Enzo Salemi

Assessment of the Intrinsic Vulnerability of Agricultural Land to Water and Nitrogen Losses via Deterministic Approach and Regression Analysis

A set of indices was developed in order to classify the vulnerability of agricultural land to water and nitrogen losses (LOS), setting a basis for the integrated water resources management in agricultural systems. To calibrate the indices using multiple regression analysis, the simulation results of Groundwater Loading Effects of Agricultural Management Systems (GLEAMS) model for combinations of different soil properties, topography, and climatic conditions of a reference field crop were used as “observed values.” GLEAMS quantified (1) the annual losses of the percolated water beneath the root zone, (2) the annual losses of the surface runoff, (3) the annual losses of the nitrogen leaching beneath the root zone, and (4) the annual losses of nitrogen through the surface runoff, which were used to calibrate the following indices LOSW-P, LOSW-R, LOSN-PN, and LOSN-RN, respectively. All the simulations to gain the LOS indices were carried out for the same reference field crop, the same nitrogen fertilization, and the same irrigation practice, in order to obtain the intrinsic vulnerability of agricultural land to water and nitrogen losses. The LOS indices were also combined to derive nitrogen concentrations in the percolated and in the runoff water. Finally, the connection of LOS indices with the groundwater was performed using an additional equation, which determines the minimum transit time of the percolated water to reach the groundwater table.

Groundwater nitrogen speciation in intensively cultivated lowland areas

The study was conducted in Ferrara Province (Italy), a lowland area covering 2636 km2, located in the southern part of the Po River Delta. It is an intensively cultivated area, with more than 50% of land cultivated with winter cereals (32.11%) and maize (22.63%). The main nitrogen fertilizer used in this area is synthetic urea which is suspected to cause nitrate leaching towards shallow groundwater. A network of 56 piezometers, homogeneously distributed throughout the whole area, was installed in order to monitor both water table fluctuations and nitrogen species distributions in the shallow aquifer, over time. Data collected at the end of November 2010 were used to obtain maps of water table, urea (CO(NH2)2), ammonium (NH4 +), nitrate (NO3 -) and nitrite (NO2 -) distributions. Maps show an accumulation of NH4 + overlapping a stagnant zone, where drained peaty soils are present. The peaty soils are characterized by a pH ranging between moderately acid and slightly acid, and by high values of organic matter content. Along the drainage line induced by peaty soils dewatering, the flow velocity is very low or almost motionless, determining anaerobic conditions. Instead, the largest accumulation NO3 - is observed in the Eastern part of the province, where the groundwater head gradient is higher and soils are characterized by values of pH that range between 8.1 and 8.3, providing the best conditions for nitrification processes.

Assessment of specific vulnerability to nitrates using LOS indices in the Ferrara Province, Italy

A set of indices (LOS), based on a deterministic approach and regression analysis were used to assess intrinsic and specific vulnerability to nitrates in Ferrara Province, in northern Italy. To calibrate the LOS indices, using multiple regression analysis, the simulation results of GLEAMS model for combinations of different soil properties, topography and climatic conditions of a reference fieldcrop were used as “observed values”. Results of model were introduced in a GIS environment to obtain the vulnerability maps. Maps of water and nitrogen losses under the root zone (LOSW-P and LOSN-PN respectively) were used to obtain the map of relative concentration of percolated water (RCPW). Data on individual crops were used to calculate specific crop evapotranspiration rates (ETc) from potential evapotranspiration (PE). ETc values replaced PE values in the indices, to obtain both specific vulnerability map for water and nitrogen losses under the root zone (LOSW-P mod; LOSN-PN mod) and for relative nitrogen concentration of percolated water (RCPW mod). The RCPW mod map shows that concentration of nitrogen losses under the root zone is under water drinking limit of 50 mg/L (WFD; 2006/118/EC) all over the territory with highest concentrations along small areas of the coastal zone, where sandy textured soil are present (coastal dunes) and lowest concentrations where ETc is higher.

Comparison of Different “S-index” Expressions to Evaluate the State of Physical Soil Properties

The S-index describes the absolute value of slope at the inflection point of the water retention curve (WRC) and it is a popular tool for the indirect description of the soil physical-hydraulic conditions. It can be calculated as a function of van Genuchten (VG) parameters by four different expressions based on (a) the use of arithmetic or ln-transformed axis of pressure head h and (b) the use of volumetric θ or gravimetric W soil water content [the four expressions are obtained from the dθ/dh, dW/dh, dθ/d(lnh) and dW/d(lnh) derivatives]. The aim of the study is to test the ability of the four S-index expressions to rank soil physical-hydraulic (SPH) indicators such as fractions of porosity, bulk density and saturated hydraulic conductivity. WRCs of an artificial soil, which was subjected to flooding and extensive destruction of macroporosity, and WRCs from the twelve USDA soil classes were used in the analysis. The results showed that the expressions of dθ/dh and dW/dh display higher correlations versus the SPH indicators in comparison to dθ/d(lnh) and dW/d(lnh). The expressions based on the gravimetric approach such as dW/dh and dW/d(lnh) do not contain the parameter of bulk density while the expressions based on the ln-transformed axis of h do not contain the parameter a of VG model. Even though dθ/dh and dW/dh showed similar performance, the first one is suggested as the safest approach to describe SPH indicators for non-deformable soils because its theoretical background is more consistent and more relevant to porosity attributes which are always described in terms of volume and not of mass, whereas the second one is suggested for the case of deformable soils in order to avoid errors introduced by the changes of bulk density.

Seismic induced variation of hydraulic conductivity distribution in a large tank

A large tank (4x8x1.4 m) equipped with 26 standard piezometers and filled with sandy sediments (35 m3), was characterized via slug tests and grain size distribution analysis in 2006. The characterization via slug tests was repeated in 2013 after the seismic events of May 2012 near Ferrara (IT) characterized by a maximum Mw of 5.9 to infer the hydrological response of the tanks hydraulic conductivity field to the stresses induced by the earthquake. Results show a general decrease in hydraulic conductivity values, due to compaction phenomena, and a variation in the spatial distribution of hydraulic conductivity values, directly correlated to the presence of the finer fraction of the sediments within the large tank. In particular, the lenses, prevalently made of sandy sediments showed a negligible reduction, while the lenses constituted prevalently by silt displayed a larger decrease. The differential hydraulic conductivity reduction inhibited the flow linkages between the sandy lenses, provoking a four times decrease in the bulk hydraulic conductivity of the tank. This laboratory experiment highlights that pronounced hydraulic conductivity shrinkages should be expected after important seismic events in heterogeneous fine alluvial units.