Interacting effects of change in climate, human population, land use, and water use on biodiversity and ecosystem services

Human population growth and resource use, mediated by changes in climate, land use, and water use, increasingly impact biodiversity and ecosystem services provision. However, impacts of these drivers on biodiversity and ecosystem services are rarely analyzed simultaneously and remain largely unknown. An emerging question is how science can improve the understanding of change in biodiversity and ecosystem service delivery and of potential feedback mechanisms of adaptive governance. We analyzed past and future change in drivers in south-central Sweden. We used the analysis to identify main research challenges and outline important research tasks. Since the 19th century, our study area has experienced substantial and interlinked changes; a 1.6°C temperature increase, rapid population growth, urbanization, and massive changes in land use and water use. Considerable future changes are also projected until the mid-21st century. However, little is known about the impacts on biodiversity and ecosystem services so far, and this in turn hampers future projections of such effects. Therefore, we urge scientists to explore interdisciplinary approaches designed to investigate change in multiple drivers, underlying mechanisms, and interactions over time, including assessment and analysis of matching-scale data from several disciplines. Such a perspective is needed for science to contribute to adaptive governance by constantly improving the understanding of linked change complexities and their impacts.

Dynamic load balancing in parallel KD-tree k-means

One among the most influential and popular data mining methods is the k-Means algorithm for cluster analysis. Techniques for improving the efficiency of k-Means have been largely explored in two main directions. The amount of computation can be significantly reduced by adopting geometrical constraints and an efficient data structure, notably a multidimensional binary search tree (KD-Tree). These techniques allow to reduce the number of distance computations the algorithm performs at each iteration. A second direction is parallel processing, where data and computation loads are distributed over many processing nodes. However, little work has been done to provide a parallel formulation of the efficient sequential techniques based on KD-Trees. Such approaches are expected to have an irregular distribution of computation load and can suffer from load imbalance. This issue has so far limited the adoption of these efficient k-Means variants in parallel computing environments. In this work, we provide a parallel formulation of the KD-Tree based k-Means algorithm for distributed memory systems and address its load balancing issue. Three solutions have been developed and tested. Two approaches are based on a static partitioning of the data set and a third solution incorporates a dynamic load balancing policy.