Preliminary results from the Submarine Gamma Imager

In this work, we present preliminary results of a novel Submarine Gamma Imager (SUGI) being developed based on pixelated CdZnTe detector modules. The instrument, mounted on a remotely operated vehicle (ROV), has been tested in a series of field deployments performed at the hydrothermal fields on the island of Milos, Greece. The analysis of the collected data demonstrate the capabilities of the instrument, while comparison with a reference gamma detector confirms the validity of the results.


Introduction
Gamma radiation imaging has applications in many fields, including, among others, medicine, scientific research, quality control in manufacturing as well as nuclear power plant decommissioning.It allows to localize and identify sources of gamma radiation, providing valuable insights into the composition and behavior of various materials and biological tissues.One of the most prevalent technologies for gamma radiation imaging is based on solid-state pixelated detectors, mainly using CdZnTe crystals.These detectors can convert the energy of incoming gamma rays into electrical signals, which are then digitized and processed to produce spherical radiation images and spectra.
Almost all the gamma imaging applications consider radiation imaging in the air or in space.In this work we present the Submarine Gamma Imager (SUGI), a novel gamma radiation imaging solution whose aim, similarly to its open-air counterpart, is the detection, localization and identification of radioactivity in underwater environments.The instrument is under development and targets numerous underwater applications, ranging from the identification and localization of possible radioactivity leaks from man-made objects or structures, to the monitoring of natural radioactivity emissions from tectonic regions of interest (e.g.hydrothermal fields).
Similarly to open-air gamma imaging detectors the goal of SUGI, besides gamma ray detection and spectroscopy, is the as-accurate-as-possible estimation of the direction of incidence of the detected gamma rays, accompanied by corresponding uncertainty quantification, based on Compton imaging.However, in comparison to open-air gamma imaging solutions there are several additional challenges caused by the underwater environment.These challenges include the design and validation of a suitable housing, and the development of suitable data processing algorithms to account for a significant amount of Compton scattering in the water.
This paper presents preliminary results of the instrument being developed, covering the challenge of developing a suitable underwater housing with a depth rating of 1000 meters that, on one hand, provides suitable protection of the detector components (CZT, electronics, HV, comms, etc.), while on the other hand affects the measurement as-little-as-possible in terms of gamma radioactivity shielding and scattering.To validate the correct operation of the instrument in the housing, we compare the measurements acquired from its modules with the ones acquired with an ordinary (non-pixelated) fully-characterized gamma radiation detector and we find that they are equivalent.The capability of the SUGI instrument to provide directionality information in the water is subject of future work.

