​

The proposal seeks to establish a National Geochronological Database as part of the GSI National Geoscience Data Centre, underpinned and expanded by the development of a novel in situ Rb–Sr dating protocol, made possible by reaction cell technology within the new multi-collector mass-spectrometer being commissioned at UCD. The database will be designed in consultation with GSI, iCRAG and the international geochronological community and will be populated initially by a quality-screened compilation. 

The database will be underpinned by new Rb–Sr dates of feldspars and micas in Leinster granite and lithium pegmatite samples, benchmarked against existing zircon U–Pb SIMS ages for different pulses of granite intrusion and then populated by a quality-screened compilation of Irish geochronological data. The proposed Rb–Sr capability aims to expand the range of minerals amenable to precise (<2%SE) in situ dating in Ireland beyond accessory phases, allowing phases such as magmatic micas and K-feldspar, hydrothermal white mica and celadonite, and diagenetic glauconite and adularia, among others, to be dated, with manifold applications across geoscience disciplines. Visualized and disseminated alongside decades of legacy geochronological data in the proposed database, this method and project aims to reshape our approach to the fourth dimension of Irish geology.

Title: EIRbSr – Establishing a national geochronological database underpinned and expanded by a new Irish LA–ICP–MC–MS/MS Rb–Sr capability

Principal Investigator: Thomas Belgrano

Host: University City Dublin (UCD)

​
While the C pool and sink function of peatlands are well known, their impact on atmospheric CH4 concentrations is unclear. Reliable estimates of current and past peatland CH4 fluxes is key to determine their impact on future climate. The MATCH4 project will bridge an important gap between the geological records and real-time GHG measurements to provide insights into the role of long-term CH4 release, climate change, and C storage in peatlands. The overarching goal MATCH4 is to (i) develop a testate amoeba-based proxy to infer long-term dynamics in CH4 fluxes, and (ii) increase our scientific understanding of drivers of CH4 fluxes. MATCH4 will develop and employ a novel approach to the investigation of long-term CH4 emissions. To achieve this, we will build an inference model based on the relationships between contemporary testate amoebae and CH4 emissions at the plot level across three peatlands with long-term monitoring. Preliminary data shows this is possible, but no reliable model exists to date. Moreover, the strategic agility and unparalleled value for money will be maximized through the integration of study fundings with an international project led by The Neuchâtel University, Switzerland, further promoting GSI as a key player in peatland research.

​
This project aims to push the boundaries of our current understanding and capabilities in interpreting, analysing, and leveraging hyperspectral core scanning data for geological applications. This project pioneers novel chemometrics including K-means clustering and Multi-way analysis to uncover hidden geological insights underlying raw hyperspectral data. A synergy of unsupervised and supervised machine learning techniques will be performed to obtain a holistic view of geological domaining and correlation. The project's applicability will be demonstrated through a case study on two Irish Carboniferous correlation sections. Furthermore, the project innovatively develops predictive models for petrophysical properties, enhancing the utility of hyperspectral data in geothermal exploration. The optimal models will be applied to predict the petrophysical attributes from hyperspectral data of other boreholes with Carboniferous sequences. As a proof of concept, this continuously predicted borehole data will then be integrated into a subsection of an existing 3D reservoir model. Ultimately, the quantitative prediction of petrophysical indices through hyperspectral core scanning not only refines our geological understanding but also boosts resource exploration and reservoir assessments. This project represents a truly interdisciplinary endeavour, leveraging substantial expertise from applicants in hyperspectral imaging, machine learning, and geological sciences, ensuring a knowledgeable approach to its execution.

​

Cliffs act as natural barriers against coastal hazards such as storm surge flooding and extreme waves and runup; but they can also be undercut and fail as a result of hydrodynamic forcing. Rockfalls and landslides from eroding cliffs represent a significant hazard to humans and infrastructure around Ireland's rocky coasts. Understanding the complex dynamics of cliff erosion is essential for assessing and mitigating coastal hazards and for understanding future response of these systems to climate change and sea level rise. Despite their ubiquity along Ireland's coastline, rates of rock cliff recession are largely unknown, and their failure mechanics remain poorly understood. The proposed research aims to address this knowledge gap by employing multiple complementary approaches to quantify cliff erosion rates at representative locations. We will use these in a comprehensive analysis of the drivers of coastal cliff recession on Ireland's rocky shores.

The proposed research will:

(1) Quantify cliff recession at different temporal and spatial scales using Persistent Scatterer Interferometry Synthetic Aperture Radar (PS-InSAR) and Structure from Motion (SfM) photogrammetry.

(2) Perform numerical analysis of the main factors driving coastal cliff erosion at selected sites through characterisation of rock mass properties and process-based modelling.

Title: Estimating shoreline recession rates for coastal hard cliffs using time-series elevation data and additional datasets

Principal Investigator: Niamh Cullen

Host: Dublin City University (DCU)

​With a reinvigorated interest in offshore geology driven by the critical need for evidence-supported strategies for marine spatial planning - including offshore renewable energy (rock anchors for floating wind devices), marine infrastructure and marine protected areas - a detailed reappraisal of the shallow marine bedrock geology of Ireland is urgently needed. At present only a low-resolution solid geology map is available on the EMODnet portal, however, with the acquisition of novel high-resolution multibeam echosounder (MBES) data from the INFOMAR programme, significant improvements can be made. The INFOMAR MBES dataset allows for an extension and improvement of the geological interpretation of the shallow marine regions but must rely on computer algorithms to effectively facilitate the task.

We propose a research project with the following objectives. Firstly, machine learning techniques (FCNNs) will be used to separate sediment cover from bedrock outcrops thus delineating the mappable areas, and discriminate bedrock into different morphotypes. Machine learning will produce confidence scores and provide repeatable methodologies for future application. Secondly, the bedrock morphotypes will be further characterised coupling semi-automated geomorphometry techniques with manual supervision, quality control and classification. Results will include a classification of the bedrock at stratigraphic Group level, major structural elements and allowing for the construction of simple cross sections. Finally, links to onshore geology mapping will be attempted with measurements of confidence included in the mapping.