# Spatial Surface Reflectance Retrievals for Visible/Shortwave Infrared Remote Sensing via Gaussian Process Priors

^{1}

^{2}

^{*}

## Abstract

**:**

## 1. Introduction

## 2. Method

#### 2.1. Optimal Estimation of Surface Reflectance

#### 2.2. Forward Model and Uncertainty

#### 2.3. Baseline Optimal Retrievals

#### 2.4. Prior

Algorithm 1: Simplified Optimal Spatial Inversion. |

#### 2.5. Naive Spatial Retrieval Structure

#### 2.6. Efficient Implementation

**x**denote the concatenated version of the latent state vector. For the update step shown in Equation (2) with $\alpha \approx {\left[{\nabla}_{x}^{2}Q\right]}^{-1}$ representing the constant matrix that results from the Levenberg-Marquardt approximation in (A3), we have

#### 2.7. Complexity

#### 2.8. Other Practical Considerations

## 3. Results

#### 3.1. Simulation Study

- Sample multiple surface reflectance states of vegetation, the most common of the priors described in Section 2.4. The atmospheric states are correlated according to their predetermined orientation following the technique outlined in Section 2.5.
- Simulate noisy AVIRIS-NG instrument radiance measurements corresponding to the sampled joint state using the built-in methods and configuration of the ISOFIT code [23]; the noise model is described in Section 2.2.
- Invert the simulated radiance measurements according to the implementation outlined in Section 2.6. Setting prior cross-pixel covariances to 0 results in individual retrievals as a special case.

Algorithm 2: Simulation Procedure: generate n pixels, compute correlated radiances, and invert. Repeat ${m}_{iter}$ times. |

