Modeling water and pollutant connectivity in pristine and human-impacted landscapes

点击翻译按钮获取中文摘要

Ecohydrologists seek to better understand, predict and mitigate various environmental issues of societal and economic significance, including floods, droughts and water pollution. While those issues are created by different processes, they can all be examined through the lens of connectivity. Indeed, "connectivity" refers to the ability of a given landscape location "A" to transmit water and other materials (e.g., pollutants or biota) to another location "B", and it is a key indicator of terrestrial and aquatic ecosystem health. Given the lack of scientific consensus on connectivity measures, models are seen as pragmatic tools for estimating and testing hypotheses about connectivity. However, while existing precipitation-runoff models are used for connectivity estimation, few were explicitly designed with this goal in mind, raising questions about the reliability of their predictions. Most existing models rely on implicit connectivity simulation approaches, whereby only some of the landscape properties relevant to connectivity are considered through spatially aggregated proxy measures. Conversely, explicit connectivity simulation approaches are rare: they should account for a large number of landscape properties and allow their influence on connectivity to vary in space and time, either via equation-based modelling (EBM) or through methods imported from complex systems theory, such as agent-based modelling (ABM). The latter are relatively common in ecology but scarce in hydrology. The goal of the proposed research program is, therefore, four-fold: 1) compare the performance of existing precipitation-runoff models in their implicit simulation of connectivity; 2) create spatially explicit models of water and water-mediated pollutant connectivity using equation-based modelling principles, as is typical in hydrology; 3) create spatially explicit models of connectivity using agent-based modelling tools borrowed from ecology; and 4) assess the value of hybrid connectivity modelling approaches at different spatial and temporal scales. The focus will be on sites ranging from small catchments to large watersheds located in forested, agricultural and urban regions of Canada and other parts of the world: this will ensure that a good variety of landscapes is considered and spatially variable issues related to model data requirements and computational efficiency are appraised. Given its strong emphasis on data-model integration in pristine and human-impacted landscapes, the proposed research program will provide unique and vital training opportunities for future ecohydrologists.