EARTHSYSTEMS: Lisbon Doctoral School on Earth System Science, PhD Projects

Biosphere plays an important role in the climate system, controlling the cycle of carbon. Additionally, the activity of ecosystems is particularly sensitive to climate variability (Ciais et al., 2003, Gouveia et al., 2008). The regional changes observed in temperature, radiation and precipitation in recent decades seems to have caused an increase in global Net Primary Productivity (NNP), increasing CO2 fixation by the ecosystems (Nemani et al., 2003). Fire is the main agent of disturbance in the global terrestrial ecosystem (Bowman et al., 2009), affecting the net carbon balance both directly and indirectly, throughout the biomass burning that introduces carbon in the atmosphere (van der Werf et al., 2010) and the carbon loss by burned and died vegetation (Amiro et al., 2010). The spatio-temporal patterns of fires on a global scale are also related to climate variability such as El Niño (LePage et al., 2008) and extreme events on regional scales (Pereira et al., 2005), looking significantly affect the regional balance of CO2 (Bond-Lamberty et al., 2007). Quantifying the impact of fire on global carbon balance is required to better understanding the carbon dynamics and their changes (Running, 2008). Tropical forests and savanas contribute with about 60% of the terrestrial NPP (Field et al., 1998), however the frequent clouds and biomass-burning aerosols may lead to misleading interpretations (Samanta et al., 2011). The emergence of new platforms, sensors and satellites has arisen remarkable efforts to develop more sophisticated methods and algorithms. There are several remote sensing datasets available that can be used to derive vegetation proprieties. Currently the major NPP databases correspond to two distinct datasets– NASA-CASA project and GPP/NPP MODIS covering the periods 1982-1998 and 2000 to present, respectively. MODIS is certainly a strong tool for remote monitoring of tropical forests. However, there is a few number of ground observations to evaluate NPP MODIS accuracy in tropical areas. Recently, more focused validation studies on tropical regions indicate that MODIS17 NPP does not perform well in modeling field-measured tropical NPP (Samanta et al. 2010). The driving behind this work is related to the need to enhance the knowledge on fire impacts on vegetation dynamics and carbon cycle. This project intends to contribute to a detailed characterization of LSA-SAF/VP, including the Fraction Vegetation Cover (FVC), Leaf Area Index (LAI), Fraction of Absorbed Photosynthetic Active Radiation (FAPAR) and NDVI. The availability of other Vegetation Parameters, among the NDVI, is expected to improve the quality of new NPP dataset, as saturation of NDVI in closed canopy was found as one of the main reasons of NPP MODIS lack of accuracy. The add value of using FVC or FAPAR will be assessed. In this project the NPP CASA algorithm will be adapted to LSA SAF/VP in order to develop a new NPP database over Africa (NPP LSA-SAF). This dataset will be compared with the available datasets (CASA and MODIS). Special attention EARTHSYSTEMS: Lisbon Doctoral School on Earth System Science, PhD Projects 8 will be devoted to Africa tropical areas. In particular, the NPP LSA-SAF will be used to analyze the impact of large fires in vegetation dynamics in order to assess their direct and indirect contribution to the carbon balance in the short and medium/long term.