#### 3.2. Application to Real Data

## 4. Discussion

## 5. Conclusions

## Author Contributions

## Funding

## Data Availability Statement

## Conflicts of Interest

## Abbreviations

VSWIR | Visible/Shortwave InfraRed |

AVIRIS-NG | Airborne Visible-Infrared Imaging Spectrometer-Next Generation |

ATREM | Atmosphere Removal |

LUT | Look-up Table |

UQ | Uncertainty Quantification |

OE | Optimal Estimation |

AOD | Aerosol Optical Depth |

RTM | Radiative Transfer Model |

## Appendix A. Appendix/Proofs

#### Appendix A.1. Iterative Optimization

#### Appendix A.2. Parameter Estimation

## References

- Space Studies Board; National Academies of Sciences, Engineering, and Medicine. Thriving on Our Changing Planet: A Decadal Strategy for Earth Observation from Space; National Academies Press: Washington, DC, USA, 2019. [Google Scholar]
- Chapman, J.W.; Thompson, D.R.; Helmlinger, M.C.; Bue, B.D.; Green, R.O.; Eastwood, M.L.; Geier, S.; Olson-Duvall, W.; Lundeen, S.R. Spectral and radiometric calibration of the next generation airborne visible infrared spectrometer (AVIRIS-NG). Remote Sens.
**2019**, 11, 2129. [Google Scholar] [CrossRef] [Green Version] - Cogliati, S.; Sarti, F.; Chiarantini, L.; Cosi, M.; Lorusso, R.; Lopinto, E.; Miglietta, F.; Genesio, L.; Guanter, L.; Damm, A.; et al. The PRISMA imaging spectroscopy mission: Overview and first performance analysis. Remote Sens. Environ.
**2021**, 262, 112499. [Google Scholar] [CrossRef] - Guanter, L.; Kaufmann, H.; Segl, K.; Foerster, S.; Rogass, C.; Chabrillat, S.; Kuester, T.; Hollstein, A.; Rossner, G.; Chlebek, C.; et al. The EnMAP spaceborne imaging spectroscopy mission for earth observation. Remote Sens.
**2015**, 7, 8830–8857. [Google Scholar] [CrossRef] [Green Version] - Connelly, D.S.; Thompson, D.R.; Mahowald, N.M.; Li, L.; Carmon, N.; Okin, G.S.; Green, R.O. The EMIT mission information yield for mineral dust radiative forcing. Remote Sens. Environ.
**2021**, 258, 112380. [Google Scholar] [CrossRef] - Nieke, J.; Rast, M. Towards the copernicus hyperspectral imaging mission for the environment (CHIME). In Proceedings of the Igarss 2018–2018 IEEE International Geoscience and Remote Sensing Symposium, Valencia, Spain, 22–27 July 2018; IEEE: Piscataway, NJ, USA, 2018; pp. 157–159. [Google Scholar]
- Alonso, K.; Bachmann, M.; Burch, K.; Carmona, E.; Cerra, D.; De los Reyes, R.; Dietrich, D.; Heiden, U.; Hölderlin, A.; Ickes, J.; et al. Data products, quality and validation of the DLR earth sensing imaging spectrometer (DESIS). Sensors
**2019**, 19, 4471. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Yokoya, N.; Iwasaki, A. Hyperspectral and multispectral data fusion mission on hyperspectral imager suite (HISUI). In Proceedings of the 2013 IEEE International Geoscience and Remote Sensing Symposium-IGARSS, Melbourne, VIC, Australia, 21–26 July 2013; IEEE: Piscataway, NJ, USA, 2013; pp. 4086–4089. [Google Scholar]
- Schaepman-Strub, G.; Schaepman, M.E.; Painter, T.H.; Dangel, S.; Martonchik, J.V. Reflectance quantities in optical remote sensing—Definitions and case studies. Remote Sens. Environ.
**2006**, 103, 27–42. [Google Scholar] [CrossRef] - Berk, A.; Conforti, P.; Kennett, R.; Perkins, T.; Hawes, F.; Van Den Bosch, J. MODTRAN
^{®}6: A major upgrade of the MODTRAN^{®}radiative transfer code. In Proceedings of the 2014 6th Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing (WHISPERS), Lausanne, Switzerland, 24–27 June 2014; IEEE: Piscataway, NJ, USA, 2014; pp. 1–4. [Google Scholar] - Emde, C.; Buras-Schnell, R.; Kylling, A.; Mayer, B.; Gasteiger, J.; Hamann, U.; Kylling, J.; Richter, B.; Pause, C.; Dowling, T.; et al. The libRadtran software package for radiative transfer calculations (version 2.0. 1). Geosci. Model Dev.
**2016**, 9, 1647–1672. [Google Scholar] [CrossRef] [Green Version] - Thompson, D.R.; Guanter, L.; Berk, A.; Gao, B.C.; Richter, R.; Schläpfer, D.; Thome, K.J. Retrieval of atmospheric parameters and surface reflectance from visible and shortwave infrared imaging spectroscopy data. Surv. Geophys.
**2019**, 40, 333–360. [Google Scholar] [CrossRef] [Green Version] - Gao, B.C.; Montes, M.J.; Li, R.R.; Dierssen, H.M.; Davis, C.O. An atmospheric correction algorithm for remote sensing of bright coastal waters using MODIS land and ocean channels in the solar spectral region. IEEE Trans. Geosci. Remote Sens.
**2007**, 45, 1835–1843. [Google Scholar] [CrossRef] - Thompson, D.R.; Gao, B.C.; Green, R.O.; Roberts, D.A.; Dennison, P.E.; Lundeen, S.R. Atmospheric correction for global mapping spectroscopy: ATREM advances for the HyspIRI preparatory campaign. Remote Sens. Environ.
**2015**, 167, 64–77. [Google Scholar] [CrossRef] - Healey, G.; Slater, D. Models and methods for automated material identification in hyperspectral imagery acquired under unknown illumination and atmospheric conditions. IEEE Trans. Geosci. Remote Sens.
**1999**, 37, 2706–2717. [Google Scholar] [CrossRef] [Green Version] - Acito, N.; Diani, M.; Corsini, G. Coupled subspace-based atmospheric compensation of LWIR hyperspectral data. IEEE Trans. Geosci. Remote Sens.
**2019**, 57, 5224–5238. [Google Scholar] [CrossRef] - Guanter, L.; Del Carmen González-Sanpedro, M.; Moreno, J. A method for the atmospheric correction of ENVISAT/MERIS data over land targets. Int. J. Remote Sens.
**2007**, 28, 709–728. [Google Scholar] [CrossRef] - Cressie, N. Statistics for Spatial Data, Revised Edition; John Wiley & Sons: New York, NY, USA, 1993. [Google Scholar]
- Thompson, D.R.; Kahn, B.H.; Brodrick, P.G.; Lebsock, M.D.; Richardson, M.; Green, R.O. Spectroscopic Imaging of Sub-Kilometer Spatial Structure in Lower Tropospheric Water Vapor. Atmos. Meas. Tech.
