Machine Learning Scientist
MetDesk Ltd

- Improved execution speed of in house model by 76%, resulting in the quickest model to market in the sector and attracting multiple new clients due to this speed up
- Developed a deep-learning based algorithm for wind energy forecasting
- Developed plan and leading project for an improved solar and cloud forecasting method using live satellite data
- Pitched and now leading a project that would allow the company to expand to a worldwide clientbase without the need for any new hires
- Interacted with company leaders to help in the planning of new offerings in the upcoming years, as well as developing methods to improve efficiency of current company practices
- Helped improve understanding and usage of company hardware, leading to more efficient processes, including version control and a clean codebase
- Helping mentor and set reasonable targets for new intern

Postdoctoral Researcher
Lawrence Livermore National Laboratory

- Developed deep learning model to detect atmospheric rivers in meteorological data with world-leading performance
- Worked with a team to create a model to perform bias correction for rainfall for the in-house climate model
- Developed a system to employ deep learning to discover new physics to improve rainfall modelling via a new parametrization for deep convection
- Handled and preprocessed large volumes of climate data (>2TB)
- Helped out various colleagues to optimize their computational workflows, thus being able to produce results much quicker
- Organized various networking events for postdocs across different sections, as well as contributing to various decision-making initiatives
- Presented results in multiple international conferences

Computational Scientist
National Centre for Atmospheric Science

- Carried out data analysis to inform on the implementation of a new data storage method when using the UK Met Office climate model
- Started work to use physically-informed neural networks to be used for emulation of certain processes in a climate model, to improve the computational cost of the climate model, while making it quicker to execute

JAVA Software Engineer
LOQUS Business Intelligence

- Improved existing algorithms to produce more efficient solutions
- Carried data analysis on existing algorithms to identify any inefficiencies
- Helped transition the company from a local MySQL database to a cloud-based MongoDB database
- Presented team results to company executives and clients

PhD Computer Science
University of Reading

- Developed deep learning model using PyTorch to detect the presence of tropical cyclones in meteorological data
- Formulated and executed a method to use this model for a data reduction method, which was implemented in the UK Met Office’s climate model
- Performed data analysis to computationally optimize the data reduction method
- Developed skills in deep learning, working with large datasets, HPC, Linux, version control, Docker and Singularity containers
- Presented findings at an international conference
- Participated in the Young Entrepreneur Scheme where various skills, including working to a deadline and working with a team, were developed

MSc Atmosphere, Oceans and Climate
University of Reading

- Developed understanding of major physical processes controlling meteorological activity across the globe
- Developed physical modelling skills, e.g. using the finite difference method to model dispersion of a trace gas
- Developed skills in presenting complex data and ideas to a non-expert audience, as well as writing weekly progress reports

BSc Computational Physics
University of Malta

- Developed various skills in computational modelling using a variety of programming languages
- Learnt various computer science fundamentals including Linux, HPC and object-oriented programming
- Experienced and developed skills in implementing machine learning techniques, e.g. genetic algorithms
- Developed skills in mathematical understanding of differential equations and other constructs

UXNet: Joint U-Net and fully connected neural network to bias correct precipitation predictions from climate models
Shuang Yu, Indrasis Chakraborty, Gemma J. Anderson, Donald D. Lucas, Yannic Lops, Daniel Galea, Artificial Intelligence for the Earth Systems

Assessment of Storm-Associated Precipitation and its Extremes using Observations and Short-Range Climate Model Hindcasts
Wen-Ying Wu, Hsi-Yen Ma, David Conway Lafferty, Zhe Feng, Paul Ullrich, Qi Tang, Jean-Christophe Golaz, Daniel Galea, Hsiang-He Lee, Journal of Geophysical Research: Atmospheres

Investigating differences between Tropical Cyclone detection systems
Daniel Galea, Kevin Hodges, Bryan N. Lawrence, Artificial Intelligence for the Earth Systems

Deep Learning Image Segmentation for Atmospheric Rivers
Daniel Galea, Hsi-Yen Ma, Wen-Ying Wu, Daigo Kobayashi, Artificial Intelligence for the Earth Systems

TCDetect: A New Method of Detecting the Presence of Tropical Cyclones Using Deep Learning
Daniel Galea, Julian Kunkel, Bryan N. Lawrence, Artificial Intelligence for the Earth Systems

Meteorological data reduction for tropical cyclones using deep learning techniques
Daniel Galea

Intercomparison of deep learning model architectures for Atmospheric River prediction
Daniel Galea, Hsi-Yen Ma, Artificial Intelligence for the Earth Systems

Evaluating Probabilistic Deep Learning Methods for Uncertainty Quantification of Precipitation Bias Correction
Yannic Lops, Gemma J. Anderson, Indrasis Chakraborty, Donald D. Lucas, Daniel Galea, Shuang Yu, Artificial Intelligence for the Earth Systems

Evaluating Probabilistic Deep Learning Methods for Uncertainty Quantification of Temperature Downscaling
Yannic Lops, Gemma J. Anderson, Indrasis Chakraborty, Donald D. Lucas, Daniel Galea, Shuang Yu, Artificial Intelligence for the Earth Systems