After decades of investigation, we still have to deal withthe problem of environmental effects of the use of de-icingsalt on roads. Lacking useful indicators of the system, westill do not know the environmental utility of ameliorativemeasures taken. The thesis aims at i) describing the system of de-icing practices and theirenvironmental effects with special reference to the exposure ofthe roadside environment to air-borne salt and damage to Norwayspruce seedlings and ii) at proposing indicators for afollow-up system. By collecting bulk deposition and relatingthe deposition pattern to factors concerning wind androad-maintenance characteristics, the influence of thesefactors on air-borne exposure is investigated. By exposingNorway spruce (Picea abies(L.) Karst.) seedlings to roadsideconditions the influence ofsalt exposure on the degree ofdamage is investigated. Based on a comparison of severalsystems of monitoring, improvements are suggested by proposingnew indicators for salt use and its environmental effects. A large part of the salt that is applied on the road surfacewill be transported by air and deposited on the ground in theroadside. While the vast majority of the salt will be depositedwithin some tens of metres of the road, some may still be windtransported several hundred metres away. The wind directionplays an important role for the deposition already at adistance of some ten metres from the road. Wind speedinfluences the distance to which the salt is transported.Chloride concentration in unwashed needles collected after thesalting season is positively related to the bulk deposition ofCl during the exposure. The degree of damage can be describedby a sigmoid curve of response to the Cl concentration inneedles. Improvements to the follow-up system are suggested byproposing indicators of the driving forces, pressures, states,impacts and possible responses as regards the undesired impactson water resources, vegetation and the roadside scenery as asocietal asset. <b>Keywords:</b>road, de-icing salt, air-borne, environmentalimpact, vegetation damage, Norway spruce, seedling, follow-up,monitoring, indicator

<p>After decades of investigation, we still have to deal withthe problem of environmental effects of the use of de-icingsalt on roads. Lacking useful indicators of the system, westill do not know the environmental utility of ameliorativemeasures taken. The thesis aims at</p><p>i) describing the system of de-icing practices and theirenvironmental effects with special reference to the exposure ofthe roadside environment to air-borne salt and damage to Norwayspruce seedlings and ii) at proposing indicators for afollow-up system. By collecting bulk deposition and relatingthe deposition pattern to factors concerning wind androad-maintenance characteristics, the influence of thesefactors on air-borne exposure is investigated. By exposingNorway spruce (<i>Picea abies</i>(L.) Karst.) seedlings to roadsideconditions the influence ofsalt exposure on the degree ofdamage is investigated. Based on a comparison of severalsystems of monitoring, improvements are suggested by proposingnew indicators for salt use and its environmental effects.</p><p>A large part of the salt that is applied on the road surfacewill be transported by air and deposited on the ground in theroadside. While the vast majority of the salt will be depositedwithin some tens of metres of the road, some may still be windtransported several hundred metres away. The wind directionplays an important role for the deposition already at adistance of some ten metres from the road. Wind speedinfluences the distance to which the salt is transported.Chloride concentration in unwashed needles collected after thesalting season is positively related to the bulk deposition ofCl during the exposure. The degree of damage can be describedby a sigmoid curve of response to the Cl concentration inneedles. Improvements to the follow-up system are suggested byproposing indicators of the driving forces, pressures, states,impacts and possible responses as regards the undesired impactson water resources, vegetation and the roadside scenery as asocietal asset.</p><p><b>Keywords:</b>road, de-icing salt, air-borne, environmentalimpact, vegetation damage, Norway spruce, seedling, follow-up,monitoring, indicator</p>

Winter road maintenance activities are crucial for maintaining the accessibility and traffic safety of the road network during winters. Common winter road maintenance activities include plowing and the use of de-icing agents (e.g. NaCl) to avoid freezing. Effective winter road maintenance strives towards keeping the roads free from snow and ice while reducing negative side effects of winter road maintenance, such as ground water contamination from road salt. Since the weather is decisive for when there is an increased risk of slipperiness, the understanding and continuous observation and forecasts of the road weather are of highest importance. Sensors are commonly installed along roads to measure road weather conditions to support the road maintenance personnel in taking appropriate actions. Different types of errors and uncertainties related to sensors used for frost warnings along roads have been investigated by using a regional scale dataset from south-western Sweden. The results from this study indicate that various types of uncertainties originate from both measurements and models which have a significant impact on the winter road maintenance efficiency. To provide better information about the road surface conditions, a new method for detecting ice formation on roads is presented. Infrared sensors were used to detect temperature patterns which may occur when ice formation take place on a road surface. The investigations demonstrate the potential to improve winter road maintenance by introducing new methods to better describe the road surface conditions. / QC 20101206

