A reproducible procedure to elaborate long groundwater level and abstraction rate time series acquired from data logger, a present-day necessity

In a climate change context, a sustainable water resource management is crucial and long time series of monitoring data become necessary. Recently, monitoring networks were improved, but the consequent extension of available data requires specific and reproducible procedures to effectively manage future big data. In groundwater studies, water suppliers could be valuable providers of long time series data. Indeed, they often store millions of groundwater level and abstraction rate data with a high temporal resolution, which often lie unexploited due to a missing dedicated operating procedure. In this work, a reproducible procedure to exploit long time series of groundwater levels and abstraction rates from data logger in operating wells is presented. This procedure allows the extraction of hydrogeological information from the data, improving monitoring networks’ efficiency and data management. This work relies on over 9 million of hourly data of groundwater level and withdrawal rate of 107 public wells over 10 years (2013-2022). These wells, managed by the water supplier Acque Bresciane S.r.l., serve 50 municipalities in the province of Brescia (N Italy). This method includes a first pre-processing phase consisting of: (1) homogenization of data acquired from different sources, (2) errors and outliers treatment, and (3) data association with the geographical localization and the well’s structure. The successive analysis phase allows to (1) separate static from dynamic groundwater levels, (2) classify groundwater levels according to abstraction rate ranges, (3) quantify the available static data and select the groundwater levels associated with the minimum withdrawal class (< 5 L/s), (4) extract and compare pluriannual, annual and seasonal trends of groundwater level, and (5) investigate the groundwater drawdown related to abstraction rates. Nowadays, in-depth analyses and elaborations of high-resolution hydrogeological data from automatic sensors are vital. Such elaborations can support and improve both monitoring networks’ efficiency and day-to-day management, pinpointing malfunctioning (errors, sensors’ misplacement, etc.) or evaluating the maintenance effectiveness on wells. As a result, a clearer evaluation of wells hydrodynamic behaviors is facilitated, alongside to the identification of tapped aquifer, of the possible interconnections among different compartments, and of the aquifer recharge and discharge assessment, also in climate change context.