Hackathon: Data Assimilation in Firedrake

Wednesday-Friday 8-10th April

Rutherford Appleton Laboratory

A hackathon to augment variational data assimilation capabilities in Firedrake.

Context

Data assimilation (DA) methods allow users to combine information from prior (numerical) models with measurement information, in order to update state estimates and initialise forecasts. Variational DA achieves this using gradient-based optimisation to minimise the misfit between the model state and the observed data. 
 
The need to differentiate the model solution process, as well as the observation operators, makes variational data assimilation challenging to implement. Consequently, much of the DA research is conducted using toy models that sacrifice realism for ease of implementation. Conversely, variational DA remains inaccessible to many simulation scientists as they lack the required derivative (adjoint) model code.
 
Firedrake is an automated forward and adjoint PDE simulation system. Firedrake users write high-level mathematical code for their PDE solver, and the resulting high-performance parallel implementation is generated and executed automatically. Further, the adjoint to a forward simulation written in Firedrake is available without further implementation work. This means that efficient forward and adjoint solvers can be created for any PDE and employing a huge range of finite element simulations, all at a tiny fraction of the development effort required using conventional software approaches.
 
This makes Firedrake a powerful tool for variational data assimilation. DA researchers can easily employ a wide range of PDEs and discretisations to test their work, while simulation scientists gain access to the gradients required to adopt variational DA. Recent advances in Firedrake’s DA capabilities include the features required for weak 4D variational DA, bringing more advanced approaches within reach of its users.

 

Goal

This hackathon is an opportunity for two-way exchange between the developers of Firedrake’s DA capabilities and current and potential users. The hackathon will incorporate:
  1. an introduction and demo to Firedrake and its adjoint capabilities aimed at new users from Josh Hope-Collins and David Ham.
  2. discussions to gather input from participants about relevant test problems, and about desired further developments.
  3. interactive sessions giving participants the opportunity to gain hands-on experience with the technology by implementing new test cases; and provide input to the VarDA-Firedrake project.

 

Audience

This hackathon is for anyone with an interest in variational DA, whether as a DA researcher or a simulation scientist who is a current or potential DA user. It will be suitable for existing Firedrake users and for those with no Firedrake experience, and for those at any professional stage.

Note: Oxford University is hosting a 2.5-day training course on solving partial differential equations with Firedrake, March 18-20 2026. Further information can be found here. 

 

Registration

All are welcome, and there is no registration fee for attendance, but registration is mandatory - registration DEADLINE: 25th March 2026. Please register using this registration form.
Please bring a valid ID (BRP, Passport, Driving Licence, etc) to allow you onto the RAL site.
 
We will go out for dinner in the early evening of Thursday. Attendees will pay for their meal. Please indicate on the form whether or not you'd like to join us when registering so we have a rough idea of the numbers.
 

Location

The postcode for the Rutherford Appleton Laboratory is OX11 0QX.

The nearest train station is Didcot, and the Thames travel X24/X34/X35 buses travel between Didcot Parkway Station and the Harwell Campus (~ 30 minute journey, with a bus every ~ 15 minutes). Taxis are also available at Didcot Parkway Station.

Detailed travel instructions are provided on the STFC "How to Find Us" page. The official website for the bus service can be used to plan your journey.

Talks will take place in the Visitor Centre on the main Rutherford Appleton Laboratory site. On arrival, please report to reception and they will direct you to the Visitor Centre. If driving, reception will direct you to a suitable place to park.

 

 

Hotels nearby

  • The Coseners House
    This hotel is in Abingdon.
    The X2 bus goes Oxford/Didcot to Abingdon.
    To go to RAL from Abingdon you need to take the 43 bus. The 43 bus has only two journeys in the morning from Abingdon to RAL and two journeys in the evening. 
  • Premier Inn Oxford South Didcot hotel
    This hotel is near Didcot.
    Take X2/X32/X36/33 from Didcot Parkway Railway to go to the hotel
    Take the ST1 bus from the nearby bus stop Turn (also called Milton Heights) to RAL and use the reverse direction on the way back
  • Ridgeway House
    This hotel is on the Harwell Campus, the campus that hosts RAL.
    Travel to RAL is explained here How to Find Us. Note that the ST1 bus service is limited to certain hours (last journey leaving Oxford is at 18:40). Please check the official website for more details.  
  • + many hotels in Oxford are also available. We recommend those being close to the High Street as they are closer to the ST1 bus stop.

Sponsors

We gratefully acknowledge the support of our sponsors, the Computational Science Centre for Research Communities, CoSeC, and the Scientific Computing Department.

Organisers

Links