Assessing a cautionary groundwater level in roadworks is crucial in minimizing the risk of flooding and of hydraulic uplift throughout the entire period of road operation.
A Monte Carlo probabilistic forecast of groundwater rise in the next 50 years from today was performed in order to determine the interaction between groundwater levels and a depressed freeway to be built on the outskirts of Milano, Italy.
This approach was favoured over deterministic ones (e.g. worst case scenarios) since it made it possible to assess the uncertainty intertwined with all the interacting variables and incorporate it in the research output. These variables, namely recharge from precipitation, well withdrawal and irrigation, were studied both in terms of their current influence and probabilistic evolution over time.
A review of all the land and water management plans compiled by Local Authorities provided information about the evolution of agricultural and green areas. Demographic forecasts by the National Institute for Statistics (ISTAT) helped to obtain the probabilistic curve of well withdrawal for drinking and industrial use.
Meteorological statistics at the study area were used to draw a probabilistic curve of groundwater recharge, using a cautionary approach, although studies by the Intergovernmental Panel on Climate Change (IPCC) forecast a potential limited decrease in precipitation.
A steady-state Modflow model was run several hundred times to determine the probability density functions of groundwater levels at control points along the road. The seasonal rise due mainly to irrigation was then sampled starting from probabilistic distributions based on recorded levels and finally summed to the model outputs.
The cumulative distribution functions of groundwater levels at the control points show a significant probability (i.e. > 5%) of exceeding the safety thresholds resulting from a standard approach adopted by road designers. This highlights the limitations in the current approaches since the future evolution of the groundwater driving variables plays a key role in the design of structures with an extended lifespan.