The following projects are available for candidates who would like to apply to study at Newcastle University.
Spatial modelling of present and future extreme wind speeds for improved design and resilience of national infrastructure.
Project application reference GEO22_01
The main research gap concerns the very outdated return level maps for extreme wind speed return levels (such as the 50-year level, used in building design) and the failure of current standards to build in any resilience to a changing climate of extreme wind speeds. We propose a network of regional spatial Bayesian models for extremes, whose power in combining information from multiple sites will greatly improve (a) the reliability of return level maps in reflecting current conditions, and (b) the ability to detect and model the climate change signal as it applies to extreme winds. This latter can be used in conjunction with recently developed high resolution regional climate models, developed and used by the Met Office, to construct models for the future risks to structures and infrastructure posed by extreme wind speeds. This work will be co-supervised by Simon Brown, Climate Extremes Science Manager at the Met Office Hadley Centre, and the applicant. High quality data already exists as a result of recent collaboration between the supervisors, carried out specifically for the purpose of building enhanced models for extreme wind speeds with a much broader scope than existing models. Skills gained as part of this training would be the ability to build and implement sophisticated Bayesian spatial models, and the data management capabilities needed for such a project.
During the first year, as part of the MRes in Geospatial Data Science, the project will include the design and implementation of a limited spatial Bayesian model for extreme wind speeds for a single region in the UK, with the possibility for incorporating a climate change signal.
This will then lead to the development into a national model, incorporating a network of regional models, and allowing for variation in possible climate change signals for extreme winds. The idea of using the model in conjunction with state-of-the-art regional climate models will be introduced and developed, with the aim of calibration of the changing extreme-wind risk within the climate models, and hence future risk-assessment as it applies to structures and national infrastructure.
Candidates should have a minimum of a 2:1 undergraduate degree (or equivalent) in Mathematics and/or Statistics. This degree would have incorporated appropriate skills in statistical inference, statistical modelling, data management and programming, for instance using R or Python. Knowledge of R would be extremely beneficial.
This project would be co-supervised by the Met Office.
Autonomous reality capture: from 3D data acquisition to digital twin creation
Project application reference GEO_02
Leica Geosystems, owned by the Swedish industrial group Hexagon, recently released two new autonomous reality capture instruments designed for advanced robots and aerial drones. The devices autonomously map areas of interest in 3D point clouds with potential to create accurate digital twins. The BLK ARC is compatible with the Boston Dynamics Spot robot and comes equipped with panoramic cameras, a high-resolution camera, and a dual-axis lidar to map its surroundings. The ARC complements Spot’s existing technology, allowing it to better navigate its surroundings. The BLK2FLY is an autonomous flying laser scanner used to explore hard-to-reach areas and map entire buildings. The lidar-based UAV charters optimal flight paths to produce 3D digital twins. For further details, see https:/blk2021.com
During the first year, as part of the MRes in Geospatial Data Science, this project will comprise preliminary background research into the BLK technology and assess both its capabilities and limitations for digital twin creation. You will apply your newfound skills in machine learning, big data, GeoBIM and visualisation, all acquired via the CDT’s MRes taught programme, to data from the BLK technology to address an example smart city related problem of mutual interest to you and project partners Hexagon (e.g. urban infrastructure monitoring). Findings from the MRes project will inform design of the subsequent PhD programme.
Applicants should have a good (2:1 or better) undergraduate or masters (merit or better) degree in computing science, mathematics, physics, engineering, surveying, informatics or related geospatial disciplines. Programming skills (e.g. in Python or C++) are desirable but not essential as training in programming will be provided via the MRes taught component and support will be available throughout the research programme. Skills and knowledge of surveying, 2D and/or 3D scene understanding is desirable but not essential. Likewise, a PfCO in drone operations is desirable, but not essential and can be obtained if needed via funding from the research training support grant.
This project will be co-supervised by Hexagon Geosystems
Networking Newcastle’s Coastal Observatory: Using underwater sensors to monitor marine environments
Project application reference GEO22_03
Newcastle University has established itself at the forefront of sensor-based observations through the establishment of the Urban Observatory. The Coastal Observatory is scheduled for deployment in Spring 2022 and will expand the sensor network into the marine ecosystem to provide real-time information about environmental conditions along Northumberland’s coastline. The proposed PhD project associated with the Coastal Observatory sits at the intersection of environmental monitoring, marine engineering, and big data analysis. The PhD student will be exposed to different professional environments including (1) working with marine scientists to deploy/recover sensors onboard the Princess Royal research vessel (2) working with the School of Engineering to package the data to enable live streaming via underwater acoustic communications (3) develop data processing protocols so that the data is displayed on the publicly available website hosted by the Urban Observatory. The studentship is expected to communicate with key stakeholders in the Coastal Observatory including wind farm operators and public bodies to demonstrate the value of these observations. Ultimately, the Coastal Observatory seeks to establish itself a low-cost platform providing underwater observations and there will be opportunities for the student to engage with UK and international institutes with similar objectives.
