History does not move in a straight line. CEPS Commentary, 5 July 2012

Eight years have passed since the EU launched its European Neighbourhood Policy, aimed at inducing its neighbours to the east to converge on modern European values and economic norms. In this Commentary, Michael Emerson reflects on the curious and circuitous turn of events in the region during this period. Michael Emerson is Senior Associate Research Fellow at CEPS.

Avian mortalities due to transmission line collisions: a review of current estimates and field methods with an emphasis on applications to the Canadian electric network

Birds are vulnerable to collisions with human-made fixed structures. Despite ongoing development and increases in infrastructure, we have few estimates of the magnitude of collision mortality. We reviewed the existing literature on avian mortality associated with transmission lines and derived an initial estimate for Canada. Estimating mortality from collisions with power lines is challenging due to the lack of studies, especially from sites within Canada, and due to uncertainty about the magnitude of detection biases. Detection of bird collisions with transmission lines varies due to habitat type, species size, and scavenging rates. In addition, birds can be crippled by the impact and subsequently die, although crippling rates are poorly known and rarely incorporated into estimates. We used existing data to derive a range of estimates of avian mortality associated with collisions with transmission lines in Canada by incorporating detection, scavenging, and crippling biases. There are 231,966 km of transmission lines across Canada, mostly in the boreal forest. Mortality estimates ranged from 1 million to 229.5 million birds per year, depending on the bias corrections applied. We consider our most realistic estimate, taking into account variation in risk across Canada, to range from 2.5 million to 25.6 million birds killed per year. Data from multiple studies across Canada and the northern U.S. indicate that the most vulnerable bird groups are (1) waterfowl, (2) grebes, (3) shorebirds, and (4) cranes, which is consistent with other studies. Populations of several groups that are vulnerable to collisions are increasing across Canada (e.g., waterfowl, raptors), which suggests that collision mortality, at current levels, is not limiting population growth. However, there may be impacts on other declining species, such as shorebirds and some species at risk, including Alberta’s Trumpeter Swans (Cygnus buccinator) and western Canada’s endangered Whooping Cranes (Grus americana). Collisions may be more common during migration, which underscores the need to understand impacts across the annual cycle. We emphasize that these estimates are preliminary, especially considering the absence of Canadian studies.

Wetland restoration within agricultural watersheds: Balancing water quality protection with habitat conservation

Peat wetlands that have been restored from agricultural Land have the potential to act as Long term sources of phosphorus (P) and, therefore have to potenital to accelerate freshwater eutrophication. During a two-year study the water table in a eutrophic fen peat that was managed by pump drainage fluctuated annually between +20 cm and -60 cm relative to ground Level. This precise management was facilitated by the high hydraulic conductivity (K) of the humified peat (1.1 x 10(-5) m s(-1)) below around 60 cm depth. However, during one week of intermittent pumping, as much as 50 g ha(-1) dissolved P entered the pumped ditch. Summer. rainfall events and autumn reflooding also triggered P losses. The P Losses were attributed to the low P sorption capacity (217 mg kg(-1)) of the saturated peat below 60 cm, combined with its high K and the reductive dissolution of Fe bound P.

Mineral and geochemical characterization of a leptic aluandic soil and a thapto aluandic-ferralsol developed on trachytes in Mount Bambouto (Cameroon volcanic line)

Mineral and geochemical investigations were carried out on soil samples and fresh rock (trachytes) from two selected soil profiles (TM profile on leptic aluandic soils and TL profile on thapto aluandic-ferralsols) from Mount Bambouto to better understand geochemical processes and mineral paragenesis involved in the development of soils in this environment. In TM profile, the hydrated halloysites and goethite occur in the weathered saprolite boulders of BC horizon while dehydrated halloysite, gibbsite and goethite dominate the soils matrices of BC and A horizons. In TL profile, the dehydrated halloysites and goethite are the most abundant secondary minerals in the weathered saprolites of C and BC horizons while gibbsite, hematite and kaolinite occur in the soil matrices of BC, B and A horizons. The highest gibbsite content is in the platy nodules of B horizon. In both soil profiles, organo-metal complexes (most likely of AI and Fe) are present in the surface A horizon. Geochemically, between the fresh rock and the weathered saprolites in both soils, SiO2, K2O, CaO, Na2O and MgO contents decrease strongly while Fe2O3 and Al2O3 tend to accumulate. The molar ratio of SiO2/Al2O3 (Ki) and the sum of Ca, Mg, K and Na ions (TRB) also decreases abruptly between fresh rocks and the weathered saprolites, but increases significantly at the soil surface. The TM profile shows intense Al enrichment whereas the TL profile highlights enrichment in both AI and Fe as the weathering progresses upwards. Both soil profiles are enriched in Ni, Cu, Ba and Co and depleted in U, Th, Ta, Hf, Y, Sr, Pb, Zr and Zn relative to fresh rock. They also show a relatively low fractionation of the rare earth elements (REE: La, Nd, Sm, Eu, Tb, Yb and Lu), except for Ce which tends to be enriched in soils compared to CI chondrite. All these results give evidence of intense hydrolysis at soil deep in Mount Bambouto resulting in the formation of halloysite which progressively transforms into gibbsite and/or dehydrated halloysite. At the soil surface, the prominent pedogenetic process refers to andosolization with formation of organo-metal complexes. In TL profile, the presence of kaolinite in soil matrices BC and B horizons is consistent with ferralitization at soil deep. In conclusion, soil forming processes in Mount Bambouto are strongly influenced by local climate: (i) in the upper mountain (>2000 m), the fresh, misty and humid climate favors andosolization; whereas (ii) in the middle lands (1700-2000 m) with a relatively dry climate, both andosolization at the soil surface and ferralitization at soil deep act together. (C) 2009 Elsevier B.V. All rights reserved.

