Satellite estimation of biophysical parameters for ecological models

• Main Author: Prieto-Blanco, Ana
• Published: Swansea University 2007
• Online Access: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.752066

Ecological models are central to understanding of hydrological and carbon cycles. These models need input from Earth Observation data to function at regional to global scales. Requirements of these models and the satellite missions designed to fulfill them are reviewed to asses the present situation. The aim is to establish a better informed framework for the design and development of future satellite missions to meet the needs of ecological modellers. Key land surface parameters that can potentially be derived by remote sensing are analysed - leaf area index, leaf chlorophyll content, the fraction of photosynthetically-active radiation absorbed by the canopy and the fractional cover - as well as the aerosol optical thickness. Three coupled models - PROSPECT, FLIGHT and 6S - are used to simulate top of the atmosphere reflectances observed in a number of viewing directions and spectral wavebands within the visible and near-infrared domains. A preliminary study provides a sensitivity analysis of the top of the atmosphere reflectances to the input parameters and to the viewing angles. Finally, a methodology that links ecological model requirements to satellite instrument capabilities is presented. The three coupled models - PROSPECT, FLIGHT and 6S - are inverted using a simple technique based on look-up tables (LUTs). The LUT is used to estimate canopy biophysical variables from remotely-sensed data observed at the top of the atmosphere with different directional and spectral sampling configurations. The retrieval uncertainty is linked with the instrument radiometric accuracy by analysing the impact of different levels of radiometric noise at the input. The parameters retrieved in the inversion are used to drive two land-surface parameterization models, Biome-BGC and JULES. The effects of different configurations and of the radiometric noise on the NPP estimated are analysed. The technique is applied to evaluate desirable sensor characteristics for driving models of boreal forest productivity. The results are discussed in view of the definition of future satellites and the selection of the best measurement configuration for accurate estimation of canopy characteristics.