolution and a spectral interpolation scheme that considerably speeds up the calculations. The model is used to demonstrate the effect of clouds on top and bottom of the atmosphere filling-in factors and differential optical depths. Cloud effects on vertical profiles of the filling-in factor are also presented. The spectral behaviour of the model is compared against measurements under thunderstorm and aerosol loaded conditions.


Sign in / Sign up

Export Citation Format

Share Document