Multi-Year Datasets of Groundwater Level and Surface Water Budget in a Central Mediterranean Site

1. Summary

Datasets including groundwater level, soil moisture, and meteorological parameters allow to estimate fluxes of water and energy across the atmosphere-land interface and to calculate water budgets [1,2,3,4,5]. Given their fundamental importance in climate and environmental sciences, series of data of hydrological and atmospheric parameters taken at specific sites are becoming increasingly freely shared [6,7,8].

Several regions in the world, including the Euro-Mediterranean, were affected by declining groundwater resources in the last decades [9,10]. Drought events linked to increases of temperature, in combination with decreasing of precipitation in the summer months, has been indicated among the main causes [11,12]. From July 2016 to June 2017 western-central Europe was affected by the most severe meteorological drought at the continental scale since at least 1979 [13]. Winter 2016/17 was drier than normal on the Central Mediterranean, with precipitation totals reached only 60% of normal precipitation [14]. In 2017, Italy experienced the driest year of the whole instrumental period of its climate history [15].

Since June 2017, the Institute of Atmospheric Sciences and Climate of the National Research Council of Italy (ISAC-CNR) takes measures of water-table depth of a shallow karst aquifer (the Miocene Aquifer of Central-Eastern Salento, hereinafter referred as MACES) located in the Salento Peninsula (Apulia region, Italy). The measures are taken from two wells located few hundreds of meters from the ISAC-CNR Micrometeorological Station (IMS) in operation since June 2002 [16,17,18]. As a whole, wells and station constitute a measurement site that allows direct comparison between hydrological and weather data. Due to their climate sensitivity and subsurface heterogeneity, the karst systems located in dry lands require particular attention in processing data related to the exchanges among air, soil and groundwater [19,20]. Series of data from this site have been used to interpret the local water consumption during the lockdown for COVID-19 [21]. Datasets of IMS were previously used to calculate recharge time and specific yield of the regional deep aquifer [22,23]. This note provides and makes available five hydrological years (from 1 October to 30 September in the Northern Hemisphere) of gathered data with the aim of promoting open science and fostering cooperation by making related description and insights shareable.

2. Data Descriptor

2.1. Physical Setting

Located in the Central Mediterranean, the Salento Peninsula constitutes a lowland plateau (maximum altitude 200 m) surrounded by the Ionian Sea to the West and South and by the Adriatic Sea to the East. Precipitations are usually limited (about 700 mm/y of mean annual precipitation) due to the low orography and the residual blocking effect of the Apennine Chain (150 kn NW of Salento), for the incoming cyclonic moisture flux coming from the Atlantic Ocean, generally accompanied by southern winds. Moderate northern winds typically blow during daytime in high pressure conditions, with clear skies and strong insolation. The warm/dry season (April–September) is characterized by temperatures that can reach 40 °C and more, and reduced precipitations, while the cold/wet season (October-March) by mild autumn–winter temperatures (generally above zero even at night) and enhanced precipitations. In recent years, there has been a generalized tendency to an increase in more intense, brief precipitation events during the spring-summer months [22].

Groundwater resources of the Salento Peninsula are contained in a complex system of aquifers, consisting of a main regional reservoir, the Cretaceous Aquifer (CA), and several shallow aquifers [24,25,26]. Among these latter, MACES is the most important in terms of stored groundwater volume. It is supplied by meteoric water and two Plio-Quaternary aquifers, and extensively exploited for irrigation and domestic purposes through several licensed and unlicensed wells. In the warm/dry season, MACES is recharged by irrigation water withdrawn from the CA (more hydrogeological features are given in Appendix A).

2.2. Measurement Site and Equipment

The measurement site is placed in a suburban area 5 km south-west from Lecce (Figure A1), within the Ecotekne Campus of the University of Salento, with mixed local vegetation like Mediterranean shrubs, pines and olive groves, scattered buildings, and quarried surfaces (Figure 1). Here, the water table of MACES lies at a depth of 20-22 m below the ground [21].

