GROSS PRIMARY PRODUCTIVITY ESTIMATION USING MULTI-ANGULAR MEASUREMENTS FROM SMALL SATELLITE CLUSTERS Sreeja Nag 1,2,3 , Charles Gatebe 2,3 , Thomas Hilker 4 , Forrest Hall 2,5 , Lars Dyrud 6 , Olivier de Weck 1 1 Massachusetts Institute of Technology, Cambridge, MA, USA 2 NASA Goddard Space Flight Center, Greenbelt, MD, USA 3 Universities Space Research Association, Columbia, MD, USA 4 Oregon State University, Corvallis, OR, USA 5 University of Maryland, Baltimore Country, MD, USA 6 Draper Laboratory, Cambridge, MA, USA ABSTRACT Gross primary productivity is an excellent metric of how much forests act as carbon dioxide sinks but currently have up to 40% uncertainty in their global estimates. A large proportion of the uncertainty has been attributed to artifacts in the sun-sensor geometry of monolithic spacecrafts leading to insufficient sampling of the bi-directional reflectance of vegetation. This paper proposes to use small satellite clusters with spectrometers as a new measurement solution to improve angular sampling locally and scale up measurements globally. Initial observing system simulations with four satellites launched as secondary payloads via the ISS and operating in different imaging modes show error estimates of less than 12% when compared to dense airborne measurements, a 50% improvement to the worst case error produced by corresponding monoliths. Index Terms— BRDF, satellites, constellation, PRI 1. INTRODUCTION TO THE SCIENCE PROBLEM Quantifying the extent to which forests and vegetation act as a sink for atmospheric carbon dioxide is very important to estimate carbon feedbacks of vegetation in response to global climate change [1]. Deforestation and forest degradation accounts for 12% of anthropogenic carbon emissions, which have nearly doubled in the past 30 years[2]. Current Gross Primary Productivity (GPP) estimates show uncertainties up to 40% in the terrestrial carbon uptake [3]. GPP is the product of photosynthetic efficiency (ε) and photosynthetically active radiation (APAR) absorbed by the plant. In recent studies, we have shown that measurements of vegetation reflectance at multiple angles can be used to estimate changes in protective leaf pigments as a function of shadow fraction [4]. These protective leaf pigments (xanthophylls), regulate light use efficiency in leaves and can be measured by means of the Photosynthetic Reflectance Index (PRI), a normalized difference index that is sensitive to the xanthophyll absorption at 531nm. Photosynthetic efficiency is the differential of PRI with respect to the shadow fraction [4],[5]. This differential can be estimated from the bi- directional reflectance function (BRDF) of PRI. BRDF describes the directional and spectral variation of reflectance of an optically thick surface element at any instant of time; it is a function of the material surface properties and roughness [7]. Measurement of the BRDF of PRI is inaccurate using existing space-borne sensors. Existing imaging spectrometers such as MODIS or MISR in sun-synchronous orbits, provide insufficient angular coverage during a single overpass. Recent studies have also shown an overestimation of the greening of Amazon forests during the dry season due to seasonal artifacts in MODIS’ sun-sensor geometry[8]. Global and frequent BRDF is impractical to estimate using towers and airborne instruments. Therefore, small satellite clusters on repeat track orbits with VNIR spectrometers have been proposed for the purpose [9]–[11]. Usage of angular reflectance data from the CHRIS instrument on the PROBA spacecraft has shown to bring GPP estimation errors down to 10% [3], however PROBA is not designed to measure GPP and does not provide the temporal resolution and global coverage required to do so. One possible solution would be to use constellations of CubeSats to obtain measurements of GPP with a frequent temporal repeat and global coverage [5],[9],[10]. 2. CLUSTER EVALUATION METHODOLOGY This paper proposes a new measurement solution to make multi-angular reflectance measurements using small satellites in close formations called clusters. It uses an observing system simulation experiment (OSSE) to demonstrate the potential improvement in GPP estimation using the new solution and design the most optimal cluster architecture. An architecture is defined here as a unique combination of orbits for the satellites in the cluster. Data from airborne campaigns of the Cloud Absorption