5. A Scientific Gateway for integrated data analysis and research on biodiversity and climate change

Author(s): 
Anders, Rufino, Seijmonsbergen, Cunha, Oliveira Galvao, Los, Canhos, Lezzi, Fiore, Aloisio, Brasileiro, Blanquer
Focus area: 
cross-border collaboration through e-infrastructure services; integrated data analysis; climate change; biodiversity, LULC change indicators.
Who stands to benefit and how: 
Earth observation and spatio-temporal analysis techniques are increasingly being made available to the public domain. We combine different data sources and analysis tools into a single, user-friendly, Scientific Gateway supporting multi-disciplinary research on biodiversity and climate change research. Professionals (practitioners, decision-makers, educators and scientists) working in governmental and commercial agencies can benefit from information that is analyzed across various temporal and spatial scales to reconstruct the past, monitor the present and predict future landscape responses, particularly in the domains of hydrology, agriculture, forestry, earth science, biology and environmental conservation.
Focus of your position paper: 
The Earth is a dynamic system which acts upon complex interactions of the atmosphere, biosphere, and soil. In order to advance our knowledge of these dynamics, and to predict what effects natural processes and human activity will have on biodiversity, ecosystem services, and livelihood, there is a need for detailed understanding of the interdependencies of such environmental systems. The increase in data availability, wealth of analysis tools, and growing computational capacity offer unique opportunities for multi-disciplinary research to tackle key questions raised in the climate change debate. In this context the cross-continental collaboration EUBrazil Cloud Connect integrates a wide variety of data sources and analysis tools into a user-friendly and web-based Scientific Gateway. This Scientific Gateway is backed by the PDAS (Parallel Data Analytics Service), which provides a framework for parallel I/O and data analysis, storage and distribution of large scientific datasets. The framework has been extended with specific data analytics operators and primitives to fully meet the use case requirements. A set of VMIs has been also prepared to deploy the PDAS in private cloud environments. The PDAS is fully interoperable with the EGI Infrastructure as well as with OGC-based environments. The Scientific Gateway provides access to various sources of data, including multi-spectral satellite imagery (Landsat), laser altimetry data (LiDAR) and plant distribution occurrences. Landsat provides global imagery (30x30m grid cells) at frequent time intervals with information of spectral reflectance in visible and infrared wavelengths. The Surface Energy Balance Algorithm for Land (SEBAL) algorithm has been implemented to automatically translate spectral responses into temporal proxies, such as surface albedo, land surface heat fluxes and vegetation products such as NDVI, EVI and leaf area index. In addition, high-resolution laser altimetry data (LiDAR) and analysis tools have been implemented to derive detailed terrain metrics (e.g. terrain elevation, slope angle, aspect) and forest metrics (e.g. canopy height, aboveground biomass, vertical biomass distribution). Moreover, vegetation occurrences provide input for Species Distribution Models (SDMs) and simulations of vegetation response to changes of biodiversity, land use and climate parameters. Experiments have been carried out to demonstrate the strong potential of the Scientific Gateway. SEBAL data provides indicators of Land Use/Cover (LULC) trends. Additionally, SEBAL output data can be input data for hydrologic models such as evapotranspiration time series. To provide such data, a method is required that combines global and local remote sensing datasets that should be calibrated with field observations and existing climate measurements. Efficient computational capacity to run the complex workflows on huge data sets is required. OpenModeller provides observations about species data occurrences in the targeted areas of Brazil. In this regard, species occurrences provide an indication about the presence of some species in a specific area. Laser altimetry data analysis is implemented for the Adolpho Ducke Forest Reserve, a 10,000 ha protected area near the city of Manaus, Brazil, representing an undisturbed tropical rainforest. We demonstrate that through point filtering techniques 3D forest structure is visualized at the scale of individual tree crowns and branches, showing the complexity of multi-layered rainforest. This forms the basis for detailed change analysis and monitoring, particularly when combined with other climate indicators and SDMs. In summary, the EUBrazil Cloud Connect Scientific Gateway provides users access to massive datasets, processing algorithms and visualization tools. It is designed to integrate these components for analyzing the multi-disciplinary nature of earth science systems, particularly dynamics related to biodiversity and climate change. In the near future, larger, more frequent and more detailed data will become available, which can be directly implemented in the current infrastructure for near-real-time data processing and visualization. This will increase public awareness of land degradation or improvement, and also aids practitioners and decision makers to respond quickly to changing environmental conditions which, in turn, increases the sustainability of land management.