The IMS [16,17] has been a data provider for the Hymex (Hydrological Mediterranean Experiment) project long term campaign [27]. It is mainly devoted to water and energy surface-atmosphere transfer, and is composed by two complementary systems collecting data based on half-hour averages: 1) a 16 m height mast equipped with a fast response eddy correlation system and basic meteorological data; 2) a meteorological station that collects also standard meteorological and soil surface data in a dedicated data logger. The mast has a typical flux footprint fetch of the order of hundreds of meters [22,28], thus taking contributions from the campus and some immediately surrounding vegetated areas. Both data files from the two systems are stored in local memory and then transferred to the web data base once per day through UMTS connection system [18]. Details of the instrumentation for the considered period of measurements are the following.

2.2.1. Telescopic mast (top level 14 m above surface)

The Eddy covariance system (measurements of turbulence, heat and evapotranspiration fluxes) is composed by one Solent-Gill 20 Hz ultrasonic anemometer and one Campbell Kh20 Krypton hygrometer. One thermohygrometer for standard air temperature and humidity measurements (Rotronic MP100). Data are firstly collected in a dedicated notebook that calculates the half hour eddy covariance statistics rotated in the ‘streamline’ coordinate system [29] before storage in the daily data file.

2.2.2. Meteorological station

It is equipped with standard meteorological sensors and soil data sensors (data are collected by a dedicated Campbel CR1000-X datalogger).

Standard sensors (2 meters height):

Cup anemometer (Campbell A100R) , windvane (Campbell W200P) thermohygrometer (Rotronic MP100), four-components net radiometer (Hukseflux NR01), precipitation gauge (EM ARG100), barometer (pressure sensor Campbell PTB101B).

Soil sensors:

2 thermistor temperature sensors (Campbell109L, 2 and 5 cm depth), 2 thermopile soil heat flux sensor (Hukseflux HFP01, 2 and 5 cm depth), 2 moisture content capacitive sensors (Decagon EC-5, 2 and 35 cm depth)

A selection of the available data can be visualized at the address [18] in graphic format. Data are freely available upon request at the address indicated in the staff area. Part of the dataset is also available in the Hymex site for subscribed users [30].

2.3. Groundwater Level Measures

The two monitoring wells of the measurement site (named Fiorini, FW, and Benessere, BW, Figure 1) were hand dug about 25 m deep decades ago, to provide water for agriculture and rural settlements. The wells are easily identifiable by their stone well heads (Figure A3, Figure A4). The measurements of the phreatic level began in June 2017 for FW, and in June 2018 for BW. They have been taken using a portable electronic water-table meter (Boviar GST-FR100), with an average frequency of one measure every about three days for both wells (more frequently in conjunction with some rainy events; for more details see see Appendix B). Despite the laboriousness of the field operations, the use a water-table meter allow to overcome problems related to the lack of funding for research and the reliability of the measurements in case of equipment malfunction [31,32,33]. Embedded in well hydrographs there is valuable information on aquifer responses to both natural and human stresses [34].

2.4. Data Records

The dataset of the water-table depths taken from June 2017 to October 2022 is supplied as text file in Supplementary Materials (Table S1). The columns represent the following items:

- date (gg/mm/yy)

- water-table depth measured in FW (m)

- water-table depth measured in BW (m).

There are only few gaps in the dataset, because sometimes adverse weather conditions prevented the measure for one of the two wells. The curves of the variations of the groundwater head in the monitored wells are shown in Figure 2.

For FW, the measure of the water-table depth started in the middle of the warm/dry season (June 2017) of the 2016-2017 hydrological year. A year later (June 2018), monitoring of BW began. Note that the two lines fit almost perfectly. Only in very few cases, the daily variations in FW and BW were of the opposite sign (Table S1). A general negative trend from the hydrological year 2017-2018 to 2021-2022 is apparent. At the end of the 2021-2022 hydrological year (June-September 2022), the groundwater level in FW was nearly coincident with that of June-September 2017.

The datasets of surface water budget are available in the Supplementary Materials (Tables S2, S3) as text files of daily/monthly data. The columns represent the following items:

- date (Julian day or fraction of year representing the end of each month)

- precipitation (daily/monthly cumulative, m)

- real evapotranspiration (daily/monthly cumulative, m)

- soil moisture 2 cm (beginning of the day/month, m³/m³)

- soil moisture 2 cm (end of the day/month, m³/m³)

- soil moisture 35 cm (beginning of the day/month, m³/m³)

- soil moisture 35 cm (end of the day/month, m³/m³)

- estimated infiltration (daily/monthly cumulative, m)

- percentage of validated data (daily/monthly cumulative)

- corrected real evapotranspiration (daily/monthly cumulative corrected by percentage of lacking data, m)

- corrected estimated infiltration (daily/monthly cumulative corrected by percentage of lacking data, m).

