Monitoring vegetation conditions is a critical activity for assessing food security in Africa. Rural populations relying on rain-fed agriculture and livestock grazing are highly exposed to large seasonal and inter-annual fluctuations in water availability [1]. Timely monitoring of the state, evolution, and productivity of vegetation (crops and pastures in particular) is important to conduct food emergency responses and plan for a long-term, resilient, development strategy [2]. In the last decades, a number of process-based and crop growth models has be used to simulate carbon and water fluxes, vegetation productivity and crop yield. Complex deterministic models are often constrained by the large number of parameters and input data needed, resulting in large uncertainties when applied over large areas where reliable parameterization is not available [3]. An alternative approach is represented by those simpler models that are able to capture the main biological processes and use a reduced and empirical parametrization that can be fine-tuned using remote sensing observations (e.g. [4-5]). In this contribution we explore the performances in tracking the gross primary production (GPP) in a semi-arid environment of a simple model assimilating remote sensing observations. Preliminary results of GPP modelling are presented for three years (2007-2009) over a sparse savanna ecosystem in the Sudan.

MERONI Michele;  REMBOLD Felix;  MIGLIAVACCA M.;  ARDO Jonas; 

2015-12-01

IEEE Geoscience and Remote Sensing Society

JRC94304

978-1-4799-7929-5,   

https://publications.jrc.ec.europa.eu/repository/handle/JRC94304,