Stochastic management of pump-and-treat strategies using surrogate functions

Typical pump-and-treat (PAT) optimization problems involve design of pumping schemes, while minimizing cost and meeting a set of constraints. Due to scarcity of information about the hydrogeological system, stochastic modeling approaches can be used to assess tradeoffs between optimality and reliability. Using a stochastic approach, the constrained, single-objective problem may be turned into a multiobjective problem by substituting constraint inequalities with an additional objective function (OF) that accounts for the reliability of the PAT process. In this work, two approaches are analyzed: in one case, the additional OF consists of the probability of failure of a given remediation policy; in another, the OF additional is represented by the recourse, namely the penalty cost induced by the violation of constraints. In order to overcome the overwhelming computational cost required by stochastic simulation, surrogate forms of the OFs are introduced. In the test case under investigation, such functions are estimated by a kriging interpolation of the OF over a series of data points obtained from stochastic simulations of flow and transport, and calibrated against stochastic optimization solutions. The analysis of the two approaches for addressing the tradeoff of cost vs. reliability indicates that recourse accounts not only for the frequency of constraint violations, as the probability of failure does, but also for the intensity with which these occur. Ultimately, the recourse method allows considering less restrictive policies, although these may be highly sensitive to the choice of penalty functions.     

Geochemistry of the Submarine Gaseous Emissions of Panarea (Aeolian Islands, Southern Italy): Magmatic vs. Hydrothermal Origin and Implications for Volcanic Surveillance

The marine sector surrounding Panarea Island (Aeolian Islands, South Italy) is affected by widespread submarine emissions of CO₂ -rich gases and thermal water discharges which have been known since the Roman Age. On November 3^rd, 2002 an anomalous degassing event affected the area, probably in response to a submarine explosion. The concentrations of minor reactive gases (CO, CH₄ and H₂) of samples collected in November and December, 2002 show drastic compositional changes when compared to previous samples collected from the same area in the 1980s. In particular the samples collected after the November 3^rd phenomenon display relative increases in H₂ and CO and a strong decrease in the CH₄ contents, while other gas species show no significant change. The interaction of the original gas with seawater explains the variable contents of CO₂, H₂S, N₂, Ar and He which characterize the different samples, but cannot explain the large variations of CO, CH₄ and H₂ which are instead compatible with changes in the redox, temperature and pressure conditions of the system. Two models, both implying an increasing input of magmatic fluids are compatible with the observed variations of minor reactive species. In the first one, the input of magmatic fluids drives the hydrothermal system towards atypical (more oxidizing) redox conditions, slowly pressurizing the system up to a critical state. In the second one, the hydrothermal system is flashed by the rising high-T volcanic fluid, suddenly released by a magmatic body at depth. The two models have different implications for volcanic surveillance and risk assessment: In the first case, the November 3^rd event may represent both the culmination of a relatively slow process which caused the overpressurization of the hydrothermal system and the beginning of a new phase of quiescence. The possible evolution of the second model is unforeseeable because it is mainly related to the thermal, baric and compositional state of the deep magmatic system that is poorly known.     

Forecasting of storm rainfall by combined use of radar, rain gages and linear models

An integrated approach to real-time prediction of point rainfall is presented. This is based on the assumption that hourly rainfall at a station can be predicted by a Multivariate AutoRegressive Integrated Moving Average (MARIMA) process. The real-time calibration of the multivariate model is performed by combining radar maps and data from rain gages. Accordingly, radar maps provide the basic information for a storm tracking procedure which enables to detect the direction and the speed of storm movement. Storm tracking is used to select those stations which are characterized by the highest Lagrangian cross-correlation of observed precipitation, and which are therefore best suitable for application of the multivariate model. The parameters of the multivariate model are finally estimated using only observed rainfall at the selected stations throughout the current event. Preliminary results of an application to some events which occurred in northern Italy show that the combined use of radar and rain gages allows for an increased efficiency of the MARIMA model performances, as compared with empirical selection of stations to be considered by the multivariate model. The multivariate approach performs better also when it is compared with simple nowcasting procedures based on rain gage data or on radar data used separately. Finally, some considerations are issued in view of a systematic use of this technique to nowcast rainfall intensity in small urban or natural catchments, with a response time of less than 1 or 2 h.     

Extensional shear zones as imaged by reflection seismic lines: the Larderello geothermal field (central Italy)

The Larderello geothermal field is located in the Inner Northern Apennines, in an area which has been subject to extension since the Early Miocene. The latest extensional episode (Pliocene–Present) has resulted in the formation of NW-trending, NE-dipping listric normal faults, whose geometry is controlled down to 3 km by borehole data. In this paper, we integrate a new interpretation of seismic reflection lines with existing seismic, field, and borehole data to analyse the relations among listric normal faults, the top of the brittle–ductile transition, and the migration of geothermal fluids.In accordance with previous interpretations, we consider the strong reflector (K-horizon) marking the top of the reflective mid-lower crust, and located at a depth of 3–5 km in the geothermal area, to represent the top of the brittle–ductile transition. Its reflectivity most probably derives from the presence of overpressured fluids. We identify three main NW-trending, NE-dipping extensional brittle shear zones, showing listric geometry and soling out in the vicinity of the K-horizon. The latter appears to be dislocated in correspondence of the soling out of the shear zones. These shear zones, because of the associated intense fracturing, represent the most natural channels of upward migration of geothermal fluids from the magmatic sources located below the K-horizon.We suggest that these two conclusions—that listric normal faults root at or near the brittle–ductile transition, and that they act as preferential upward migration paths for magmatic fluids—may be of general validity for geothermal fields located in extensional settings.     

Factors Controlling Seismic Susceptibility of the Sele Valley Slopes: The Case of the 1980 Irpinia Earthquake Re-Examined

Most of the documented slope failures triggered by the 1980 Irpinia earthquake (M_s 6.9) occurred in the upper Sele valley epicentral area (southern Italy). The early investigations revealed some puzzling characteristics of the slope failure distribution, i.e., (i) the higher landslide concentration on the valley slopes located farther away from the earthquake fault; (ii) the predominance of re-activations over first-time movements. The analyses of factors controlling the landslide concentrations indicates that the differences in hydrological setting and in slope were the two main causal factors whereas the seismic shaking, according to the radiation pattern modelling, could have been characterised by a relatively low rate of decrease across the valley. The aspect of the slopes did not play a significant role. The differences in groundwater conditions between the western and eastern valley sides were probably enhanced by the earthquake. In addition to the probable pore-water pressure rise, the seismic shaking caused large increases in the flow of springs draining the western aquifer, and this made the adjacent flysch slopes more prone to landsliding. Data from the available literature suggest that the effects of earthquake-induced groundwater release on seismic landslide distribution is especially important for normal-fault events. The Sele valley case also indicates that the slope of the pre-existing landslides is an important factor controlling their susceptibility to seismic re-activations.