Dynamic load balancing for the distributed mining of molecular structures

In molecular biology, it is often desirable to find common properties in large numbers of drug candidates. One family of methods stems from the data mining community, where algorithms to find frequent graphs have received increasing attention over the past years. However, the computational complexity of the underlying problem and the large amount of data to be explored essentially render sequential algorithms useless. In this paper, we present a distributed approach to the frequent subgraph mining problem to discover interesting patterns in molecular compounds. This problem is characterized by a highly irregular search tree, whereby no reliable workload prediction is available. We describe the three main aspects of the proposed distributed algorithm, namely, a dynamic partitioning of the search space, a distribution process based on a peer-to-peer communication framework, and a novel receiverinitiated load balancing algorithm. The effectiveness of the distributed method has been evaluated on the well-known National Cancer Institute’s HIV-screening data set, where we were able to show close-to linear speedup in a network of workstations. The proposed approach also allows for dynamic resource aggregation in a non dedicated computational environment. These features make it suitable for large-scale, multi-domain, heterogeneous environments, such as computational grids.

Decentralized load balancing for highly irregular search problems

In this paper, we present a distributed computing framework for problems characterized by a highly irregular search tree, whereby no reliable workload prediction is available. The framework is based on a peer-to-peer computing environment and dynamic load balancing. The system allows for dynamic resource aggregation, does not depend on any specific meta-computing middleware and is suitable for large-scale, multi-domain, heterogeneous environments, such as computational Grids. Dynamic load balancing policies based on global statistics are known to provide optimal load balancing performance, while randomized techniques provide high scalability. The proposed method combines both advantages and adopts distributed job-pools and a randomized polling technique. The framework has been successfully adopted in a parallel search algorithm for subgraph mining and evaluated on a molecular compounds dataset. The parallel application has shown good calability and close-to linear speedup in a distributed network of workstations.

DTM generation from LIDAR data using skewness balancing

LIght Detection And Ranging (LIDAR) data for terrain and land surveying has contributed to many environmental, engineering and civil applications. However, the analysis of Digital Surface Models (DSMs) from complex LIDAR data is still challenging. Commonly, the first task to investigate LIDAR data point clouds is to separate ground and object points as a preparatory step for further object classification. In this paper, the authors present a novel unsupervised segmentation algorithm-skewness balancing to separate object and ground points efficiently from high resolution LIDAR point clouds by exploiting statistical moments. The results presented in this paper have shown its robustness and its potential for commercial applications.

Assessment of the consistency of near-infrared water vapor line intensities using high-spectral-resolution ground-based Fourier transform measurements of solar radiation

Calculations of the absorption of solar radiation by atmospheric gases, and water vapor in particular, are dependent on the quality of databases of spectral line parameters. There has been increasing scrutiny of databases such as HITRAN in recent years, but this has mostly been performed on a band-by-band basis. We report nine high-spectral-resolution (0.03 cm(-1)) measurements of the solar radiation reaching the surface in southern England over the wave number range 2000 to 12,500 cm(-1) (0.8 to 5 mm) that allow a unique assessment of the consistency of the spectral line databases over this entire spectral region. The data are assessed in terms of the modeled water vapor column that is required to bring calculations and observations into agreement; for an entirely consistent database, this water vapor column should be constant with frequency. For the HITRAN01 database, the spread in water vapor column is about 11%, with distinct shifts between different spectral regions. The HITRAN04 database is in significantly better agreement (about 5% spread) in the completely updated 3000 to 8000 cm(-1) spectral region, but inconsistencies between individual spectral regions remain: for example, in the 8000 to 9500 cm(-1) spectral region, the results indicate an 18% (+/- 1%) underestimate in line intensities with respect to the 3000 to 8000 cm(-1) region. These measurements also indicate the impact of isotopic fractionation of water vapor in the 2500 to 2900 cm(-1) range, where HDO lines dominate over the lines of the most abundant isotope of H2O.

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.