In a previous project, mobile exposures in road environments have been performed in various areas of the world. A number of materials and coatings have been attached under trailers which travels long distances in different types of road environments. One of the materials, zinc, demonstrated various corrosion rates in different parts of the world. In this investigation two various accelerated corrosion tests have been performed at the laboratory of Swerea Kimab in order to realize parameter influence on zinc. Even though the corrosion rates are known, the relation to formed corrosion products has not been investigated earlier. In the present study, corrosion products on zinc have been analyzed using XRD and FTIR. The results from analyses of formed corrosion products have been evaluated together with measured corrosion rates, both from the mobile exposure and the accelerated tests, in order to try to understand under what conditions certain corrosion products are formed and how it affects the corrosion rate. A cold humid climate was found to be the most corrosive. In these environments, de ‐icing salts are used which promotes formation of Simonkolleite and in a subsequent step; Gordaite (if there is a not negligible deposition rate of SO2/SO42‐) which is often observed as the end product in marine environments.

Roads and traffic are a major non-point source of pollutants and may have severe impacts on surface water, groundwater, soil and vegetation. In cold climates, de-icing salt is one such pollutant that may cause increased chloride concentrations and induce other effects on the environment. Monitoring and quantifying environmental effects are crucial for governing decisions towards more suitable use of de-icing salt in order to achieve and maintain good environmental status around roads. This thesis presents an operational modelling tool for monitoring the transport and fate of de-icing salt in the roadside environment in order to quantify changes in the environment at various spatial and temporal scales, using salt application data, meteorological data, geology and generic descriptions of hydrogeological environments as main inputs. A combination of modelling and various independent field measurements provided an efficient means for evaluating and describing the spread of de-icing salt from the road to the surroundings, the deposition of salt and ploughed snow in the roadside, and the corresponding increase in chloride concentration in soil and groundwater. Both the spatial and seasonal variation in soil chloride concentration were significantly affected by de-icing salt application. The importance of type of soil, vegetation type, groundwater conditions and distance from the road was clearly demonstrated for modelling the transport and fate of de-icing salt in the roadside environment. Salt emissions from the road by surface runoff were estimated at 50-80% of applied salt and transport by snow ploughing and air emissions at 20-50%. The uncertainty in the spatial distribution of snow and salt deposition close to the road was high and a previous proposed exponential decline in salt deposition with distance from the road could not be justified within a couple of metres from the road. Future monitoring should include both modelling and systematic data collection in order to reduce the uncertainty in predictions of the environmental impact of de-icing salt. Modelling of chloride concentration, soil water content and soil temperature and measurements of electrical resistivity may be a cost-effective solution for quantifying changes in the roadside environment. / QC 20100526

Growth performance of Bur oak (Q. macrocarpa Michx.) and Red oak (Q. rubra L.) under salinity conditions was assessed by growing seedlings in the presence of increasing levels of NaCl. Salinity reduced root growth in both species, although its repressive effect was more pronounced in Red oak. Exposure to 75 mM NaCl for three weeks almost arrested root growth in Red oak, while it reduced it only by 40 % in Bur oak. Red oak roots showed extensive necrosis and limited branching. Salinity also induced leaf injury, which at a NaCl level of 25 mM was less severe in Bur oak possibly due the higher expression of dehydroascorbate reductase (DHAR) and ascorbate peroxidase (APX), enzymes participating in the detoxification of reactive oxygen species (ROS). Salinity also altered nutrient uptake and accumulation in root and leaf tissue. Compared to Red oak, the relative calcium level in Bur oak roots exposed to increased salinity remained elevated, while an opposite trend was observed in leaf tissue. This was in contrast to nitrogen and potassium, the relative level of which was higher in Red oak leaves grown in the presence of NaCl. The better performance of Bur oak root tissue under salinity conditions was ascribed to structural modifications of the root system with maturation of casparian bands and suberinization occurring closer to the root tip. These structures are known to act as barriers enhancing ion selectivity. Collectively this study demonstrates that relative to Red oak, Bur oak is more tolerant to NaCl induced salinity conditions. / February 2017

Mémoire numérisé par la Division de la gestion de documents et des archives de l'Université de Montréal.

