10-13 November 2015
Villa Romanazzi Carducci
Europe/Rome timezone

High-throughput Processing of SEBAL in the EUBrazilCC Federated Cloud

Not scheduled
Villa Romanazzi Carducci

Villa Romanazzi Carducci

Via G. Capruzzi, 326 70124 Bari Italy


Francisco Brasileiro (Universidade Federal de Campina Grande)


EU-Brazil Cloud Connect (EUBrazilCC) is a project funded under the second EU-Brazil coordinated call under the topic “Cloud Computing for Science”. The overarching objective of the project is to drive cooperation between Europe and Brazil by strengthening the scientific and knowledge-based society as key to sustainable and equitable socioeconomic development. This is being achieved through the deployment of an intercontinental federated e-Infrastructure for scientific usage and the development of a platform of components for federation and application adaptation. Three applications were selected to test this infrastructure and its potential to foster collaboration between the two regions. All of them require the collaboration between Brazil and Europe in the provision of data, services and expertise. One of these applications aims to gain a better understanding of the mutual interaction between climate change and biodiversity dynamics through joint actions aimed at integrating data at a global scale. Current lack of knowledge means that key parameters on the impact of the biodiversity system on the climate system (and reverse) are missing and currently entered as assumptions in climate models. Another approach would require a holistic methodology to identify and unravel patterns and processes in the system-system interactions. This methodological approach assumes the availability of large-scale data sets with observations and measurements of the systems components, together with advanced analytical and modeling software. Climate change research usually involves the processing of large amounts of environmental data spanning huge geographical areas and long periods of time (eg. a few decades). On the other hand, this processing is usually highly parallelizable, since the processing of the data covering a specific area on a particular point in time can be done completely independent of the processing of other areas or the same area in other points in time. In this demonstration we show how we have setup a high-throughput processing service for the execution of SEBAL – a state-of-the-art algorithm to quantify the energy balance using satellite data as an input. We are currently running a data challenge for the processing of 30 years of satellite images covering an area of 1 million square kilometers in the Northeast of Brazil. The processing was started at the UFCG private cloud, and as these shown to be insufficient to cope with the processing load. By means of fogbow– a middleware for the transparent federation of private clouds developed in EUBrazilCC, the processing was automatically and transparently cloudbursted to other cloud providers in the EUBrazilCC federation.

Additional information

The Surface Energy Balance Algorithm for Land (SEBAL) uses the “surface” energy balance to estimate aspects of the hydrological cycle. SEBAL maps evapotranspiration, biomass growth, water deficit and soil moisture.

The basis of SEBAL is the energy balance: the energy driving the hydrological cycle is equal to the incoming energy minus the energy going to heating of the soil and air, and the energy reflected back to space.

SEBAL quantifies the energy balance using satellite data as an input. Land surface characteristics such as surface albedo, leaf area index, vegetation index and surface temperature are derived from satellite imagery. In addition to satellite images, the SEBAL model requires meteorological data, such as wind speed, humidity, solar radiation and air temperature. It uses meteorological data from the moment of the recording of the satellite data to solve the 'instantaneous' energy balance, and uses extrapolation to calculate daily evapotranspiration. Using a time series of satellite and meteorological data, periodic cumulative (e.g. weekly, monthly, yearly) evapotranspiration data can be calculated.

Links, references, publications, etc.


Primary authors

Mr Barros Abmar (Universidade Federal de Campina Grande) Dr Carlos Galvão (Universidade Federal de Campina Grande) Francisco Brasileiro (Universidade Federal de Campina Grande) Mr Giovanni Silva (Universidade Federal de Campina Grande) Dr Iana Rufino (Universidade Federal de Campina Grande) Dr Ignacio Blanquer (UPVLC) Mr John Cunha (Universidade Federal de Campina Grande)

Presentation Materials

There are no materials yet.