Andrew Banks

Novel techniques for the analysis of the TOA radiometric uncertainty

In the framework of the European Copernicus programme, the European Space Agency (ESA) has launched the Sentinel-2 (S2) Earth Observation (EO) mission which provides optical high spatial -resolution imagery over land and coastal areas. As part of this mission, a tool (named S2-RUT, from Sentinel-2 Radiometric Uncertainty Tool) estimates the radiometric uncertainties associated to each pixel using as input the top-of-atmosphere (TOA) reflectance factor images provided by ESA. The initial version of the tool has been implemented — code and user guide available1 — and integrated as part of the Sentinel Toolbox. The tool required the study of several radiometric uncertainty sources as well as the calculation and validation of the combined standard uncertainty in order to estimate the TOA reflectance factor uncertainty per pixel. Here we describe the recent research in order to accommodate novel uncertainty contributions to the TOA reflectance uncertainty estimates in future versions of the tool. The two contributions that we explore are the radiometric impact of the spectral knowledge and the uncertainty propagation of the resampling associated to the orthorectification process. The former is produced by the uncertainty associated to the spectral calibration as well as the spectral variations across the instrument focal plane and the instrument degradation. The latter results of the focal plane image propagation into the provided orthoimage. The uncertainty propagation depends on the radiance levels on the pixel neighbourhood and the pixel correlation in the temporal and spatial dimensions. Special effort has been made studying non-stable scenarios and the comparison with different interpolation methods.

Stray light effects in above-water remote-sensing reflectance from hyperspectral radiometers.

Stray light perturbations are unwanted distortions of the measured spectrum due to the nonideal performance of optical radiometers. Because of this, stray light characterization and correction is essential when accurate radiometric measurements are a necessity. In agreement with such a need, this study focused on stray light correction of hyperspectral radiometers widely applied for above-water measurements to determine the remote-sensing reflectance (RRS). Stray light of sample radiometers was experimentally characterized and a correction algorithm was developed and applied to field measurements performed in the Mediterranean Sea. Results indicate that mean stray light corrections are appreciable, with values generally varying from -1% to +1% in the 400-700 nm spectral region for downward irradiance and sky radiance, and from -1% to +4% for total radiance from the sea. Mean corrections for data products such as RRS exhibit values that depend on water type varying between -0.5% and +1% in the blue-green spectral region, with peaks up to 9% in the red in eutrophic waters. The possibility of using one common stray light correction matrix for the analyzed class of radiometers was also investigated. Results centered on RRS support such a feasibility at the expense of an increment of the uncertainty typically well below 0.5% in the blue-green and up to 1% in the red, assuming sensors are based on spectrographs from the same production batch.

Assessment of Satellite-Derived Surface Reflectances by NASA’s CAR Airborne Radiometer over Railroad Valley Playa

CAR (Cloud Absorption Radiometer) is a multi-angular and multi-spectral airborne radiometer instrument, whose radiometric and geometric characteristics are well calibrated and adjusted before and after each flight campaign. CAR was built by NASA (National Aeronautics and Space Administration) in 1984. On 16 May 2008, a CAR flight campaign took place over the well-known calibration and validation site of Railroad Valley in Nevada, USA (38.504°N, 115.692°W). The campaign coincided with the overpasses of several key EO (Earth Observation) satellites such as Landsat-7, Envisat and Terra. Thus, there are nearly simultaneous measurements from these satellites and the CAR airborne sensor over the same calibration site. The CAR spectral bands are close to those of most EO satellites. CAR has the ability to cover the whole range of azimuth view angles and a variety of zenith angles depending on altitude and, as a consequence, the biases seen between satellite and CAR measurements due to both unmatched spectral bands and unmatched angles can be significantly reduced. A comparison is presented here between CAR’s land surface reflectance (BRF or Bidirectional Reflectance Factor) with those derived from Terra/MODIS (MOD09 and MAIAC), Terra/MISR, Envisat/MERIS and Landsat-7. In this study, we utilized CAR data from low altitude flights (approx. 180 m above the surface) in order to minimize the effects of the atmosphere on these measurements and then obtain a valuable ground-truth data set of surface reflectance. Furthermore, this study shows that differences between measurements caused by surface heterogeneity can be tolerated, thanks to the high homogeneity of the study site on the one hand, and on the other hand, to the spatial sampling and the large number of CAR samples. These results demonstrate that satellite BRF measurements over this site are in good agreement with CAR with variable biases across different spectral bands. This is most likely due to residual aerosol effects in the EO derived reflectances.

Providing uncertainty estimates of the Sentinel-2 top-of-atmosphere measurements for radiometric validation activities

ABSTRACT As part of the Sentinel-2 mission, a Radiometric Uncertainty Tool (RUT) has been recently released to the community. This tool estimates the Sentinel-2 radiometric uncertainty associated with each pixel in the top-of-atmosphere (TOA) reflectance factor images provided by the European Space Agency (ESA). The use of such information enables users to assess the “fitness for purpose” of the data to their specific application. The work described here summarises the efforts and results of integrating the RUT for radiometric validation activities for the Sentinel-2 mission. Starting from the results provided by the RUT, the focus will be on providing a methodology to calculate the uncertainty associated with the mean TOA reflectance factor in a Region of Interest (ROI). Two different methods – one simple method directly using the RUT and a more rigorous one based on Monte Carlo method (MCM) propagation – are proposed and compared. These two methods focus on the effect of the spectral, spatial and temporal correlation of the errors in different ROI pixels and the impact of correlation on the uncertainty associated with the mean TOA reflectance factor. The study has also considered the impact of uncertainty contributions not included in the first version of the RUT.

A comparison of validation and vicarious calibration of high and medium resolution satellite-borne sensors using RadCalNet

The Radiometric Calibration Network (RadCalNet, www.radcalnet.org) routinely provides top-of-atmosphere (TOA) reflectance data from instrumented ground sites. The data represents the nadir view of the ground for different sites that cover areas ranging from 50 m × 50 m to 1 km x 1 km. The smaller sites can only be used with high resolution sensors (≤ 30 m), but the larger sites, such as Railroad Valley (RRV) in Nevada can also be used for the validation or vicarious calibration of medium resolution sensors (> 250 m spatial resolution). Prior to utilising RadCalNet data in this manner, this paper describes the application of a high and a medium resolution sensor to assess potential biases between the RadCalNet data and satellite data at two different spatial resolutions. Results are shown for initial comparisons over RRV for the high resolution Sentinel-2 MultiSpectral Instrument (S2-MSI) and the medium resolution Sentinel-3 Ocean and Land Colour Instrument (S3-OLCI), and indicate the potential for RadCalNet to validate and vicariously calibrate sensors with differing spatial resolutions. The comparison analysis includes taking into account the temporal differences between the Sentinel-2 and Sentinel-3 overpasses and the time of RadCalNet data collection, as well as the spectral response functions (SRF) of the bands for both instruments. The comparison against the RRV site has shown there are significant biases between the RadCalNet data and S2-MSI and S3-OLCI for non-nadir viewing geometries that may be due to directional viewing and illumination effects and the non-Lambertian character of the RadCalNet RRV site.