Adaptive Data Screening for Multi-Angle Polarimetric Aerosol and Ocean Color Remote Sensing Accelerated by Deep Learning

Abstract

Remote sensing measurements from multi-angle polarimeters (MAPs) contain rich aerosol microphysical property information, and these sensors have been used to perform retrievals in optically complex atmosphere and ocean systems. Previous studies have concluded that, generally, five moderately separated viewing angles in each spectral band provide sufficient accuracy for aerosol property retrievals, with performance gradually saturating as angles are added above that threshold. The Hyper-Angular Rainbow Polarimeter (HARP) instruments provide high angular sampling with a total of 90–120 unique angles across four bands, a capability developed mainly for liquid cloud retrievals. In practice, not all view angles are optimal for aerosol retrievals due to impacts of clouds, sunglint, and other impediments. The many viewing angles of HARP can provide resilience to these effects, if the impacted views are screened from the dataset, as the remaining views may be sufficient for successful analysis. In this study, we discuss how the number of available viewing angles impacts aerosol and ocean color retrieval uncertainties, as applied to two versions of the HARP instrument. AirHARP is an airborne prototype that was deployed in the ACEPOL field campaign, while HARP2 is an instrument in development for the upcoming NASA Plankton, Aerosol, Cloud, ocean Ecosystem (PACE) mission. Based on synthetic data, we find that a total of 20–30 angles across all bands (i.e., five to eight viewing angles per band) are sufficient to achieve good retrieval performance. Following from this result, we develop an adaptive multi-angle polarimetric data screening (MAPDS) approach to evaluate data quality by comparing measurements with their best-fitted forward model. The FastMAPOL retrieval algorithm is used to retrieve scene geophysical values, by matching an efficient, deep learning-based, radiative transfer emulator to observations. The data screening method effectively identifies and removes viewing angles affected by thin cirrus clouds and other anomalies, improving retrieval performance. This was tested with AirHARP data, and we found agreement with the High Spectral Resolution Lidar-2 (HSRL-2) aerosol data. The data screening approach can be applied to modern satellite remote sensing missions, such as PACE, where a large amount of multi-angle, hyperspectral, polarimetric measurements will be collected.

Multiple Angle Observations Would Benefit Visible Band Remote Sensing Using Night Lights

“…Multi‐angle views would therefore allow better detection of boats on moonlit nights, as the lunar reflection only affects some observing angles. In addition, in areas with frequent broken cloud cover, multi‐angle views increase the chance that at least one of the observation angles will have a clear view of the surface (Gao et al., 2021).…”

Effective uncertainty quantification for multi-angle polarimetric aerosol remote sensing over ocean, Part 1: performance evaluation and speed improvement

“…Joint aerosol/ocean color retrieval algorithms have been developed for a variety of spaceborne and airborne MAPs such as the Polarization and Directionality of the Earth's Reflectances (POLDER) instruments (Hasekamp et al, 2011;Dubovik et al, 2011Dubovik et al, , 2014Li et al, 2019;Hasekamp et al, 2019b;Chen et al, 2020), the Airborne Multiangle SpectroPolarimetric Imager (AirMSPI) (Xu et al, 2016(Xu et al, , 2019, the Spectro-Polarimeter for Planetary EXploration (SPEX) airborne (Fu and Hasekamp, 2018;Fu et al, 2020;Fan et al, 2019), SPEXone (spaceborne version of SPEX airborne) (Hasekamp et al, 2019b), the Research Scanning Polarimeter (RSP) (Chowdhary et al, 2005;Wu et al, 2015;Stamnes et al, 2018;Gao et al, 2018Gao et al, , 2019Gao et al, , 2020, the Directional Polarimetric Camera (DPC)/GaoFen-5 (Wang et al, 2014;Li et al, 2018), Airborne HyperAngular Rainbow Polarimeter (AirHARP) (Puthukkudy et al, 2020;Gao et al, 2021a, b), and HARP2 (the spaceborne version of AirHARP) (Gao et al, 2021b). The algorithms typically follow iterative optimization approaches utilizing a vector radiative transfer forward model, and simultaneously retrieve a suite of geophysical parameters.…”

“…In the absence of independent external truth, simulated retrievals are a useful tool. With MAPs, real uncertainties have been discussed for aerosols over ocean, land, and cloud by comparing retrievals with synthetic data and in-situ measurements, such as for POLDER (Hasekamp et al, 2011;Dubovik et al, 2011;Chen et al, 2020), RSP (Chowdhary et al, 2012;Stamnes et al, 2018;Gao et al, 2019;Fu et al, 2020), AirMSPI (Xu et al, 2016), SPEX Airborne (Fu et al, 2020), SPEXone (Hasekamp et al, 2019a), AirHARP Puthukkudy:2020aa, Gao:2021aa, Gao:2021bb and HARP2 (Gao et al, 2021b).…”

“…As Part 1 of a study series, this work focuses on the first two topics. The third topic has been partially addressed using an adaptive angular screening approach, described in (Gao et al, 2021b), to automatically remove MAP angles where the input uncertainty model is insufficient to describe forward model uncertainty due to contamination by cirrus clouds and other anomalies (Gao et al, 2021b). Note that noise correlation in the uncertainty model may impact retrieval results, though it is often ignored as assumed in this study (Knobelspiesse et al, 2012), and will be further addressed in the forthcoming Part 2 of this work.…”

“…Section 2 of this paper describes the FastMAPOL retrieval algorithm used in the study; Section 3 discusses the methodology in the retrieval uncertainty evaluation; Section 4. quantified the performance of retrievals uncertainties based on synthetic AirHARP and HARP2 data ; Section 5. applied the pixel-wise uncertainties on the retrievals from AirHARP field measurements; and Section 6 provides discussions and conclusions. The FastMAPOL algorithm (Gao et al, 2021a) uses neural network forward models of a coupled atmosphere-ocean system, and has been used to perform retrievals on synthetic and observed AirHARP measurements (Gao et al, 2021a) and synthetic HARP2 measurements (Gao et al, 2021b). In this section, we will first introduce the MAP measurements from the PACE mission and then review key components of the retrieval algorithm.…”