Forecasting daily pollen concentrations using data-driven modeling methods in Thessaloniki,Greece

Airborne pollen have been associated with allergic symptoms in sensitized individuals, having a direct impact on the overall quality of life of a considerable fraction of the population. Therefore, forecasting elevated airborne pollen concentrations and communicating this piece of information to the public are key issues in prophylaxis and safeguarding the quality of life of the overall population. In this study, we adopt a data-oriented approach in order to develop operational forecasting models (1–7 days ahead) of daily average airborne pollen concentrations of the highly allergenic taxa: Poaceae, Oleaceae and Urticaceae. The models are developed using a representative dataset consisting of pollen and meteorological time-series recorded during the years 1987–2002, in the city of Thessaloniki, Greece. The input variables (features) of the models have been optimized by making use of genetic algorithms, whereas we evaluate the performance of three algorithms: i) multi-Layer Perceptron, ii) support vector regression and iii) regression trees originating from distinct domains of Computational Intelligence (CI), and compare the resulting models with traditional multiple linear regression models. Results show the superiority of CI methods, especially when forecasting several days ahead, compared to traditional multiple linear regression models. Furthermore, the CI models complement each other, resulting to a combined model that performs better than each one separately. The overall performance ranges, in terms of the index of agreement, from 0.85 to 0.93 clearly suggesting the potential operational use of the models. The latter ones can be utilized in provision of personalized and on-time information services, which can improve quality of life of sensitized citizens.     

From economic to environmental sustainability: the forest management debate in 20th century Finland and Sweden

The issues of sustainability, efficiency, and profitability in private forestry dominated debate on Finnish and Swedish forest policy throughout the 20th century. The management and utilization of private forests have been significant for these countries since more than half of productive forest land has been owned by families. Forests have also played an important role in Finland’s and Sweden’s national economies. There was however no consensus among forest owners and forestry experts—forest scientists and forestry authorities—about the methods of achieving forestry sustainability and profitability, particularly in the early 1900s. This article focuses on the types of argument on which perceptions of good forest management were based, and the reasons for their use. The article also explores how disagreements and occasional open conflicts between private forest owners and forest authorities were solved, and what kind of regulative means were developed to avoid such differences. There are three main reasons for making a case study of Finland and Sweden: Firstly, the forest sector has played an important role in their national economies, secondly, the objectives and content of their forestry laws are quite similar, and thirdly, the tradition of family forest ownership has remained strong. The debate on good forest management in this period revolves around the two silvicultural paradigms: even-aged and uneven-aged systems. The former concerns formal, scientific knowledge of forest management, and the latter takes a traditional approach.     

Production and sedimentation of peptide toxins nodularin-R and microcystin-LR in the northern Baltic Sea

This seven-year survey was primarily targeted to quantification of production of nodularin-R (NOD-R), a cyclic pentapeptide hepatotoxin, in Baltic Sea cyanobacteria waterblooms. Additionally, NOD-R and microcystin-LR (MC-LR; a cyclic heptapeptide toxin) sedimentation rates and NOD-R sediment storage were estimated. NOD-R production (70-2450 μg m⁻³; ∼1 kg km⁻² per season) and sedimentation rates (particles; 0.03-5.7 μg m⁻² d⁻¹; ∼0.3 kg km⁻² per season) were highly variable over space and time. Cell numbers of Nodularia spumigena did not correlate with NOD-R quantities. Dissolved NOD-R comprised 57-100% of total NOD-R in the predominantly senescent, low-intensity phytoplankton blooms and seston. Unprecedentedly intensive MC-LR sedimentation (0.56 μg m⁻² d⁻¹) occurred in 2004. Hepatotoxin sedimentation rates highly exceeded those of anthropogenic xenobiotics. NOD-R storage in surficial sediments was 0.4-20 μg kg⁻¹ (∼0.1 kg km⁻²). Loss of NOD-R within the chain consisting of phytoplankton, seston and soft sediments seemed very effective.     

Quantification of source-level turbulence during LNG spills onto a water pond

The use of computational fluid dynamics (CFD) models to simulate LNG vapor dispersion scenarios has been growing steadily over the last few years, with applications to LNG spills on land as well as on water. Before a CFD model may be used to predict the vapor dispersion hazard distances for a hypothetical LNG spill scenario, it is necessary for the model to be validated with respect to relevant experimental data. As part of a joint-industry project aimed at validating the CFD methodology, the LNG vapor source term, including the turbulence level associated with the evaporation process vapors was quantified for one of the Falcon tests.This paper presents the method that was used to quantify the turbulent intensity of evaporating LNG, by analyzing the video images of one of the Falcon tests, which involved LNG spills onto a water pond. The measured rate of LNG pool growth and spreading and the quantified turbulence intensity that were obtained from the image analysis were used as the LNG vapor source term in the CFD model to simulate the Falcon-1 LNG spill test. Several CFD simulations were performed, using a vaporization flux of 0.127 kg/m² s, radial and outward spreading velocities of 1.53 and 0.55 m/s respectively, and a range of turbulence kinetic energy values between 2.9 and 28.8 m²/s². The resulting growth and spread of the vapor cloud within the impounded area and outside of it were found to match the observed behavior and the experimental measured data.The results of the analysis presented in this paper demonstrate that a detailed and accurate definition of the LNG vapor source term is critical in order for any vapor cloud dispersion simulation to provide useful and reliable results.     

