Assessing the Performance of HYPERION in Relation to Eucalypt Biochemistry: Preliminary Project Design and Specifications Nicholas C Coops 1 , Marie-Louise Smith 2 , Mary E Martin 3 , Scott V Ollinger 3 , Alex Held 4 and Steve J Dury 5 1 CSIRO Forestry and Forest Products Private Bag 10, Clayton South, Victoria 3169, Australia. E-mail: n.coops@ffp.csiro.au 2 USDA Forest Service Northeastern Research Station. P.O. Box 640, Durham, NH 03824, USA. 3 Complex Systems Research Center, University of New Hampshire Durham, NH 03824, USA. 4 CSIRO Land and Water. GPO Box 1666, Canberra ACT 2601, Australia. 5 CSIRO Forestry and Forest Products PO Box E4008, Kingston ACT 2604, Australia. Abstract- Vegetation function and dynamics are key parameters in terrestrial carbon cycle models. The strong linkages between foliar nitrogen, photosynthetic capacity and ecosystem productivity makes the development of methods to characterize spatial patterns of canopy bio-chemistry a potentially powerful approach for estimating forest carbon fluxes at a variety of scales. The challenge is to extrapolate results from individual leaves to regional scales to estimate carbon cycles across the landscape using combinations of inverse modeling and remote sensing. Hyperspectral remote sensing methods are advancing rapidly and offer the promise of estimating canopy pigment, bio-chemistry and water content dynamics, which can in turn be linked to carbon assimilation, forest growth and photosynthetic capacity models. This study was undertaken across eucalypt forest near Tumbarumba (Bago-Maragle State Forest), Australia which has a number of eucalypt species, ranging in productivity and age. EO-1 Hyperion imagery has been obtained and a detailed field program undertaken in February 2001. This program involved plot establishment, collected of standard forestry inventory data and the collection of leaf samples. From the sampled eucalypt leaves, individual leaf spectra were recorded, samples dried and a number of foliage bio-physical and bio-chemistry variables analysed. This dataset will form the basis of a comparison with spectral information available from the HYPERION sensor. I. INTRODUCTION The carbon and nitrogen cycle in forest ecosystems are linked through a number of common ecological processes [1]. Research in the past decade has demonstrated consistent, strong and generalisable relationships between foliar nitrogen and rates of net photosynthesis and leaf respiration [2]. At broader temporal and spatial scales, canopy nitrogen has been related to annual net primary production, litterfall N and N mineralization [3,4]. As a result, estimates of foliar nitrogen chemistry can provide an insight into terrestrial C and N cycles and thus be an useful indicator of ecosystem productivity [5]. Recent advancements in hyperspectral remote sensing indicate that a variety of independent measures of canopy properties are possible from satellite and airborne sensors including the concentration and total amount of nitrogen in canopies, leaf water content, lignin concentrations and other foliage nutrient status indicators [6,7,8]. Incorporation of these estimates of chemical composition into process models, as input parameters or constraints, will improve predictive capability and enable validation of regional assessments of growth and carbon accumulation. The use of hyperspectral remote sensing to predict canopy properties has been relatively limited, owing in part, to limited availability of hyperspectral data and to large data processing requirements. Nevertheless, interest in the approach has persisted and several new hyperspectral earth-observation sensors, such as NASA’s Hyperion instrument, which was successfully launched in November 2000 as part of the NASA New Millennium Program EO-1 satellite platform, are now available. II. STUDY AREA AND DATA COLLECTION A. Study Area The Tumbarumba (or Bago-Maragle) study area (E 148º 15' S 35º 45') is in southern New South Wales, Australia and