Integrated hydrologic and hydrogeologic modelling for basin scale subsidence analysis

Groundwater overexploitation and depletion of aquifers represent a challenging issue that impacts many areas of the World, especially during dry seasons and in arid regions. Water obtained from aquifers is considered one of the main sources of freshwater for drinkable and agricultural purposes and its uncontrolled usage can lead to many long-term consequences. Ground subsidence related to water levels decreasing is a common hazard especially occurring in large alluvial basins filled up with hundreds of meters of fine and compressible sediments, that can lead to several social and economic repercussions. Therefore, groundwater management policies for the sustainable use of the resource are needed, in order to reduce the effect of the phenomena associated with its over exploitation. Firenze – Prato – Pistoia basin (Tuscany region, central Italy) has a long experience of land subsidence and ground deformation, that has been observed in the area since the early 1960s. During the last three decades, with the development of InSAR technique, it has been possible to better analyze and identify ground displacement patterns occurring in Firenze-Prato-Pistoia basin, to characterize the effect of subsidence on structures and environment. Many studies ascribe Firenze – Prato – Pistoia basin subsidence to the over exploitation of groundwater resources, but neither quantitative analysis nor numerical modelling has been carried out so far to deeply investigate their relation. In last 20 years, PS-InSAR data identified a substantial decreasing of ground deformation patterns all over the basin. However, Pistoia area keeps showing alarming subsidence rates, enhancing the importance of clearly identify the causes of the phenomenon, in order to reduce the risk associated with it. Nowadays, numerical models represent one of the main available tools to support the development of good groundwater management policies. Relation between groundwater exploitation and ground deformation have been deeply studied in the past in many areas of the World and several modelling approaches have been proposed. One of the biggest uncertainties in these methodologies is represented by the traditional approach in water resource modelling, that used to represent in detail groundwater reservoir, simplifying the surface network only as boundary conditions. Even in aquifers where fluxes between surface and groundwater are limited, in long term condition is essential to consider the entire hydrologic cycle as a single system, in order to quantify the water exchanges between different reservoirs. Many coupled surface water – groundwater models can be found in literature, but many of them use a semi-distributed approach to describe the hydrological and geographical features of the simulated watershed. Fully distributed models are harder to set up and to configurate, but they guarantee a more realistic representation of the spatial distribution of the input variables all over the model domain. In the present study, a new time lagged integrated modelling approach has been proposed, in order to adequately represent the system of complex interaction occurring between groundwater and surficial processes. The modelling procedure is based on the integration between MOBIDIC hydrological fully distributed model and the USGS’s Modular Ground-Water Flow Model MODFLOW. MOBIDIC (MOdello di Bilancio Idrologico DIstribuito e Continuo) is a fully distributed, physically based hydrologic model developed by Castelli et al., 2009. MOBIDIC allows the quantitative and qualitative evaluation of the hydrological cycle components in the surface water networks, the soil-vegetation system and the subsurface layer. The model presents many innovations, comparing it with others hydrologic balance tools, as it couples the water balance in soil and vegetation with the surface energy balance for evapotranspiration computation and it guarantees detailed outputs for the groundwater-surface water interaction analysis. MOBIDIC is currently in place as flood early warning system for the Arno river basinTuscany, Central Italy) after an agreement between the national Civil Protection and the Arno river basin Authority. The model is born to supply hazard forecasts and predictions of likely flooded areas, but it can be coupled with others groundwater models in order to perform integrated simulation of the whole water cycle. The integrated methodology has been developed and firstly tested in the alluvial basin of Pesa river (central Tuscany) to prove its validity, and then it has been applied to simulate the water processes dynamic in Pistoia basin. Once the hydrologic model has been calibrated and validated, subsidence affecting the area has been modelled, by means of MODFLOW capabilities in simulating aquifer compaction. Nowadays, in the analysis of natural systems behavior, the impact of climate change on the environment is a variable that cannot be neglected. Climate change is an evolving phenomenon that is affecting all natural systems of the Earth, becoming one of the main threats to be faced in recent years, for its huge impact on environment and human lives. As a direct consequence of climate change, long term variation of meteorological variables may also affect river discharge, seasonal and local water availability, and groundwater states. Global Climate Models (GCMs) and Regional Climate Models (RCMs) are the most powerful tool for simulating future climate scenarios, providing all the basic information needed for climate change impact assessment. GCMs and RCMs output have been used in this study as input data to perform hydrologic forecasting with MOBIDIC model for the next decades. Then, applying the presented integrated modelling approach and by means of the calibrated subsidence model built with MODFLOW code, future predictions of land subsidence affecting Pistoia area in the years to come have been performed, based on different pumping rate and water dynamics scenarios. In order to detect the level of damage affecting buildings in Pistoia city caused by subsidence effects, two surveys campaigns were carried out. By means of the combination between damage information and ground displacement rate detected by means of PS-InSAR data, fragility curves of the surveyed buildings have been developed. Damage fragility curves were used to generate the subsidence vulnerability map of Pistoia city, in order to identify the damage probability distribution of building and structures as a function of ground displacement patterns. Finally, using the forecasted subsidence rate obtained by GCMS and RCMs simulations and the developed fragility curves for the area of interest, it has been possible to define the potential damage probability distribution that could affect Pistoia city in the future.