**2021**, 14, 2827–2840. [Google Scholar] [CrossRef] - Rodgers, C.D. Retrieval of atmospheric temperature and composition from remote measurements of thermal radiation. Rev. Geophys.
**1976**, 14, 609–624. [Google Scholar] [CrossRef] - Rodgers, C.D. Inverse Methods for Atmospheric Sounding: Theory and Practice; World Scientific: Singapore, 2000; Volume 2. [Google Scholar]
- Mouroulis, P.Z. Spectral and spatial uniformity in pushbroom imaging spectrometers. In Proceedings of the Imaging Spectrometry V. International Society for Optics and Photonics, Denver, CO, USA, 18–23 July 1999; Volume 3753, pp. 133–141. [Google Scholar]
- Thompson, D.R.; Natraj, V.; Green, R.O.; Helmlinger, M.C.; Gao, B.C.; Eastwood, M.L. Optimal estimation for imaging spectrometer atmospheric correction. Remote Sens. Environ.
**2018**, 216, 355–373. [Google Scholar] [CrossRef] - Dubovik, O.; Herman, M.; Holdak, A.; Lapyonok, T.; Tanré, D.; Deuzé, J.; Ducos, F.; Sinyuk, A.; Lopatin, A. Statistically optimized inversion algorithm for enhanced retrieval of aerosol properties from spectral multi-angle polarimetric satellite observations. Atmos. Meas. Tech.
**2011**, 4, 975–1018. [Google Scholar] [CrossRef] [Green Version] - Xu, F.; Diner, D.J.; Dubovik, O.; Schechner, Y. A correlated multi-pixel inversion approach for aerosol remote sensing. Remote Sens.
**2019**, 11, 746. [Google Scholar] [CrossRef] [Green Version] - Hobbs, J.; Katzfuss, M.; Zilber, D.; Brynjarsdóttir, J.; Mondal, A.; Berrocal, V. Spatial retrievals of atmospheric carbon dioxide from satellite observations. Remote Sens.
**2021**, 13, 571. [Google Scholar] [CrossRef] - Diner, D.J.; Martonchik, J.V.; Kahn, R.A.; Pinty, B.; Gobron, N.; Nelson, D.L.; Holben, B.N. Using angular and spectral shape similarity constraints to improve MISR aerosol and surface retrievals over land. Remote Sens. Environ.
**2005**, 94, 155–171. [Google Scholar] [CrossRef] - Lyapustin, A.; Wang, Y.; Laszlo, I.; Kahn, R.; Korkin, S.; Remer, L.; Levy, R.; Reid, J. Multiangle implementation of atmospheric correction (MAIAC): 2. Aerosol algorithm. J. Geophys. Res. Atmos.
**2011**, 116. [Google Scholar] [CrossRef] - Servera, J.V.; Rivera-Caicedo, J.P.; Verrelst, J.; Muñoz-Marí, J.; Sabater, N.; Berthelot, B.; Camps-Valls, G.; Moreno, J. Systematic Assessment of MODTRAN Emulators for Atmospheric Correction. IEEE Trans. Geosci. Remote Sens.
**2021**, 60, 1–17. [Google Scholar] [CrossRef] - Brodrick, P.G.; Thompson, D.R.; Fahlen, J.E.; Eastwood, M.L.; Sarture, C.M.; Lundeen, S.R.; Olson-Duvall, W.; Carmon, N.; Green, R.O. Generalized radiative transfer emulation for imaging spectroscopy reflectance retrievals. Remote Sens. Environ.
**2021**, 261, 112476. [Google Scholar] [CrossRef] - Thompson, D.R.; Braverman, A.; Brodrick, P.G.; Candela, A.; Carmon, N.; Clark, R.N.; Connelly, D.; Green, R.O.; Kokaly, R.F.; Li, L.; et al. Quantifying uncertainty for remote spectroscopy of surface composition. Remote Sens. Environ.
**2020**, 247, 111898. [Google Scholar] [CrossRef] - Ranganathan, A. The levenberg-marquardt algorithm. Tutoral LM Algorithm
**2004**, 11, 101–110. [Google Scholar] - Kinne, S. The MACv2 aerosol climatology. Tellus B Chem. Phys. Meteorol.
**2019**, 71, 1–21. [Google Scholar] [CrossRef] [Green Version] - Gneiting, T.; Raftery, A.E. Strictly proper scoring rules, prediction, and estimation. J. Am. Stat. Assoc.
**2007**, 102, 359–378. [Google Scholar] [CrossRef] - Gneiting, T.; Katzfuss, M. Probabilistic forecasting. Annu. Rev. Stat. Appl.
**2014**, 1, 125–151. [Google Scholar] [CrossRef] - Lindgren, F.; Rue, H.; Lindström, J. An explicit link between Gaussian fields and Gaussian Markov random fields: The stochastic partial differential equation approach. J. R. Stat. Soc. Ser. B
**2011**, 73, 423–498. [Google Scholar] [CrossRef] [Green Version] - Nychka, D.W.; Bandyopadhyay, S.; Hammerling, D.; Lindgren, F.; Sain, S.R. A multi-resolution Gaussian process model for the analysis of large spatial data sets. J. Comput. Graph. Stat.
**2015**, 24, 579–599. [Google Scholar] [CrossRef] [Green Version] - Katzfuss, M.; Guinness, J. A general framework for Vecchia approximations of Gaussian processes. Stat. Sci.
**2021**, 36, 124–141. [Google Scholar] [CrossRef] - Quiñonero-Candela, J.; Rasmussen, C.E. A unifying view of sparse approximate Gaussian process regression. J. Mach. Learn. Res.
**2005**, 6, 1939–1959. [Google Scholar] - Banerjee, S.; Gelfand, A.E.; Finley, A.O.; Sang, H. Gaussian predictive process models for large spatial data sets. J. R. Stat. Soc. Ser. B
**2008**, 70, 825–848. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Katzfuss, M.; Cressie, N. Spatio-temporal smoothing and EM estimation for massive remote-sensing data sets. J. Time Ser. Anal.
**2011**, 32, 430–446. [Google Scholar] [CrossRef]

