Future changes in climate and land use are likely to affect catchment hydrological responses and consequently influence the amount of runoff reaching roads. Blockages and damage to under-dimensioned infrastructure can be extremely costly for the regions affected. This study aims to produce scientifically well-founded suggestions on adaptation of road drainage systems to climate changes resulting in more frequent floods. This thesis demonstrates the need to integrate aspects of climate change and land use impacts into the planning and practice of road construction and maintenance in Sweden. Tools such as hydrological models are needed to assess impacts on discharge dynamics. Identifying a ‘best’ practically performing hydrological model is often difficult due to the potential influence of modeller subjectivity on calibration procedure, parameter selection, etc. Hydrological models may need to be selected on a case-by-case basis and have their performance evaluated on an application-by-application basis.

The work presented here began by examining current practice for road drainage systems in Sweden. Various hydrological models were then used to calculate the runoff from a catchment adjacent to a road and estimate changes in peak discharge and total runoff resulting from simulated land use measures. Overall, the results indicate that the specific effect of land use measures on catchment discharge depend on their spatial distribution and on the size and timing of storm events. Scenarios comprising a changing climate up to 2050 or to 2100 and forest clear-cutting were used to determine whether the current design of road drainage construction is sufficient for future conditions. Based on the findings, the approach developed can be used for similar studies, e.g. by the Swedish Transport Administration in dimensioning future road drainage structures to provide safe and robust infrastructure.

Furthermore, a statistical method was developed for estimating and mapping flood hazard probability along roads using road and catchment characteristics. The method allows flood hazards to be estimated and provides insight into the relative roles of landscape characteristics in determining road-related flood hazards. Overall, this method provides an efficient way to estimate flooding hazards and to inform the planning of future roadways and the maintenance of existing roadways.

Understanding the role of road networks in alteration of hydrological responses is crucial for maintaining the accessibility and durability of road infrastructures. Road construction is one of the most common man made disturbances to a landscape. However, still the quantitative role of road topographical and geo-morphological properties on the hydrological response

of storms in catchments is only partially understood. The aim of this study was to use new methods to estimate and quantify the flood hazard probability with reference to the most influential physical catchment descriptors and road characteristics. In addition physical based modelling was used to estimate the effect of road topography on the hydrological responses of watersheds to storms with different intensities. A simple method was developed and discussed to address the most susceptible locations to flooding along the roads. Multivariate statistical analysis (PLS) employed to quantify the flood risk probability in the road-stream crossings concerning the correlation between the quantities of the physical catchment descriptors and occurrence/absence of flooding. The most influential factors in describing the probability of flooding along the roads were topographic wetness index, soil properties, road density and channel slopes. A detailed study of simulated flow duration curves showed differences between 20 watersheds for three different storms based on a digital elevation data with and without roads. An increase in peak flow and reduced delay occurred with increased storm intensity. However, the impact of the roads was much smaller and only possible to identify by detailed examination of statistical descriptors.

An efficient maintenance of roads to ensure high accessibility and durability of the transport capacity requires an understanding of how the hydrological response depends on both the road and the landscape characteristics. New methods and data were used to identify and explain interaction between roads and surrounding environment and their influence on hydrologic responses both in watershed scale and road-section scale. In the watershed scale, flood hazard probability was made with reference to the most influential physical catchment descriptors and road characteristics. Additionally, a physical based model was used to estimate the effect of road topography on the hydrological responses of 20 watersheds to storms with different intensities. A simple method was developed and discussed to address flood risk probability in the road-stream crossings concerning the correlation between the quantities of the physical catchment descriptors and occurrence/absence of flooding. The most influential factors in describing the probability of flooding along the roads were topographic wetness index, soil properties, road density and channel slopes. A detailed study of simulated flow duration curves showed differences between the 20 watersheds for three different storms based on topography with and without roads. An increase in peak flow and reduced time to pick occurred with existence of roads and increased storm intensity.In the road-section scale, an uncertainty-based simulation approach was used to identify the most influencing processes in controlling the dynamics of the groundwater level. A model (CoupModel) set up with four different geological stratifications was made to model two positions in a slope upstream of a road with drainage pipes and ditches. Results from the simulations indicate the significance of precipitation rate, road drainage and position in hillslope, and soil properties and stratifications in controlling groundwater levels. The same model was also applied to simulate soil moisture and temperature dynamics in two road sections by using groundwater and climate data. Porous media properties were obtained as statistical distribution function that provided the best performance of moisture and temperature dynamic in the road layers and underlying soil.

Ett effektivt vägunderhåll som garanterar tillgänglig och varaktig transportkapacitet förutsätter kunskap om hydrologin i ett område beror av både vägens och landskapets egenskaper. Nya metoder och data utvecklades i projektet och användes för att identifiera och förklara samspelet mellan vägar och omgivande miljö, samt hur detta inverkar på hydrologiska svar på nederbörd, i skalor som avser hela avrinningsområden och/eller enskilda vägsektioner. I avrinningsområdet identifierades de mest betydelsefulla faktorerna för översvämningsrisker, så som fysikaliska avrinningsområdesegenskaper och vägegenskaper. Dessutom användes en fysikalisk modell för att beräkna effekten av hur vägarnas topografi påverkar avrinningens dynamik vid nederbörd för 20 olika avrinningsområden med olika nederbördsintensitet. En metod för detta utvecklades och användes för att diskutera översvämningsrisker i sektioner där vattendrag och väg korsas, genom att sammanföra egenskaper som är korrelerade med förekomst eller avsaknad av översvämnningar. De mest betydelsefulla faktorerna för förekomst av översvämning var topografiskt fuktindex, markegenskaper, vägtäthet och lutning hos vattendragen. En detaljerad studie av simulerad varaktighet av avrinningsintensiteter visade skillnader för de 20 olika områdena och 3 olika nederbördsintensiteter beroende på om områdena innfattade vägar eller ej. En ökning av toppflöden och en reducerad tid för att nå toppflödet erhölls för områden med vägar.För vägsektioner användes en osäkerhetsbaserad metod för att identifiera de mest betydelsfulla processerna som reglerar dynamiken hos grundvattennivån, genom modellen (CoupModel). För ändamålet definierades 4 olika geologiska lagerföljder för 2 positioner i en sluttning uppströms en väg med dräneringsrör och diken. Resultaten från simuleringarna beskrev hur betydelsen av nederbördsintensitet, vägdränering och vägens position i sluttningem samt markegenskaper och dess lagerföljder påverkar grundvattennivåns dynamik. Samma modell användes också för simulera dynamiken hos markvatten och marktemperaturer i 2 vägsektioner genom att använda data om grundvattennivå och klimat som dynamiska styrande randvillkor. Egenskaper hos porösa media erhölls genom statistiska fördelningsfunktioner av parametervärden som på bästa sätt återgav dynamiken av fuktighet och temperatur i vägens olika lager och underliggande mark.