Mémoire numérisé par la Division de la gestion de documents et des archives de l'Université de Montréal

Les systèmes de biorétention sont de plus en plus utilisés pour gérer le ruissellement des eaux pluviales urbaines. Bien que les plantes soient une composante essentielle des biorétentions, il existe à ce jour peu de preuves de la contribution des espèces choisies, en particulier dans les climats tempérés à grande variation saisonnière. Aussi, les bactéries et les champignons peuvent jouer un rôle majeur dans la performance des biorétentions en améliorant la qualité de l'eau ou en soutenant l'efficacité des plantes. Mais à ce jour, on sait peu de choses sur les microorganismes peuplant les biorétentions et encore moins sur l'influence des choix de conception de ce système et ses conséquences sur la contribution de ces organismes. D’autre part, en climat froid, l'utilisation de sels de déglaçage génère des ruissellements salins qui pourraient affecter les performances des biorétentions, notamment en impactant les processus biologiques en cours dans le système. Le but de mon étude était de comparer la contribution de la végétation à la performance des biorétentions en fonction de l'espèce plantée, de tester l'impact du ruissellement salin sur l'hydrologie et la filtration des contaminants par le système en fonction des espèces végétales utilisées, et de caractériser l'influence de ces 2 facteurs sur la diversité, la composition et l'abondance relative des bactéries et champignons au sein des biorétentions. J’ai réalisé une expérience en mésocosmes avec 4 espèces végétales couramment utilisées en biorétention et couvrant un large éventail de types biologiques (Cornus sericea, Juncus effusus, Iris versicolor, Sesleria autumnalis). Pour simuler un ruissellement printanier chargé en sel de déglaçage, l'eau de ruissellement semi-synthétique utilisée pour l'irrigation à était complétée au printemps avec quatre concentrations de NaCl (0, 250, 1000 ou 4000 mg Cl/L). Le séquençage des gènes 16S et ITS ont été utilisé comme indicateur de diversité, de composition et d'abondance relative bactérienne et fongique. En général, tous les mésocosmes de biorétention ont significativement réduit les volumes d'eau, les débits de pointe et les charges de contaminants. Certaines plantes améliorent significativement les performances en augmentant la perte d'eau par évapotranspiration pendant la période de croissance (jusqu'à 2,5 fois) et en réduisant le débit d'eau (jusqu'à 2,7 fois). Les plantes ont amélioré l’enlèvement des macronutriments, avec un enlèvement massique moyen de 55 % pour l’Azote total, 81 % pour le Phosphore total et 61 % pour le K, contre -6 % (relargage), 61 % et 22 % respectivement pour les non plantés. À l'exception des Sesleria, l’enlèvement des éléments traces par les mésocosmes plantés était généralement plus élevée que chez les non plantés (jusqu'à 8,7 %). Leur niveau de contribution suivait le même ordre que leur taux d'évapotranspiration et leur taille globale (Cornus > Juncus > Iris > Sesleria). Même à la concentration la plus élevée de NaCl testée, aucun effet sur la réduction du volume d'eau et les débits n'a été détecté. En revanche, le ruissellement chargé en sel a temporairement augmenté l’enlèvement de certains métaux tels que Cr, Ni, Pb et Zn. Dans l’ensemble, les plantes ont très bien toléré le passage du ruissellement chargé de sel au printemps. Tous les mésocosmes ont naturellement été colonisés par des bactéries et des champignons adaptés à un environnement humide et contaminé. Parmi les taxons dominants, plusieurs ont des fonctions en lien avec la performance des biorétentions, telles que l'implication dans le cycle de l'azote, la dégradation des hydrocarbures, la tolérance et la remédiation de métaux, ou des symbiotes végétaux. Ma thèse souligne l'importance du choix des espèces végétales sur la performance des biorétentions. D’une part en raison de leur contribution différentielle à la réduction du volume d'eau, des débits de pointe et l'enlèvement des contaminants, en particulier des macronutriments, et d’autre part à cause de leur influence sur la composition et l’abondance des microorganismes. Même si plusieurs des bactéries et champignons