The Soil Water Index, or SWI, is a hydrological indicator used to describe soil moisture conditions. In the French CatNat drought framework, it is one of the elements used to assess drought conditions, alongside other criteria such as the presence of clay in the municipality and the request submitted by the mayor.

CatNat drought recognition requests correspond to an administrative and insurance-related signal. Here, they are analysed over the 1988–2025 period. They depend on observed damage, municipal procedures, regulatory criteria and recognition decisions.

These two pieces of information are therefore not independent. But they do not measure the same thing. One describes a hydrological constraint; the other describes the administrative and territorial translation of a damaging phenomenon.

The map crosses two axes.

Horizontally, it represents the growth of CatNat drought recognition requests. Toward the right, this growth is stronger.

Vertically, it represents the evolution of the hydrological signal. Toward the top, drying measured through the SWI is stronger.

Each grid cell is therefore classified according to a combination of these two dimensions. Darker colours correspond to areas where both signals are strong. More contrasted colours highlight situations of divergence.

Class 4-4 groups the grid cells where CatNat drought requests show strong growth and where the SWI signal also indicates marked drying.

This is the most intuitive convergence case: the hydrological signal and the administrative signal evolve in the same direction.

However, the geography of this class shows that clay shrink–swell risk cannot be reduced to a simple “dry south versus humid north” pattern. Significant pockets also appear in central-eastern France, the Rhône corridor, Burgundy, the Grand Est region and some Alpine areas.

Class 4-1 corresponds to territories where CatNat drought requests increase strongly, while the SWI signal does not show an equivalent intensification of drying.

Many grid cells appear in western France, including Brittany, Pays de la Loire, Normandy, the Atlantic façade and the Mediterranean coast.

This configuration shows that the insurance-related signal cannot be read as a simple direct reflection of the hydrological signal. CatNat requests also depend on the presence of clay, building exposure, construction vulnerability, reporting practices, municipal decisions and the recognition procedure itself.

Class 1-4 shows the opposite situation to class 4-1.

The SWI signal indicates marked drying, but the growth of CatNat drought requests remains relatively low.

Several factors may explain this divergence: less built-up territories, soils less sensitive to clay shrink–swell, less vulnerable buildings, lower reporting of damage, or a time lag between hydrological stress and administrative recognition.

These areas should therefore not be interpreted too quickly as “resilient” territories. They mainly indicate sectors where the relationship between physical drought and the CatNat signal deserves finer local analysis.

The main value of this map is that it reveals several territorial configurations: areas where the hydrological signal and the CatNat signal evolve together, and others where they clearly diverge.

If the CatNat signal were driven only by SWI-measured drying, divergence classes would be much less visible. Instead, they form coherent geographical patterns.

This suggests that clay shrink–swell should be read as a territorial risk, not only as a climatic phenomenon. Local geology, building exposure, construction vulnerability, reporting practices and the recognition procedure may also play a role.

This map does not predict future damage and does not demonstrate direct causality. The SWI is a modelled hydrological indicator. CatNat requests are an administrative signal, dependent on observed damage, municipal procedures and recognition criteria.

The map classes are relative. They allow territories to be compared with each other, but they should not be interpreted as physical, regulatory or insurance thresholds.

This analysis is based on two main sources over the 1988–2025 period: the soil moisture indicator produced by Météo-France for the French CatNat drought framework, and the CatNat drought recognition data published by the Caisse Centrale de Réassurance (CCR).

These sources do not directly measure the same thing: one describes a modelled hydrological state, while the other documents an administrative and insurance-related recognition process. The map should therefore be read as a spatial exploration tool, not as direct causal proof.

These articles extend this reading: understanding clay soils, understanding what InSAR actually measures, then comparing the 2026 RGA map update.

Understanding clay soils and shrink–swell

Understanding InSAR and ground motion

2026 clay shrink–swell map: geology has not changed, risk has