arXiv subjects

James Chappell

Publications and source records attributed to James Chappell.

7 recordsLinked to original sources

arXiv2025

Technical overview and architecture of the FastNet Machine Learning weather prediction model, version 1.0

We present FastNet version 1.0, a data-driven medium range numerical weather prediction (NWP) model based on a Graph Neural Network architecture, developed jointly between the Alan Turing Institute and the Met Office. FastNet uses an encode-process-decode structure to produce deterministic global weather predictions out to 10 days. The architecture is independent of spatial resolution and we have trained models at 1$^{\circ}$ and 0.25$^{\circ}$ resolution, with a six hour time step. FastNet uses a multi-level mesh in the processor, which is able to capture both short-range and long-range patterns in the spatial structure of the atmosphere. The model is pre-trained on ECMWF's ERA5 reanalysis data and then fine-tuned on additional autoregressive rollout steps, which improves accuracy over longer time horizons. We evaluate the model performance at 1.5$^{\circ}$ resolution using 2022 as a hold-out year and compare with the Met Office Global Model, finding that FastNet surpasses the skill of the current Met Office Global Model NWP system using a variety of evaluation metrics on a number of atmospheric variables. Our results show that both our 1$^{\circ}$ and 0.25$^{\circ}$ FastNet models outperform the current Global Model and produce results with predictive skill approaching those of other data-driven models trained on 0.25$^{\circ}$ ERA5.

physics.ao-ph↗

arXiv2025

FastNet: Improving the physical consistency of machine-learning weather prediction models through loss function design

Machine learning weather prediction (MLWP) models have demonstrated remarkable potential in delivering accurate forecasts at significantly reduced computational cost compared to traditional numerical weather prediction (NWP) systems. However, challenges remain in ensuring the physical consistency of MLWP outputs, particularly in deterministic settings. This study presents FastNet, a graph neural network (GNN)-based global prediction model, and investigates the impact of alternative loss function designs on improving the physical realism of its forecasts. We explore three key modifications to the standard mean squared error (MSE) loss: (1) a modified spherical harmonic (MSH) loss that penalises spectral amplitude errors to reduce blurring and enhance small-scale structure retention; (2) inclusion of horizontal gradient terms in the loss to suppress non-physical artefacts; and (3) an alternative wind representation that decouples speed and direction to better capture extreme wind events. Results show that while the MSH and gradient-based losses \textit{alone} may slightly degrade RMSE scores, when trained in combination the model exhibits very similar MSE performance to an MSE-trained model while at the same time significantly improving spectral fidelity and physical consistency. The alternative wind representation further improves wind speed accuracy and reduces directional bias. Collectively, these findings highlight the importance of loss function design as a mechanism for embedding domain knowledge into MLWP models and advancing their operational readiness.

physics.ao-ph↗

arXiv2025

High brightness, symmetric electron bunch generation in a plasma wakefield accelerator via a radially-polarized plasma photocathode

The plasma photocathode has previously been proposed as a source of ultra-high-brightness electron bunches within plasma accelerators. Here, the scheme is extended by using a radially-polarized ionizing laser pulse to generate high-charge, high-brightness electron bunches with symmetric transverse emittance. Efficient start-to-end modelling of the scheme, from ionization and trapping until drive bunch depletion, enables a multi-objective Bayesian optimisation routine to be performed to understand the performance of the radially-polarized plasma photocathode, quantify the stability of the scheme, and explore the fundamental relation between the witness bunch charge and its emittance. Comparison of plasma photocathodes driven by radially- and linearly-polarized laser pulses show that the former yields higher brightness electron bunches when operating in the optimally-loaded regime.

physics.acc-ph↗

arXiv2023

Resonant excitation of plasma waves in a plasma channel

We demonstrate resonant excitation of a plasma wave by a train of short laser pulses guided in a pre-formed plasma channel, for parameters relevant to a plasma-modulated plasma accelerator (P-MoPA). We show experimentally that a train of $N \approx 10$ short pulses, of total energy $\sim 1$ J, can be guided through $110$ mm long plasma channels with on-axis densities in the range $10^{17} - 10^{18}$ cm$^{-3}$. The spectrum of the transmitted train is found to be strongly red-shifted when the plasma period is tuned to the intra-train pulse spacing. Numerical simulations are found to be in excellent agreement with the measurements and indicate that the resonantly excited plasma waves have an amplitude in the range $3$ - $10$ GV m$^{-1}$, corresponding to an accelerator stage energy gain of order $1$ GeV.

physics.acc-ph↗

arXiv2022

The AWAKE Run 2 programme and beyond

Plasma wakefield acceleration is a promising technology to reduce the size of particle accelerators. Use of high energy protons to drive wakefields in plasma has been demonstrated during Run 1 of the AWAKE programme at CERN. Protons of energy 400 GeV drove wakefields that accelerated electrons to 2 GeV in under 10 m of plasma. The AWAKE collaboration is now embarking on Run 2 with the main aims to demonstrate stable accelerating gradients of 0.5-1 GV/m, preserve emittance of the electron bunches during acceleration and develop plasma sources scalable to 100s of metres and beyond. By the end of Run 2, the AWAKE scheme should be able to provide electron beams for particle physics experiments and several possible experiments have already been evaluated. This article summarises the programme of AWAKE Run 2 and how it will be achieved as well as the possible application of the AWAKE scheme to novel particle physics experiments.

physics.acc-ph↗

arXiv2020

Controlled density-downramp injection in a beam-driven plasma wakefield accelerator

This paper describes the utilization of beam-driven plasma wakefield acceleration to implement a high-quality plasma cathode via density-downramp injection in a short injector stage at the FLASHForward facility at DESY. Electron beams with charge of up to 105 pC and energy spread of a few percent were accelerated by a tunable effective accelerating field of up to 2.7 GV/m. The plasma cathode was operated drift-free with very high injection efficiency. Sources of jitter, the emittance and divergence of the resulting beam were investigated and modeled, as were strategies for performance improvements that would further increase the wide-ranging applications for a plasma cathode with the demonstrated operational stability

physics.acc-ph↗

arXiv2019

A compact electron injector for the EIC based on plasma wakefields driven by the RHIC-EIC proton beam

Initial simulations investigating using the RHIC-EIC proton beam as the drive beam in a plasma wakefield acceleration experiment are presented. The proton beam enters the plasma and undergoes self-modulation, forming a series of microbunches. These microbunches resonantly drive electron density perturbations within the plasma, exciting a longitudinal electric field with accelerating gradients in excess of $\mathrm{GVm^{-1}}$. Injecting electrons into the resulting wakefield offers an efficient method for accelerating electron bunches for use in the proposed EIC collider.

physics.plasm-ph↗