dominants ce système auraient la capacité de contribuer à la performance des biorétentions, d’autres études devront être menées pour vérifier leur activité. Finalement, ma thèse a démontré que la présence de sel de déverglaçage dans le ruissellement ne devrait pas être un frein à la mise en place de cellules de biorétention, puisque qu'aux concentrations habituellement observées à la suite d’épandages, aucun effet négatif sur la performance, la végétation ou les communautés bactérienne et fongique n’a été observé dans la présente thèse. Cependant, la libération des cations structurants du sol et l’exfiltration du sels devrait être faire l’objet d’un suivi à plus long terme. Bien que ces expériences aient permis d'analyser l'effet des espèces végétales et des sels de déverglaçage sur la réduction du volume d'eau, des débits de pointe, l’enlèvement des contaminants, et les micro-organismes, l'extrapolation de ces résultats à grande échelle doit être effectuée avec prudence. En effet, les mésocosmes sont des systèmes artificiels qui ne reproduisent pas entièrement les conditions réelles des biorétentions à grande échelle, notamment en ce qui concerne les températures hivernales inférieures à zéro, et ne sont étudiés qu’un lapse de temps. Aussi, les biorétentions réelles sont rarement plantées en monoculture, de sorte que les interactions entre les espèces végétales et les microorganismes pourraient influencer la performance du système. Pour une application réussie des conclusions de cette étude à l'échelle pratique, il serait essentiel que des projets pilotes valident ces résultats. / Bioretention systems are increasingly used to manage urban stormwater runoff. Plants are an essential component of bioretention, improving water quality and reducing runoff volume and peak flows. Although plants are an essential component of bioretentions, there is little evidence on how their contribution changes according to the species chosen, especially in temperate climates with large seasonal variations. In addition, bacteria and fungi could play a major role in bioretention performance by improving water quality or supporting plant efficiency. But to date, little is known about the microorganisms living in bioretentions and even less about the influence of the design choices of this system and therefore its consequences on the contribution of these organisms. In cold climates, the use of de-icing salts generates saline runoff which could affect the bioretentions performance, especially by reducing biological processes. The aim of my study was to compare the contribution of vegetation to the performance of bioretentions according to the species planted, to test the impact of saline runoff on the hydrology and the filtration of contaminants by the system according to the plant species used, and to characterize the influence of these 2 factors on the diversity, composition and relative abundance of bacteria and fungi within bioretentions. I performed a mesocosm experiment with 4 plant species commonly used in bioretention and covering a wide range of biological types (Cornus sericea, Juncus effusus, Iris versicolor, Sesleria autumnalis). To simulate spring salt-laden runoffs, the semi-synthetic runoff water used for irrigation was supplemented in spring with four concentrations of NaCl (0, 250, 1000 or 4000 mg Cl/L). Sequencing of 16S and ITS genes was used as bacterial and fungal diversity, composition and relative abondance indicator. In general, all bioretention mesocosms significantly reduced water volumes, peak flows, and contaminant loads. Some plants significantly increased the performances, by increasing water loss through evapotranspiration during the growing period (up to 2.5 times) and reducing water flow (up to 2.7 times). Plants improved macronutrients removal, with an average mass removal of 55% for total nitrogen, 81% for total phosphorus and 61% for K compared to –6% (release), 61% and 22% respectively for the unplanted. Except for Sesleria, mass removal of trace elements in planted mesocosms was generally higher than in unplanted ones (up to 8.7%). Their contribution level followed the same order as their evapotranspiration rate and overall size (Cornus > Juncus > Iris > Sesleria). Even at the highest concentration of NaCl tested, no impact on water volume reduction and flow rates were detected. In contrast, salt-laden runoffs temporarily increased removal of some metals such as Cr, Ni, Pb, and Zn. Overall, the plants tolerated the passage of salt runoff in spring. All bioretention mesocosms were naturally colonized by bacteria and fungi adapted to a humid and contaminated environment. Among the dominant taxa, several described functions related to the bioretention performances, such as involvement in the nitrogen cycle, the degradation of hydrocarbons, the tolerance and remediation of metals, or plant symbionts. My thesis emphasizes the importance of the plant species choice on the bioretentions performance. firstly because of their differential contribution to the water volume reduction, peak flows, and the contaminants removal, in particular macronutrients, secondly through their influence on the composition and abundance of microorganisms. Even though several of the dominant bacteria and fungi found would have the ability to contribute to the performance of bioretentions, further studies will be needed to verify their activity. Finally, my thesis demonstrated that the presence of de-icing salt in the runoff should not be a hindrance to the bioretention cells implementation, since at concentrations usually observed following salt spreading, no negative effect on the bioretentions performance, on the vegetation or the bacterial and fungal communities were observed in this thesis. However, the release of soil-structuring cations as well as salts exfiltration should be monitored in the longer term. Although these experiments allowed for the analysis of the effect of plant species and de-icing salts on the reduction of water volume, peak flow rates, contaminant removal, and microorganisms, the extrapolation of these results to a larger scale must be done with caution. Despite the advantage of characterizing the performance of bioretention in mesocosms in terms of replication and factor control, it is important to recognize that this approach also has its limitations. Mesocosms are artificial systems that do not fully replicate the real-world conditions of large-scale bioretention, especially concerning sub-zero winter temperatures and the limited time frame of study. The size of the mesocosms could lead to edge effects, such as preferential flows, while real-world scenarios involve slopes that create heterogeneous conditions within these basins, influencing factors like frequency and quantity of water received. Additionally, real bioretention systems are seldom planted as monocultures, thus interactions between plant species and microorganisms may impact system performance. For successful implementation of the findings of this study on a practical scale, validation through pilot projects would be essential.

The increasing highway traffic is leading to higher and higher levels of pollutants in the stormwater, such as suspended solids, metals, oils, nutrients and PAH. This runoff water causes the degradation of water bodies quality, leading to increased interest in the treatment of highway runoff for the removal of contaminants. Traditional treatments mostly consist of sedimentation, aimed primarily at the removal of suspended solids and particlebound contaminants. Recently the concern has shifted to the dissolved and colloidal fractions of contaminants as well, metals in particular, which makes reactive media filters a potentially interesting technology for runoff water treatment. However, very little research has been carried out on this technology so far, mostly in batch laboratory experiments with synthetic runoff water. This study aims at investigating the performance of four reactive media filters (Petrit T, Polonite, D-Rainclean and Filtralite-P) and a defined filter sand used for water processing at removing dissolved and particulate-bound metals. To achieve the goal, a pilot plant has been constructed in Gröndal, south of Stockholm, inside the existing Gröndalsmagasin, a full-scale stormwater treatment plant. The catchment area is located mostly on the E4 and Essingeleden, the highest trafficked motorway in