**Figure 1.**Representative radiance and reflectance spectra. Red, green, and blue lines indicate visible color channels.

**Figure 2.**Inversions of simulated data showing the water vapor (units of g/cm${}^{2}$) and aerosol optical depth estimates across 10 pixels in 1D. The retrieved fields are more realistic for spatial (Spatial_Post) than for individual retrievals (Posterior). The post-hoc smoothing (Smooth_Post) can improve the individual retrievals but cannot overcome the bias.

**Figure 3.**Inversions of simulated data showing the aerosol optical depth estimates across 9 pixels on a $3\times 3$ grid. The spatial prior is more accurate than the independent inversions and comparable to post-hoc smoothing.

**Figure 4.**Inversions of simulated data showing the water vapor estimates (units of g/cm${}^{2}$) across 9 pixels on a $3\times 3$ grid. The spatial field better represents the truth; smoothing cannot overcome the bias in the independent estimates.

**Figure 5.**Prior score plots for 25 simulated realizations. The posterior estimates for the spatial model are usually closer to their priors than the independent models. The effect is weaker for the 2D case, suggesting that the improvement tends to be most pronounced with highly correlated data. (

**a**) Prior score results for a 1D array of 10 pixels. (

**b**) Prior score results for a 2D grid of 9 pixels.

**Figure 6.**The top box plots show the difference in log score between the spatial and independent models across 25 simulations. The $(25\%,50\%,75\%)$ quantile values are $(21.8,52.8,69.1)$ for the 1D case and $(-19.2,10.4,39.6)$ for the 2D case. The two lower box plots show the difference in log score between the spatial and smoothed independent model. Quantiles are $(-37.9,91.8,128.0)$ and $(-3326.4,16.9,67.3)$ for 1D and 2D respectively. Note the discontinuous x-axis.

**Figure 7.**The aerosol optical depth prediction for validation data at Ivanpah. The predictions are effectively identical, but the spatial retrievals are closer to the in situ measurement of 0.043.

**Figure 8.**The surface reflectance profiles are nearly identical for the Cuprite data, with scaling changes due to the estimation of atmospheric parameters. This suggests that independent inversions may be overestimating reflectance. Pixels 105, 254, and 255 are adjacent and the reflectance can be interpreted as a percent, so at a particular wavelength a reflectance of 0.4 means 40% of the incoming radiant energy is reflected.

**Figure 9.**The water vapor (units of g/cm${}^{2}$) estimates are noticeably smoother under the spatial models. The predicted fields are qualitatively more realistic and are a principled alternative to post-hoc smoothing.

**Figure 10.**For the Cuprite dataset, the aerosol optical depth prediction is susceptible to the surface state prediction (

**bottom right**), but smoothing with a spatial prior decreases the noise. The

**top left**figure shows inversions done 1 pixel at a time, compared to stripes of 1 × 5 pixels in the

**top right**figure and squares of 2 × 2 pixels in the

**bottom left**figure. The

**bottom right**figure shows topography visible at a single wavelength, 1600 nm.

**Figure 11.**A retrieved aerosol field under a spatial model is smoother than the independent retrievals and spreads out large estimates.

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |

© 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Zilber, D.; Thompson, D.R.; Katzfuss, M.; Natraj, V.; Hobbs, J.; Braverman, A.
Spatial Surface Reflectance Retrievals for Visible/Shortwave Infrared Remote Sensing via Gaussian Process Priors. *Remote Sens.* **2022**, *14*, 2183.
https://doi.org/10.3390/rs14092183

**AMA Style**

Zilber D, Thompson DR, Katzfuss M, Natraj V, Hobbs J, Braverman A.
Spatial Surface Reflectance Retrievals for Visible/Shortwave Infrared Remote Sensing via Gaussian Process Priors. *Remote Sensing*. 2022; 14(9):2183.
https://doi.org/10.3390/rs14092183

**Chicago/Turabian Style**

Zilber, Daniel, David R. Thompson, Matthias Katzfuss, Vijay Natraj, Jonathan Hobbs, and Amy Braverman.
2022. "Spatial Surface Reflectance Retrievals for Visible/Shortwave Infrared Remote Sensing via Gaussian Process Priors" *Remote Sensing* 14, no. 9: 2183.
https://doi.org/10.3390/rs14092183