The focus of Year 1 is to ensure that the student understands the entirety of the research project and is comfortable with all areas of operation. For example, it is anticipated that the student will not have previously exposed to working with underwater sensors. Therefore, during Year 1 the student will learn how to maintain and calibrate sensors for deployment underwater. Similarly, it is unlikely that the student has previously made scientific datasets publicly available and therefore the individual will work alongside the Urban Observatory to learn the best practices. At the end of the MRes, the student is expected to demonstrate a completely operational system for at least a single sensor beginning with data generation, followed by data transmission, data processing, and concluding with data publication online.
Candidates should have a minimum of a 2:1 undergraduate degree (or equivalent) in Computing, Engineering, Geography, Geomatics, Environmental Sciences, Mathematics and Statistics, or related disciplines. It is also anticipated that the ideal student will have a background in data sciences, with a strong interest in applying their data skills to solving environmental problems. There are currently very few underwater coastal observatories around the world and it is likely that the student draws upon the lessons learned from observatories on land. Therefore, the student should develop strong communication with their Geospatial Systems colleagues and receive the project-specific training related to the sensors, acoustic communications, and marine sciences. The student is expected to understand the ethics and scientific importance of data transparency and to be able to communicate its relevance to a range of stakeholders including academics, industry, and public bodies. Above all, the student should be excited about the unique opportunity to help develop Newcastle’s first coastal observatory and establish the foundations for data communication.
This project will be co-supervised by the UK Hydrographic Office
Detecting high-mountain hazard cascades using remote sensing and real-time data streams
Project application reference GEO22_11
Landslides from high mountain peaks, glacier collapses, glacial lake outburst floods (GLOF) and landslide-dam outburst floods (LDOF) cause far travelled hazard cascades – often in the form of deadly debris flows and floods. Even the more dilute sediment plumes that can be detected 1000+km from the initial source are capable of extreme disruption for Hydropower turbines and water quality for irrigation/drinking. We think these hazard cascades are becoming more common ins a warming world, or, at least interacting with us more often, but, can we test that, and, can we produce tools to reduce disaster impacts using remote sensing, real time sensors and big data? Here you would produce a tool to detect far travelled sediment plumes triggered by hazard cascades in the satellite era, and, make that tool run live as new imagery comes in to detect new events – sometimes otherwise unknown. In tandem, you would work with us (project with UN-World Food Programme and Royal Government of Bhutan) to design, test and deploy real-time streaming sensors in risk hotspots capable of detecting the passage of hazard cascades and near-live refining the magnitude/type of event using a pre-run set of model ensembles.
During the first year, as part of the MRes in Geospatial Data Science, you will learn skills in Google Earth Engine cloud computing to detect and quantify the far travelled sediment plumes generated by known hazard cascades (linked to a NERC Urgency project), testing different sources and types of imagery, for example Planet Labs and Sentinel 2. You will then run your tool forwards from the initial event to determine if the impact was short-lived, or, if the deposits become chronic issues for down valley.
Fusion of remote sensing data with hydraulic modelling to extend capabilities for predicting flow in river systems.
Project application reference GEO22_12
Hydraulic models are often used to predict downstream levels and inundation extents in river systems. Development and validation of these models can be costly, time-consuming, and typically require acquisition of high-resolution datasets that are spatially and temporally distributed. Recent technological advances offer the potential for hydraulic models to be driven and validated using a diverse range of information, which may be particularly beneficial for understanding hydrological processes in traditionally data scarce environments. This research will seek to harness advances in remote sensing capabilities to produce highly constrained river flow simulations.
This project will seek to:
(i) Develop a workflow whereby a 2D hydraulic model can be constructed and validated primarily using remotely sensed datasets acquired using UASs. The topographic and bathymetric conditions will be generated through application of structure from motion multiview stereo (SfM-MVS), and outputs will be calibrated/validated using UAS observations of water extent and 2D surface velocities. At this stage, observed river flows will act as the upstream boundary conditions for the model.
(ii) Alternate data sources will be used to represent the inflow boundary conditions. Rather than using in-situ flow rate observations, remotely sensed water levels will be used to inform the upstream inflow conditions. The source of this data will be UAS ranging sensors, and satellite altimeters (e.g. NASA SWOT Mission).
(iii) A range of virtual stations will be established using this approach that will facilitate predictions of hydrological fluxes and inundation extents using remotely sensed data. The outputs generated will be validated using UAS-borne methods outlined in (i).
Year-1 will involve the application of UAS-based SfM-MVS, image velocimetry, and hydraulic modeling to predict inundation extents in an experimental research catchment. This information will be used to understand how flow conditions are affected by river management techniques. In years 2-4, this skill-set will be used to extend the workflow to a range of environmental conditions and using a diverse range of source/validation data.
Suitable candidate will be familiar with concepts relating to hydrological modeling and/or hydraulic modeling. Experience of analysing river system processes using remote sensing systems and/or products. Should also be willing to travel and conduct fieldwork.
The University of Nottingham
The following projects are available for candidates who would like to apply to study at the University of Nottingham.
