Laboratory of Hydrological forecasting by Artificial Intelligence – Hydr.IA

The LabCom project associating the Hydrosciences Montpellier (HSM) research unit and the SYNAPSE company aims at developing a suite of hydrometeorological forecasting services based on artificial intelligence (AI) techniques. HSM is the leader research unit in the French environment for using artificial intelligence in hydrology whereas SYNAPSE is a SME specialized in the concentration and provision of hydrological data online. The range of the considered services brings solutions for the cases without any existing solution that would allow the users to anticipate phenomena to protect themselves better by any means. AI, by decreasing the costs of field studies, will enable providing reliable and affordable services for all the private and public water managers in order to significantly decrease the costs of flooding, which is the more impacting natural hazard. Indeed, the losses related to flooding are significant, first regarding human lives (sometimes tens of fatalities for one event, like Aude 1999, Gard 2002, Var 2010, French Atlantic coast 2010, Aude 2018 or recently in the Alpes Maritimes area) and psychological impact. To these fatalities, economic losses for companies and direct and indirect costs (insured or infrastructures) must be added. These costs reach on average 5.5 billion euros per year in the European Union and could exceed 50 to 80 billion euros per year in 2080 due to both socio-economic and climate changes (Rojas et al., 2013). In France, they reach 650 to 800 million euros and about one in four inhabitants and one in three employs are potentially exposed to this risk (DGPR, 2014).
This project relies on a specifically French statement: despite the increasing development of AI related solutions in many complex fields, like health, production engineering or aeronautics, and the demonstration of the relevance of these solutions for hydrologic and hydrogeological issues, a gap remains in the transfer of these technologies to end users. Among them, at the forefront, local authorities are the main prospects of SYNAPSE for hydro(geo)logical forecasting. It appears that two items are missing for a more massive transfer of these solutions to the market. The first is the difficulty, for research, to directly confront itself to the real operational conditions, that comes with all the legitimate, accurate and specific requests of the users. The second is a relative lack of confidence in solutions often seen as obscure and whose operational implementation is rare, inducing a vicious cycle that this project aims at breaking.
The working programme considered by this project is formatted to answer the following goals:
1) Definition of the range of services of real time forecasting and expected associated features, corresponding to a need of risk impact reduction and with market acceptance.
2) Realization of a demonstration online platform replaying past flood events to bring the proof of the AI efficiency to end users.
3) Development of methodologies of AI modelling design adapted to the range of services previously defined.
4) Lead time extension, so far limited to the response time of the hydrosystem, by coupling AI modelling with meteorological forecast that could be post-treated.