Design and development
The proposed underwater gamma imager instrument (SUGI) is based on the GDS-100 platform developed by IDEAS AS.GDS-100, being an integrated solution comprising detector modules, acquisition electronics and management PCB in an OEM product, offers an optimal trade-off between customization ability and product maturity.GDS-100 supports up to four GDS-10 detector modules.Each detector module is paired with a 11 × 11 pixelated CdZnTe (CZT) crystal and has 121+2 detection channels, 121 anode channels for the 11 × 11 array and 2 special (cathode) channels.This configuration allows to perform Compton imaging by analyzing the spatial and temporal correlation amid the events registered by the device, either within the same detector or across different ones [1].The use of multiple pixelated detector modules is advantageous as it speeds up the collection of sufficient statistics for a reliable localization and identification of nearby radioactivity sources.This is particularly crucial in the underwater setting as the majority of gamma rays arrive at the detector after multiple scattering events.
The SUGI instrument, besides the GDS-100 system, comprises a High Voltage module, necessary for providing suitable bias voltage to the CZT detectors, a Raspberry Pi 4 single-board computer hosting all software necessary for logging, processing, and analyzing all data collected from GDS-100, and auxiliary electronic and network components (i.e., power modules, relays, network switches, etc).All these components are contained in a custom-made underwater housing rated for 1000 depth.The housing thickness and material, as well as the placement of the components in the housing, have been chosen based on simulations performed by the NuSTRAP research group of National and Kapodistrian University of Athens, in the context of the RAMONES project.A solution was chosen that balances between as-small-as-possible gamma radiation absorption while ensuring the aforementioned depth rating.Specifically, an aluminum alloy 6061-T6 of 4 mm thickness has been chosen with a hard anodized treatment for withstanding the expected pressure (100 bar) and the corrosiveness of sea water, especially in highly acid environments as the ones encountered in hydrothermal vents.Figure 1 shows the setup of the internal components of SUGI on the additive manufacturing base designed to fit the dimensions of the underwater housing.
The design of SUGI allows the mounting of the instrument both on a static benthic laboratory or on a remotely operated vehicles (ROV).Regarding the first solution, by mounting SUGI on a static underwater platform lying on the seabed, SUGI can perform prolonged measurements collecting data regarding the surrounding radioactivity.Mounting SUGI on an ROV, on the other hand, gives the possibility of moving the instrument in the vicinity of regions of interest where significant radioactivity events are located or suspected.The multipurpose design of SUGI enhances its versatility making it suitable across diverse radioactivity monitoring and mapping scenarios.
Data logging, processing, and communication with the GDS-100 is achieved using the TCP/IP protocol.TCP packets are used for system configuration and UDP packets for data readout.A custom python API has been developed, according to the GDS-100 TCP/UDP packet specifications (Doppio format), to allow real-time configuration, system health monitoring (e.g., key component voltage, temperatures, etc.), and data readout.A python implementation was favored, to allow easy deployment both on x86-64 desktop/workstation systems for development purposes, and on AARCH64 embedded computers, like Raspberry Pi 4, for testing and final deployment.
Preprocessing software modules have been developed to achieve basic tasks as reporting of total event counts and counts per second, the assignment of recorded events to ADC channels and the construction of the corresponding histogram, as well as the specific anode channels (i.e., pixels) where -2 - the event(s) were registered.Figure 2 shows the raw data registered for a single detection event and cumulative pixel activations from background radioactivity registered in the lab using all four ASIC modules of SUGI corresponding to 1000 registered events.From the digitized signals shown on the left, one can observe the charge measured at location (4,5) of the pixelated array as well the corresponding values for the cathode and the neighboring pixels.The latter is due to charge "leaking" from the central pixel caused by the fact that part of the generated electron cloud affects the neighboring pixels too.That phenomenon is dependent on the high-voltage steering grid of the sensor and is exploited in Compton imaging to provide sub-pixel accuracy for the location of the Compton interaction/event.-3 -Regarding the cumulative pixel activations shown on the right of figure 2, more activations are observed in the fourth detector module (ASIC 3).This is due to applying different gain settings between the detector modules to explore the capabilities of the newly developed system.Specifically, the fourth module uses gain settings that focus on gamma energies from 150keV up to 700keV, with respect to the other three modules that focus on higher energies (i.e., 700keV to 3MeV).

Field tests
Field tests with the SUGI instrument have been performed at the island of Milos, Greece from the 25th to the 27th of March 2023.Besides basic engineering testing demonstrating the correct operation of all the components of SUGI in a real environment, the goal of the test was to validate the readings of the instrument via comparison with a fully characterized reference CZT-based gamma detector based on the RITEC SPEC 4000 (Sniffer, figure 3).Both SUGI and the Sniffer have been mounted on an mini ROV at nearby positions and at the same height with respect to the ROV frame (see figure 3).For the validation in a real underwater scenario, measurements with both instruments have been collected from hydrothermal fields at the Paleochori and Alykes beaches (figure 4).During these experiments, different gas emitting sources (vents) have been visited with the ROV and radioactivity measurements were collected both at the seabed and at different water depths, to estimate radioactivity from the sediments and in the water column.
Tests at Paleochori beach.Field tests at Paleochori beach took place from the 25th to the 26th of March 2023.This location presents confirmed underwater sources of radioactivity [2], mainly daughter products of 222  contained in the gaseous vent emissions.We used both instruments (SUGI and Sniffer) to measure the activity of these underwater sources located at the hydrothermal vents, visually identified by the presence of bubbles as well as discolorations of the seabed (figure 5).
The first day regarded mostly engineering tests and some qualitative control of the instrument readings which were executed by standing on the seabed with the ROV for some seconds and then slowly ascending until reaching the surface.During this time, readings of the two instruments reported on-line were compared and the configuration of the instruments was adjusted in order to make the readings match as much as possible.On the second day, more detailed tests were executed, both using the touch--4 -  stay-and-rise setting described above, as well as slowly traversing the suspected sources horizontally at different depths, in an attempt to examine the distribution of radioactivity around the sources.