Plantlife – remote capture across platforms for wild plant conservation
Project application reference GEO22_05
Plantlife’s long term ambition is of significant grassland and rainforest creation and restoration for the benefit of creating more robust environmental landscapes to mitigate against Climate Change and biodiversity loss. To optimise management decisions in pursuit of creation and restoration, data on how habitats are responding to management is required. This PhD will design optimal data capture and visualisation strategies to provide the information required. In a true partnership between the PhD candidate and Plantlife a range of remote sensor types will be explored, including that from satellites, UAS and terrestrial platforms (e.g., dashcams) and passive and active citizen scientists.
During the first year, as part of the MRes in Geospatial Data Science, the project will work alongside The UK Green Infrastructure Partnership which aims to transform rural road verges into wildlife havens. There are nearly 313,500 miles of rural road verge in the UK which need monitoring. This is a monitoring challenge. However, there are 23 million people commuting to work by road every day, many of which will carry a dashcam. Many roadside verges are passively sensed using webcams. This MRes project will explore the use of web and dashcam technology to extract metrics relating to grass verge biodiversity properties (e.g., species mix, grass health) to establish the success of the road verges project.
This will lead into the PhD which will take the principle of using remote sensing (RS) technology to capture data on grassland biodiversity demonstrated in the MRes project and be scaled across other Plantlife projects.
Applicants should hold a minimum of a UK Honours degree at 2:1 level, or equivalent, in Physical Geography, Environmental/Natural Sciences or Computer/Mathematical/Statistical Sciences. Good conceptual and practical knowledge of remote sensing/GIS is desirable. Programming (ideally in R) and machine learning skills are assets. However, enthusiasm for nature and curiosity about the best ways to conserve it under environmental change using geospatial science are by far the most important requirements.
This project would be co-supervised by Plantlife.
Precise autonomous satellite navigation for high altitude and scientific missions
Project application reference GEO22_08
With more than 40 funded Lunar missions due to launch in the next 10 years and the number of earth orbiting satellite increasing rapidly, the need for precise and reliable autonomous navigation has never been greater. Current methods of navigation at high altitudes are costly and have limited accuracy, while scientific and crewed missions have increasing more stringent navigation and timing requirements.
Initial studies by NGI and others are showing the potential for GNSS to be used at Lunar altitudes, in many cases outperforming ground based tracking techniques. The addition of navigation sources in Lunar orbit are showing vast improvements in navigation performance enabling many novel scientific missions and autonomous rendezvous and landing. NGI have developed a simulation software for the analysis of spacecraft navigation performance with GNSS and other systems.
This research will build on the initial studies performed at NGI to define different augmentation systems and the potential for additional sensors to improve performance and enable autonomous navigation for Lunar orbits.
Applicants should have a minimum of a 2:1 undergraduate degree (or equivalent) in Computing, Engineering, Geography, Geomatics, Environmental Sciences, Mathematics and Statistics, or related disciplines.
Applicants would ideally also have experience of GNSS basics (concept, signal parameters, navigation using GNSS), Space systems knowledge (equipment, missions and payloads), Satellite orbit basics and Matlab or Python programming experience
This project will be co-supervised TO BE CONFIRMED
Optimising methodologies for remote sensing-based river management
Project application reference GEO22_13
River environments are among the most threatened ecosystems on the planet. Earth observation is increasingly the preferred method for river managers to capture data on environmental change within river environments. However, the sinuous, complex nature of rivers means that ‘conventional’ remote sensing data is often sub-optimal for the extraction of river habitat data. This PhD aims to develop best-practise methodologies for the generation of river habitat data from historical and new earth observation data that would normally be sub-optimal for river management purposes. This study will ultimately help generate data needed to improve management of threatened river environments.
During the first year, as part of the MRes you will focus on the development of methods to extract high quality river corridor topography from ‘consumer’-grade aerial photography (drones and piloted aircraft) acquired without ground control data. Structure from Motion photogrammetry will be applied to pre-acquired non-metric airborne imagery collected for purposes other than topographic retrieval. This project will examine whether this error can be corrected using a combination of polynomial surface fitting and LiDAR data fusion to generate accurate 3D models of river corridors.
This will then lead on to the PhD where methods developed during the MRes will be applied to historical aerial photography (eg. NERC CEDA portal) and new data acquired from drones to map spatio-temporal change in river environments over the last century. Alongside methods developed for retrieving riverscape topography (see MRes), the project will also develop new techniques for extracting accurate other metrics (eg. temperature) from available remote sensing data sources. Data generated will generate new insights into the ecological and physical status trajectories of key rivers across Europe and North America and provide infrastructural planners, river managers and the utilities sector with the data they need to manage rivers in a targeted, sustainable manner.
Candidates should have a minimum of a 2:1 undergratuate degree in Geography, Environmental/Natural Sciences, Engineering (in a related subdiscipline) or Computer/Mathematical Sciences. Theoretical and conceptual knowledge of remote sensing and coding (preferably Python or MATLAB) are desirable, as are experience in image processing and/or machine learning.