Sweden with an AADT of 140 000 vehicles. A pump feeds the pilot plant from the Gröndalsmagasin outlet, after the stormwater has been treated with flocculation and sedimentation, and collects such water in a collection vessel. The five filter media columns are fed during runoff events with a flow scaled proportionally to the influent in the Gröndalsmagasin to mimic real conditions of flow and pollutant concentrations. Field measurements of pH, turbidity and conductivity, have been carried out during the trial. Samples have also been collected with flow proportional samplers connected to the collection vessel and the outlet of the five columns, as well as in the form of grab samples from sample ports at different depths in the columns. These samples have been sent to ALS Scandinavia AB to perform lab analysis. The pH in the effluent of the reactive filter media increased to values between 10 and 13, and slowly decreased during the trial. No difference was found between the pH in Filtersand outlet and the influent water. All materials showed excellent performance at removal of dissolved and particulate-bound Zn as well as particulate-bound Cu. Dissolved Cu was removed effectively by reactive media, and to a minor extent by Filtersand. Leaching of metals was observed from three materials: Filtersand released Mn during April, Polonite released Cr, with a very good correlation with the Chloride content and Petrit-T released Ba during the entire trials. / Den ökande vägtrafiken leder till allt högre halter av föroreningar i dagvattnet, till exampel suspenderade ämnen, metaller, oljor, närings ämnen och PAH. Detta dagvatten försämrar degradering av vattenförekomsternas kvalitet, vilket har lett till ett ökat intresse för rening av dagvatten fr ̊ån motorvägar för att avlägsna föroreningar. Traditionella behandlingar består oftast av sedimentering, som främst syftar till att reducera suspenderade ämnen och partikelbundna föroreningar. På senaste tiden har intresset förskjutits till att även omfatta lösta och kolloidala fraktioner av föroreningar, speciellt metaller, vilket gör reaktiva filtermaterial till en potentiellt intressant teknik f ̈or behandling av dagvatten. Mycket lite forskning har dock gjorts om denna teknik hittills, då oftast i laboratorieexperiment med syntetiskt dagvatten. Denna studie syftar till att undersöka hur fyra reaktiva mediafilter (Petrit T, Polonite, D-Rainclean och Filtralite-P) och en typ av sandfilter fungerar när det gäller att avlägsna lösta och partikelbundna metaller. F ̈or att uppnå målet har en pilotanläggning byggts i Gröndal, söder om Stockholm, inuti det existerande Gröndalsmagasinet, ett dagvattenreningsverk i full skala. Avrinningsområdet ligger vid E4 och Essingeleden, den mest trafikerade motorvägen i Sverige med en ÅDT på 140 000 fordon. En pump matar pilotanläggningen fr ̊ån Gröndalsmagasinets utlopp, efter att dagvattnet har behandlats med flockning och sedimentering, och samlas upp detta vatten i ett uppsamlingskärl. De fem filtreringskolonnerna matas under avrinningshändelserna med ett flöde som är proportionellt till inflödet till Gröndalsmagasinet för att imitera verkliga förhållanden av flöde och föroreningskoncentrationer. Mätningar av pH, turbiditet och konduktivitet har utförts under försöket. Prover har också samlats in med flödesproportionella provtagare som är kopplade till uppsamlingskärlet och utloppet fr ̊an de fem kolonnerna, samt i form av grabbprover fr ̊an provtagningsöppningar på olika djup i kolonnerna. Dessa prover har skickats till ALS Scandinavia AB för laboratorieanalys. pH-värdet i utflödet fr ̊ån det reaktiva filtermediet ökade till värden mellan 10 och 13 och sjönk långsamt under försöket. Ingen skillnad hittades mellan pH-värdet i filtrets utlopp och inflödesvattnet. Alla material visade mycket goda resultat när det gäller avlägsnande av löst och partikulät bundet Zn samt partikulärt bundet Cu. Lösta Cu avlägsnades mycket effektivt av reaktiva medier och i mindre utsträckning av Filtersand. Läckage av metaller observerades fr ̊ån tre material: Filtersand släppte ut Mn under April, Polonite släppte ut Cr, med en mycket god korrelation med kloridhalten och Petrit-T släppte ut Ba under hela försöket.