JINST 19 C03012
Tests at Alykes beach.On the third day of field experiments, we collected measurements at Alykes.For these experiments we used the ROV as a benthic station, anchoring it on the seabed for five minutes, collecting measurements each time at different positions on a circular pattern with a radius of about half a meter around a very active vent in shallow waters (around 1 meter depth, see figure 6).This experiment was conducted to validate the difference between background measurements, the measurements right above the source and those gathered at a certain horizontal distance from it.

Experimental results
At Paleochori multiple radioactivity hot spots have been identified.Figure 7 shows the evolution of the count-per-second (CPS) values for both SUGI and Sniffer instruments for a continuous acquisition while the ROV visited four different sources in the study area designated in figure 5.For robustness, CPS values have been calculated for a temporal window of 10 seconds.One can observe that the measurements taken from both instrument are matching to a very high degree.Both instruments detected the same sources and had similar background levels of radiation.Small differences can be attributed both to the differences in instrument efficiency, as well as differences due to their different placement on the ROV frame, exacerbated also by the significant radiation shielding of water as a sensor could be slightly closer to the source with respect to the other.Similar observations can be made on the data captured at Alykes. Figure 8 presents the energy spectrum and the corresponding activations at the level of detector pixels measured at source 'C'.Experiments at Paleochori focused also on the dependence of radioactivity levels on the vertical distance from the source in the water.For this analysis, the touch-stay-and-rise maneuvers where considered.In this settings, the CPS values computed based on the timestamped events registered by SUGI have been paired with the vertical distance from the sources (vents), computed based on -6 -  the depth readings of the ROV platform.The results from different iterations are shown in figure 9.One can note that radioactivity readings are high when the ROV lies on the seabed and they go back to background levels within less than one meter distance.

Conclusions
This work presented the development and initial field testing of an innovative underwater gamma radiation imaging instrument.The instrument aims to cover a significant gap in detecting and monitoring marine radioactivity which, although relevant in different scientific fields and civil protection activities, is vastly under-studied.The introduction of an underwater gamma imager can advance the study of gamma radiation emissions in various marine geology [3,4] and marine biology contexts [5,6] as well as emissions resulting from human activities such as leakage detection from drilling activities and/or nuclear power plants, nuclear waste management, to name a few.
The first round of field testing of the developed instrument comprised multi-day deployments of the instrument in areas of Milos island with high hydrothermal activity where high radioactivity -7 -levels were also expected and confirmed by the results.The results have verified the quality of the instrument readings and have also led to useful observations regarding the mapping of radioactivity underwater.After the successful field testing of SUGI, the development will continue with the design of suitable Compton imaging methods for estimating the direction of gamma rays, providing an integrated underwater gamma imaging solution.

Figure 1 .
Figure 1.SUGI component setup inside the underwater housing.

Figure 2 .
Figure 2. Left: raw digitized signals provided by the GDS-10 modules; Right: pixel activations on all four detector modules.

Figure 3 .
Figure 3. Left: SPEC 4000 gamma detector (inset) in its housing connected to control electronics; Right: SUGI and Sniffer instruments mounted on the ROV "METIS".The blue tapes in the inset image indicate the orientation of the detector modules at the ROV bottom side.

Figure 4 .
Figure 4. Locations of SUGI engineering tests performed at Paleochori and Alykes beaches in Milos.

Figure 5 .
Figure 5. Left: GPS track of the boat from which the ROV was deployed and area with hydrothermal vents visited with the ROV; Right: view from the ROV camera on top of a source located in the study area where seabed discoloration is also evident.

Figure 6 .
Figure 6.Left: location of hydrothermal vent studied at Alykes beach; Right: view from the ROV camera at Alykes where high degasing activity is observed.

Figure 7 .
Figure 7.Comparison of counts-per-second (CPS) readings from the baseline (Sniffer) sensor with CPS computed based on events registered by SUGI for Paleochori (left) and Alykes (right) locations.

Figure 9 .
Figure 9. Counts-per-second in relation to the vertical distance from the source.