Search

The purpose of this document is to advance and define requirements for the eLTER RI to be able to serve validation processes of the major Earth Observation (EO) data providers. The document describethe general requirements for in-situ data and identified current eLTER Standard Observations (SOs) that are mostly needed in the calibration and validation processes of the main EO data providers. The docume givea view to the current validation protocols and tools used in the cal/val processes as well as described a framework to evaluate the maturity of in-situ observations for EO validation. Finally, key steps for the eLTER RI required to develop research infrastructure activities to be able serve these validation processes are identified

We present a modified version of an existing lake ecosystem model,describing a trophic chain generated by nutrients, phytoplankton andzooplankton (NPZ model). The NPZ model takes into account the verticaldynamics of the biomasses of the main species.We tailor the model to specific ecosystems by including seasonalityin the dynamics of the various compartments. Moreover, different speciesexhibit a different behaviour with respect to diffusion and to the rate ofvertical movement. With this model, we simulate the ecosystem dynamicsof Alpine lakes located in study sites of the H2020 ECOPOTENTIALproject.

This poster has been presented at the first ILTER Open Science Meetingin Skukuza, Kruger National Park, South Africa, 9-13October 2016 (https://na.eventscloud.com/ehome/156435), and describes the general purposes and organization of the H2020 project ECOPOTENTIAL (http://www.ecopotential-project.eu/)&nbsp

One of the core European Union environmental policies is the creation and monitoring of the Natura 2000 network of protected areas. This network has been explicitly established for the preservation of conservation priority habitat types and species. Still the concept of habitat is a key concept for ecologists that remains ill defined and is notoriously hard to quantify and measure. Several classification schemes have been put forward, but their relative strengths and weaknesses remain less well examined. In this study we analyzed 8 different Natura 2000 sites (3 Greek, 2 Italian, 2 Portuguese, 1 British). Our study sites reflect a variety of ecosystems, most of them are Mediterranean (7 of the 8) and most of them are wetlands (6 of the 8). In each site, we classified habitats according to 4 different classification schemes (Annex I of the Habitats Directive, Corine Biotopes, EUNIS and General Habitat categories). Also, we used three other widely used land cover classification schemes (namely Corine Land Cover, FAO Land Cover Classification System and IGBP DIS scheme). We found that the different schemes produced considerably different values of landscape diversity leading even to different ranking of the sites according to their diversity. Furthermore, when comparing the landscape composition among sites according to the different schemes, they led to different inferences. Our results imply that the classification scheme used for estimating habitat composition plays an important role for the monitoring of protected areas, perhaps more important than previously assumed.

At a global level, protected sites have been established for the primary purpose of conserving biodiversity, with survey and monitoring of habitats undertaken largely within their boundaries. However, because of increasing human populations with greater access to resources, there is a need to now consider monitoring anthropic activities in the surrounding landscapes as pressures and disturbances are impacting on the functioning and biodiversity values of many protected sites. Earth Observation (EO) data acquired across a range of spatial and temporal scales offer new opportunities for monitoring biodiversity over varying time-scales, either through direct or indirect mapping of species or habitats. However, Land Cover (LC) and/or Land Use (LU), rather than habitat maps are generated in many national and international programs and, whilst the translation from one classification to the other is desirable, differences in definitions and criteria have so far limited the establishment of a unified approach. Focusing on both natural and non-natural environments associated with Natura 2000 sites in the Mediterranean, this paper considers the extent to which three common LC/LU taxonomies (CORINE, the Food and Agricultural Organisation (FAO) Land Cover Classification System (FAO-LCCS) and the IGBP) can be translated to habitat taxonomies with minimum use of additional environmental attributes and/or in situ data. A qualitative and quantitative analysis based on the Jaccard’s index established the FAOLCCS as being the most useful taxonomy for harmonizing LC/LU maps with different legends and dealing with the complexity of habitat description and as a framework for translating EO-derived LC/LU to habitat categories. As demonstration, a habitat map of a wetland site is obtained through translation of the LCCS taxonomy.

The full text can be downloaded at http://journal.embnet.org/index.php/embnetjournal/article/view/205/460

We analyze the generalization ability of a simple perceptron acting on a structured input distribution for the simple case of two clusters of input data and a linearly separable rule. The generalization ability computed for three learning scenarios: maximal stability, Gibbs, and optimal learning, is found to improve with the separation between the clusters, and is bounded from below by the result for the unstructured case, recovered as the separation between clusters vanishes. The asymptotic behavior for large training sets is the same for structured and unstructured input distributions. For small training sets, the generalization error of the maximally stable perceptron exhibits a nonmonotonic dependence on the number of examples for certain values of the model parameters.

We analyse on-line learning of a linearly separable rule with a simple perceptron. Example inputs are taken from two overlapping clusters of data and the rule is defined through a teacher vector which is in general not aligned with the connection line of the cluster centers. We find that the Hebb algorithm cannot learn the rule perfectly in general. Moreover the dependence of the generalization error on the number of examples is nonmonotonic for certain choices of the model parameters. Perceptron and AdaTron training, however, approach perfect generalization with increasing size of the training set, and the asymptotic behavior is the same as for unstructured input data.