Heavy metal pollution disturbs immune response in wild ant populations

Concern about the effects of environmental contaminants on immune function in both humans and wildlife is growing and practically nothing is known about this impact on terrestrial invertebrates, even though they are known to easily accumulate pollutants. We studied the effect of industrial heavy metal contamination on immune defense of a free-living wood ant (Formica aquilonia). To find out whether ants show an adapted immune function in a polluted environment, we compared encapsulation responses between local and translocated colonies. Local colonies showed higher heavy metal levels than the translocated ones but the encapsulation response was similar between the two groups, indicating that the immune system of local ants has not adapted to high contamination level. The encapsulation response was elevated in moderate whereas suppressed in high heavy metal levels suggesting higher risk for infections in heavily polluted areas.     

Role of sea-ice biota in nutrient and organic material cycles in the northern Baltic Sea

This paper compiles biological and chemical sea-ice data from three areas of the Baltic Sea: the Bothnian Bay (Hailuoto, Finland), the Bothnian Sea (Norrby, Sweden), and the Gulf of Finland (Tv?rminne, Finland). The data consist mainly of field measurements and experiments conducted during the BIREME project from 2003 to 2006, supplemented with relevant published data. Our main focus was to analyze whether the biological activity in Baltic Sea sea-ice shows clear regional variability. Sea-ice in the Bothnian Bay has low chlorophyll a concentrations, and the bacterial turnover rates are low. However, we have sampled mainly land-fast level first-year sea-ice and apparently missed the most active biological system, which may reside in deformed ice (such as ice ridges). Our limited data set shows high concentrations of algae in keel blocks and keel block interstitial water under the consolidated layer of the pressure ridges in the northernmost part of the Baltic Sea. In land-fast level sea-ice in the Bothnian Sea and the Gulf of Finland, the lowermost layer appears to be the center of biological activity, though elevated biomasses can also be found occasionally in the top and interior parts of the ice. Ice algae are light limited during periods of snow cover, and phosphate is generally the limiting nutrient for ice bottom algae. Bacterial growth is evidently controlled by the production of labile dissolved organic matter by algae because low growth rates were recorded in the Bothnian Bay with high concentrations of allochthonous dissolved organic matter. Bacterial communities in the Bothnian Sea and the Gulf of Finland show high turnover rates, and activities comparable with those of open water communities during plankton blooms, which implies that sea-ice bacterial communities have high capacity to process matter during the winter period.     

Modeling the vapor source term associated with the spill of LNG into a sump or impoundment area

The recent publication of evaluation protocols for vapor source term models and vapor dispersion models have influenced the modeling approaches that can be used for approval of new and expansion projects at LNG receiving terminals. In the past few years the scientific basis of integral vapor source term models has been questioned with growing concerns regarding their validity. In this paper, the shallow water equations (SWEs) were solved to study the characteristics of the evaporating LNG pool associated with a constant flow rate spill of LNG into a concrete sump. In the early stages of pool spreading, the leading edge thickness profile of the SWE model scales with the square root of the distance from the leading edge as the pool spreads. After the edge of the pool reaches the wall, the reflected wave forms a hydraulic jump that travels back towards the center of the pool at a speed that is considerably slower than the initial spreading of the pool. Once the hydraulic jump reaches the center, the pool assumes a nearly flat free surface for the rest of the spill. The pool spreading and the rate of evaporation from the SWEs were then compared to the solution provided by the integral model, PHAST. The two approaches were found to agree well with one another. The SWE model was also used to demonstrate the influence of an elevated spill source. With an elevated source, the LNG pool spreads faster, significantly increasing the initial rate of vaporization and peak vaporization rate. This increase in the initial rate of vaporization could lead to an increase in the vapor cloud hazard distance. The SWE model was also used to demonstrate the influence of an inclined sump floor in the shape of an inverted cone where the spilling LNG accumulates in the low vertex of the cone. Inclined sump floors can be used to significantly reduce the cumulative evaporation, making them attractive as a possible mitigation approach in cases where a containment sump is located close to a property boundary.     

A procedure for generating synthetic weather records in conjunction of climatic scenario for modelling of ecological impacts of changing climate in boreal conditions

A simulation procedure for the calculation of temperature, cloudiness, radiation, precipitation, air humidity, windiness and atmospheric carbon is presented. The procedure generates the hourly or daily values of the weather factors based on long-term weather statistics. The basic factor behind the weather pattern indicated by different weather factors is air temperature, which modifies the cloudiness produced by a stochastic process in other respects. Consequently, the radiation and precipitation coming onto a site and the air humidity are also modified by temperature. This facilitates the future weather patterns to be calculated on the basis of the temperature increase allowing to introduce the effect of the suggested climatic change also into the other weather factors, assuming that the basic variability of the weather factors remains unchanged. The future windiness is calculated as random process without correlation to other weather factors.     

The process of burnout in white‐collar and blue‐collar jobs: eight‐year prospective study of exhaustion

This study tested three contrasting theories about the sequential process of the three dimensions of burnout (i.e., exhaustion, cynicism and professional efficacy) as measured by the Maslach Burnout Inventory—General Survey (MBI‐GS). The causal order of the burnout components was investigated by including job stressors as antecedents of burnout in structural equation models (LISREL). The predictive effect of exhaustion on burnout dimensions eight years later was also investigated. The longitudinal models were then tested in two occupational subgroups. The participants were 713 Finnish employees (415 white‐collar and 298 blue‐collar workers) from an international industry enterprise. The best fitting model of the associations between the three burnout dimensions was obtained by a path model where exhaustion predicted cynicism, and cynicism in turn predicted lack of professional efficacy. The symptoms of exhaustion were persistent over time. There were no differences between the two occupational groups in the process of burnout. Also the work‐related antecedents of burnout (job stressors) were very similar in both groups. Copyright © 2002 John Wiley & Sons, Ltd.