UNCLASSIFIED

INTRODUCTION

This paper introduces a new capability for performing the vicarious radiometric calibration of high, medium and low spatial resolution sensors. The SPecular Array Radiometric Calibration (SPARC) method employs convex mirrors to create two arrays of calibration targets for deriving absolute calibration coefficients of Earth remote sensing systems in the solar reflective spectrum. The first is an array of single mirrors used to oversample the sensor’s point spread function (PSF) providing necessary spatial quality information needed to perform the radiometric calibration of a sensor when viewing small targets. The second is a set of panels consisting of multiple mirrors designed to stimulate detector response with known at-sensor irradiance traceable to the exo-atmospheric solar spectral constant. The combination of these arrays with a targeting station is the basis for a new, on-demand commercial calibration network called FLARE.

SPARC METHOD – AN INNOVATION IN CALIBRATION

Current best practice for verifying absolute radiometric requirements of remote sensing imaging systems is always based on viewing a traceable extended source that uniformly illuminates many if not all detectors on the focal plane. This holds true whether the calibration images are recorded in pre-flight testing or in-flight viewing on-board or vicarious sources. Illuminating a large number of detectors establishes a mean response to the calibration radiance with high precision and averages out the blurring effects induced by the sensor’s optical system. The procedure derives radiometric gain coefficients with uncertainties that can be maintained through the image processing chain to higher level products but only for extended uniform scenes for which the response is unaffected by the sensor’s image quality performance.

Many targets of interest are small, typically subpixel up to a few pixels in diameter. As a result, the spatial performance of the sensor and any pixel resampling performed in the processing chain can become a significant error contributor to the radiometric knowledge of small targets in the scene. This leads to several factors to be accounted for in the accuracy and traceability book-keeping specific to small targets:

1.) Non-uniform energy-on-detector sensitivity: In this effect, the shape of the small target image profile projected on the focal plane detectors produces a high contrast relative to the scene background in which a single detector may ultimately respond differently because of high spatial frequency illumination gradients across the detector not present when under calibration.

2.)  Sensor blurring: With sensor blurring, the energy originating geometrically within the projected ground location of a single detector, containing the small target, becomes recorded by adjacent detectors outside the target area because of the system point-spread function (PSF). This induces a probabilistic uncertainty in the location of the signal’s origin and, thus, also the target radiance without some a priori knowledge of the targets shape and size.

3.)  Effects of post processing on image quality: In post processing, band-to-band registration and geometric calibration apply resampling methods that ultimately introduce increased radiometric uncertainty. The result is additional spatial smearing of information between the small target and its background specific to the resampling kernel and routine used to redefine individual pixel digital values, size and orientation.

Due to these factors, there is a need to develop a comprehensive radiometric error propagation analysis process for small targets that can benefit by introducing vicarious ground references that reveal and validate these artifacts within image data collected under operational conditions.

The vicarious calibration technique known as the Specular Array Radiometric Calibration (SPARC) method is demonstrated to provide a highly flexible and robust approach for characterizing the performance of a sensor imaging small targets. Its general use is shown in Figure 1. This technology is the invention of Dr. Stephen Schiller of Raytheon and has a well-documented record of published performance (see Appendix 1). SPARC delivers small intensity reference targets, made from a collection of facets consisting of convex mirrors, which can be built to any size, shape, brightness or spectral composition with accurate radiometric traceability to the solar radiometric scale. These types of synthetic targets may be a dynamic replacement for naturally observed phenomena (PICS sites, deserts, etc).

THE FLARE NETWORK – ON DEMAND LOW COST CALIBRATION

Labsphere has created a new automated calibration network called FLARE based on the SPARC mirrors and their established radiometric technique. FLARE is an acronym that stands for Field, Line- of-sight Automated Radiance Exposure. The benefit of FLARE is a point-able system with improved radiometric performance knowledge compared to other in-flight vicarious techniques through reduced uncertainties in target reflectance, atmospheric effects, and temporal variability. The only ground truth needed is the measurement of atmospheric transmittance (although upwelling and downwelling radiance, surround reflectance, and mirror characteristics are also collected by FLARE stations). These mirror arrays are sub-pixel size. The simplification of calibration targets and ground truth collection in the FLARE Network makes the deployment more cost effective and/or portable. FLARE creates the opportunity to strategically locate and imbed spectral, spatial, and radiometric targets at study sites providing references that improve a sensor’s interactivity as a phenomenological tool. This technology is

patented by Raytheon Inc. and licensed to Labsphere, Inc. in order to create a commercially available network for calibration.

The FLARE Network concept is an innovative, automated, on-demand, ground-based radiometric calibration station for in-flight sensors. These stations will track and point to sensor (satellite/airborne) locations and provide absolute ground-truth calibrations and direct evaluation of image quality. They hold the promise to provide standardization among imaging satellites and Analysis Ready Data (ARD) at a first level of image capture (L0 or L1). An extended network of these systems will be the basis for fractional-cost calibrations, digital subscription services and worldwide common calibration operation for all airborne and satellite imagers as shown in Figures 2 and 3.

Initial development of the commercial FLARE network is targeting an Alpha system for evaluation by university and commercial partners in early CY2020, which will be followed by a more feature-rich Beta system by Summer of 2020 that includes the complete subscription-based revenue platform. The network will seek commercial imaging and Government customers who are looking for a cost-controlled, on-demand, traceable calibration system for their satellite and airborne instruments. The network will expand as quickly as the market uptake of the concept allows. A model of this expansion is shown in Figure 4. The intent of this information is to seek customers as a faster means for expanding this capable network to speed the possibility of Analysis Ready Data from all sensor platforms.

ENHANCING CURRENT VICARIOUS CALIBRATION METHODS WITH FLARE

The FLARE network will offer a breakthrough technology to the satellite and airborne imagery community in many ways:

• Tailored to imagers, FOVs & instruments (GSD, angle independent, & bands)
• Agile system configuration, innovative uses and universal placement capability
• Tied to “Big Satellite” methods: Same or better traceable uncertainty & quality metrics
• Traceably link satellite (lower resolution) to higher resolution airborne, UAV high resolution “under-fly” data.
• Provide on-demand calibration services: when and where needed…and as often as needed
• Subscription and per-look business model: Fractional & controlled cost versus conventional calibration methods.
• Enables Analysis Ready Data as close as possible to the initial image processing levels.

Additionally, FLARE can offer some revolutionary capability to solve problems that are not currently addressable by common techniques. A discussion of specific problems and solutions is embedded in Table 1 through Table 5 below.