The corrected estimated monthly infiltration values for the MACES are shown in Figure 3, where still possible negative values have been set to zero to properly represent the groundwater contribution (see also Section 3). Other data are plotted in Figure A6, Figure A7 (Appendix C).

3. Methods

To collect the water-table depth, reference points were fixed on the stone casing of each monitored well. A rod was placed on the well head to reference the depth read by the water-table meter (Appendix B, Figure A3, Figure A5). In this way it was possible to measure the depth with an accuracy of 0.5 cm. Furthermore, this system allowed the use of several operators without affecting the reliability of the data.

For the soil water balance, in order to estimate the net infiltration, the following mass conservation equation is considered [22,35,36]:

$$S(t+dt)=S\left(t\right)+P-E-I-R+Ir,$$

(1)

where S(t) is the soil water content per unit area, and P, E, I, R and Ir are the total precipitation, actual evapotranspiration, net infiltration, surface-subsurface runoff, and irrigation water in the time interval dt, respectively. Due to the hydrological features of the measurement site and its surroundings, the runoff term R has been neglected. In fact, the site is placed on a large sub-horizontal ground without any appreciable slope, thus neither stream systems nor persistent pond waters are present.

To quantify the surface water transfer data from the IMS the following data have been used:

1) cumulative precipitation P;

2) averaged evapotranspiration flux;

3) averaged soil moisture at 2 cm (s1) and 35 cm (s2) depth.

All data are taken on a 30 minutes base interval (averages, and cumulative for precipitation). The averaged evapotranspiration flux has then been transformed in 30-minutes cumulative evapotranspiration E.

After calculating the total monthly cumulative values for the above quantities the monthly net infiltration I has been calculated by the following expression between the initial (t) and final (t+dt) time of each day/month [22]:

$$I=P-E-S(t+dt)+S\left(t\right),$$

(2)

In equation (2) the Ir term has been neglected (as well as R). The anthropic irrigation is indeed absent in the measurement area but still present over the whole MACES region and will be the object of further studies. In equation (2) P and E are directly derived by the measurements, while S(t) has been calculated approximating an average soil moisture content per square meter as:

$$S\left(t\right)=h\left(s1\right(t)+s2(t\left)\right)/2,$$

(3)

where s1 and s2 are the available soil moisture density measurements (in m³/m³ at 2 cm and 35 cm underground respectively) and h represents an effective soil thickness estimated as 0.4 m approximately [22]. Negative values of I (eq. 2) are still possible in very dry conditions because of the approximations in equation (3) but they are to be considered as vanishing contributes to the MACES water balance (see also [22]). In the daily/monthly averaging of the evapotranspiration, where it is more likely to have some missing data, the actual percentage of the available data has been also calculated, resulting generally over 80%. The results have then been corrected adding to the cumulative value the product of the lacking data percentage times the calculated daily/monthly average, to approximately compensate the effect of the missing data. The estimated infiltration for MACES is shown in Figure 3 with the related linear trend. A more complete data view can be found in the Supplementary Materials (Tables S2, S3).

4. User Notes

The trend of the water-table fluctuations (Figure 2) apparently suggest a decline of the groundwater resource and is in agreement with the negative trend in the estimated infiltration for the considered five hydrological years (Figure 3). The infiltration decrease is associated to a decrease in precipitation in the considered five years, while the real evapotranspiration shows a less significant negative trend in the same period (see also Appendix C). Thus, the reduced surface water input affects the MACES groundwater input. This finding gives an insight about the natural stress that affected MACES from 2017 to 2022. However, no data are available about the anthropogenic stress [21]. Moreover, the recharge by irrigation water (Ir) withdrawn from the CA during the warm/dry season is unknown. This currently prevents the calculation of the actual aquifer water budget. The measurements in the site herein shown and the sharing of the updated datasets are scheduled to continue for some more years.