People & Contacts

DassFlow (Data Assimilation for Free Surface Flows) is a computational software originally developped at INSA Toulouse / IMT (from a former version at IMAG Grenoble) is now developed in a few institutes / corp.: INRAe, INSA-IMT, CS Group (see the complete list of collaborators below).

Your contacts

  • Concerning the mathematical models, inversions capabilities and algorithms fundamentals, you may contact J. Monnier (Prof. INSA Toulouse).
  • Concerning hydrology and flood dynamics, contact P.-A. Garambois (INRAe).
  • Concerning the H2iVDI algorithm (Hybrid Hierarchical Variational Discharge Identification), spatial hydrology, contact K. Larnier or J. Monnier.
  • Concerning other models and applications (DassFlow-Py version): glaciology, lavas, etc, contact J. Monnier.
  • Concerning the installation, codes programming and knowledge, you may contact K. Larnier (CS group - CNES funds) or L. Villenave (INRAe).

Current contributors (2022):
K. Larnier, Eng. CS group (developements, validations, datasets; since 2017), P.-A. Garambois, Res. INRAE (hydrology, hydraulics, applications; since 2017), L. Pujol, Eng./PhD ICUBE Strasbourg (coupling, wrapping, applications; since 2018), L. Villenave, Eng./PhD INRAe (wrapping, applications; since 2021), J. Monnier, Prof. INSA Toulouse (equations, methods, hybrid A.I., architecture; since... long).

The GR4 hydrological model which has been integrated into DassFlow has been developed by L. Santos and G. Thirel from INRAe, HYCAR research Unit.

Current applications & significant results

River flows modeling 1D St-Venant, 2D shallow-water, low-complexity (0.5D); coupling with rain-fall models.
The codes are written in Fortran (MPI for 2D model) and Python.
Features: Inference of the river properties (discharge, bathymetry, friction) from available observations.

Significant result: the HiVDI algorithm; discharge and bathymetry estimations from altimetry data (e.g. from the forthcoming SWOT mission). See e.g. poster 1 and poster 2 from AGU dec. 2018.
See also the SWOT mission website (NASA-CNES).

Ice flows modeling Classical 2D shallow ice flows models (xSIA, SSA), original ones (RU-SIA, SMISA) and 3D ones (Stokes). Power-law fluids.
The codes (written in Fortran and Python) are experimental codes aiming at assessing new asymptotical models, numerical schemes and original inversion methods.
These computational codes are interfaced with public ice-sheet data sets: altimetry, InSAR, bedmaps, climatic source terms.

Significant result (INSA-IMT): estimations of the bed topography elevation inland East Antarctica Ice Sheets (EAIS) areas from surface satellite data (altimetry, InSAR). See e.g. SWOT mission website (NASA-CNES).
A few know-hows of the VDA methodology have been transfered to ISSM team developers (UCI - Nasa JPL, USA) too.

Complex flows modeling e.g. lavas (Herschel-Buckley type rheology). 2D shallow (lubrication type model and SW type model) are designed for "virtual rheometry" (identification of rheology parameters). Joint work with Univ. Canterbury, New-Zealand (Marsden project). In progress.

Greetings to the great former contributors, developers of DassFlow !

  • Leo Pujol, PhD ICUBE Strasbourg-IMT-CNES. (Amazing update! 2D-1D integrated version, hydrology coupling, wrapping, applications) (2018-2022)
  • Thibault Malou, PhD IMT-INSA CLS corp. (covariance matrices, diffusive wave equations adapted tp spatial observations) (2019-2021).
  • David Kibe, PhD U. Christchurch NZ - INSA IMT (Herschel-Buckley rheology term (2021).
  • Shenuyan MA and Ngo Truyen Huynh, INSA students, wrapping of the 2D version (2020-21).
  • J. Zhu, postdoc INSA-IMT (CNES funds), DassFlow-Py including deep learning (2018).
    Jiamin is R&D Eng. in optimal control, I.A. in a private company.
  • Jonas Verley INSA-IMT (CNES funds) (2017). Salut l'artiste !
  • Pierre Brisset, Eng. INSA-IMT (CNES funds), DassFlow 1D, 1st main architect (2014-16).
    Pierre is Res. eng. in computational sciences (HPC, inverse methods) in a private company.
  • Ronan Madec, Res. Eng. IMT Toulouse and ANR AMAC. Adjoint MPI codes, DassFlow2D & 3D FE version (2010-13).
    Ronan is Res. eng. in computational sciences (HPC, inverse methods) in a private company.
  • F. Couderc, Res. Eng. CNRS IMT & INSA Toulouse. MPI features in DassFlow 2D, Perl scripts (2010-12).
  • J.-P. Vila, Prof. INSA Toulouse (2010-12). Finite Volume schemes order 2 (MUSCL) for 2D SWE (2010-12).
  • N. Martin, IMT & INSA Toulouse (PhD 2013). Adjoint code / gradient accuracy, FE solvers (2010-13).
    Nathan became R&D Eng. in computational sciences (FE software) in a private company.
  • J. Marin, INRIA Grenoble (Res. engineer 2007). 2nd main architect, DassFlow2D (2006-08).
    Joel is Res. eng. in computational sciences in a private company.
  • I. Gejadze, INRIA, lab. LJK (postdoc 2006). coupled version 1D-2D.
    Igor is Senior Researcher at IRSTEA Montpellier (SIC2 hydraulic model).
  • X. Lai, INRIA, lab. LJK (postdoc 2005) 1st real test cases with image data (version non maintained).
    Xijun is researcher at NIGLAS, Nanking, China
  • M. Honnorat, PhD INPGrenoble & INRIA, lab. LJK (PhD 2007). 1st main architect, DassFlow2D (2004-06).
    Marc is Res. eng. in computational sciences (sometimes as a private consultant).

How to obtain DassFlow source code ?

Please consult the webpage "Download".

How to cite ?

Please consult personal webpages of collaborators for publications references.
The PI personal references page contain some of these publications (but not all).