Competence centre LifeWatch Ecological Observatories

From EGIWiki
Jump to: navigation, search
EGI-Engage Competence centres: Main page ELIXIR BBMRI MoBrain DARIAH LifeWatch EISCAT_3D EPOS Disaster Mitigation | EGI-Engage Knowledge Commons

LifeWatch CC menu: Introduction Community Meetings Task:Support Task:Ecological Observatories Task:Workflows and VLabs Task:Citizen Science
Lifewatch logo.jpg

Ecological Observatories


Updated 27 Feb 2016

Next related meetings:

EGI-LW CC meeting at EGI Conf. 2016 in Amsterdam on 6-8 April (exact session and details to be confirmed)

Deliverable D6.6 submitted


Ecological Observatories are one of the key components for LifeWatch as data providers. The objective of this mini-project is to integrate the tools required to support data management, data processing and modelling for Ecological Observatories in the framework provided by To achieve this objective, several use cases directly related to on-going LifeWatch projects that require the manipulation of data streams from different Ecological Observatories will be analysed. Ecological Observatories data processes require more and more Big Data techniques: from support to real-time data streams to handling the post-processing of large volumes of diverse data from multiple disciplines: meteorology, geophysics, hydrology, chemistry, social and of course life sciences (biology, ecology, omics). Moreover, in the last years new and powerful software packages are starting to allow the simulation of these complex multidisciplinary systems. This mini project will provide the support to use these techniques in the EGI framework, exploiting the experience being developed in different ecosystems and with different software. At the same time this mini-project will be also useful as an example of the support that can be offered to specific LifeWatch national initiatives, including the design of this support to be effective. The first use case will explore the support to an ongoing experience by LifeWatch Belgium on marine biodiversity research.

This field of research is very dependent on specific data types: species descriptions and identifications, species behaviour, occurrences, presence/absence, biomass, abundance and many others similar. For a long time the collection of this type of data has been mainly a manual process: sampling, sample preparation, identifying species, counting, weighing, typing the data in spreadsheets or databases etc. The use of biosensors and sensor networks for in-situ observation seems to be one of the most promising approaches as this method eliminates the need for taking physical sampling and avoids labour intensive sample preparation processes; moreover the dataflow can be automated and requires less workload from the scientists. Following this promising approach, Flanders Marine Institute (VLIZ) has started the installation of a number of biosensors on board of the Research Vessel Simon Stevin, as part of the Flanders Marine LifeWatch Observatory. This project has a series of needs that require the use of a powerful e-infrastructure able to handle a Big Data problem: about 50Tb of data per year, mainly video and images, will be collected by the vessel in quasi real time requiring a substantial computational power. The project also needs to incorporate a framework based in R for the final researcher.

  • Similar use cases, in particular for analytic tools with simple yet powerful user interfaces, apply to other Ecological Observatories considered: the LTER Observatory of Sierra Nevada in Spain, the Sabor-LTER site in Portugal, a lake (Sanabria lake also in Spain), and another LTER (ZATU) studying areas under chronic radiation exposure on the French territory.
  • The LTER site at the Observatory of Sierra Nevada has implemented an ontological system called Savia that can describe both the ecosystem functioning and the behaviour of abiotic factors in a Natura 2000 habitat. This system is able to automatically download images from MODIS products, create indicators and compute temporal trends for them. All this information is described through an ontology that takes into account the spatio-temporal component of the datasets. All the information is structured according to the ontology, and is uploaded to a knowledge base. Users can formulate complex questions using a SPARQL end-point. This system has been tested and validated in a case study that uses Quercus pyrenaica wild forests as a target habitat. The ecosystem functioning is assessed using NDVI, as the selected abiotic factor is snow cover. Savia provides useful data regarding these two variables and reflects relationships between them, but full exploitation of these data requires additional computing tools and resources.
  • ZATU, a Long Term Ecological Research in Massif Central and Massif Armoricain labelled as one of the 13 French “Zones-Ateliers”, is dedicated to understand how life is impacted by natural ionizing radiation. Its objectives include the analysis of the social impact of natural radioactivity linked to uranium extraction, the migration of radioelements out of storage sites and the impact of radiation on living systems. ZATU will explore how large volumes of diverse data coming from multiple disciplines (chemistry, ecology, molecular biology, physics, social sciences) can be stored and made available using EGI services and infrastructure.
  • In Sabor-LTER, the construction of a dam by the Portuguese Electricity Company EDP is used as an opportunity to set a long-term and large-scale habitat fragmentation “experiment”, whereby the ecological consequences of infrastructure development and its associated biodiversity offsets can be analysed in detail over a range of spatial and temporal scales. The site is set for a wide range of biodiversity offsets and monitoring programs for about 70 years, spanning from the early stages of dam construction through the lifetime of the exploitation concession. The dam will be built in a Natura 2000 site and within a region where several large dams are already in operation or will be built in the near future (FozTua dam), allowing the examination of cumulative impacts of infrastructure development in areas of high natural value.

A second objective in the mini-project will be the support to large simulations using existing software packages, requiring the integration of different data, and producing large data outputs that may require further post-processing. In particular two use cases are proposed: the first the complete simulation of a very relevant process for an ecological system, water eutrophication, using a large simulation suite, Delft3D; the second use case aims to understand the Global Carbon cycle using another large program, the Community Land Model (CLM, see ).

The third objective of the mini-project will be to identify other existing software suites of interest in the Biodiversity and Ecological Systems research, understand its requirements and support its integration in a LifeWatch catalog of services and applications.

Description of work

  1. Task 2.1 (UGr as JRU-LW-ES, VLZ as LW-BE, NGI-FR, CIBIO as LW-PT) Handling Data Streams from Ecological Observatories: Flanders Marine Ship (BE), Mountain Observatory in Sierra Nevada (ES), Life under natural radiation (ZATU, FR), Lakes and Water Reservoirs (Sanabria Lake and CdP Water Reservoir, ES) Task 2.2 (CSIC as JRU-LW-ES) Supporting large software suites for Modelling Ecosystems: Delft3D (on water quality and eutrophication), Community Land Model on Global Carbon. Task 2.3 (CSIC as JRU-LW-ES) Towards an integrated framework/toolbox at international level including a catalogue of applications and final user interfaces based in R and Python.


File:EGI Engage LW-CC-D1.1 v01.pdf

  1. D2.1 Proposal for a data flow handler to support integration of the information from Ecological Observatories. File:EGI Engage LW-CC-D2.1 v01.pdf Type: Prototype+Report. Due: M6 (Aug 2015)
  1. D2.2 Deployment of basic R tools to process data from Ecological Observatories using HTC/HPC infrastructure available in EGI. File:EGI Engage LW-CC-D2.2 v01.pdf Type: Tools+Report Due: M12 (Feb 2016)
  1. D2.3 Support (installation, definition of images and context, connection to HTC/HPC/Data resources) to the execution of simulation packages Delft3D and CLM. Type: Report. Due: M12 (Feb 2016)
  1. D3.4 Report on the applications installed and usage record. Type: Report. Due: M24

(Feb 2017)