Introduction

During February 2026, Senegal experienced a particularly severe dust episode, which peaked in the middle of the month. This phenomenon was measured by the network of air quality microsensors deployed by Kaikai. Installed since 2025 in schools and various public infrastructures as part of Kaikai's air quality monitoring project, this system monitors fine particle concentrations at local and national levels. The following analysis is based on data collected between 1 February and 1 March 2026.

Spatial dynamics of the dust episode across Senegal

In order to obtain an overview of the event, several strategic sites were selected, ranging from Richard-Toll in the north to Ziguinchor and Bignona in the south, via the Dakar-Thiès axis.

Observation of the readings reveals an identical time sequence across the entire territory. After a moderate start to the month, a gradual increase was observed from 14 February onwards. This increase culminated in a widespread pollution peak on 16 and 17 February, before gradually subsiding. This perfect simultaneity between regions that are nevertheless far apart confirms that this was a nationwide dust episode.

Region-by-region analysis

North: Saint-Louis and Richard-Toll

In Saint-Louis (Goxu Mbacc 2 School), daily concentrations peaked at around 50 µg/m³ on 16 February, compared with values generally ranging between 15 and 35 µg/m³ at the beginning of the month. In Richard-Toll (Ndiangué Primary School), the peak was even more pronounced, with values reaching around 65 µg/m³ on 16 February. The monthly average there is around 20–21 µg/m³.

At both sites, the increase was visible from 15 February, confirming the arrival of the dust plume from the north of the country.

Central West: Thiès and Diourbel

There is a slight difference here compared to the northern sites: the maximum concentrations were measured between 17 and 18 February, suggesting that the dust wave moved from north to south.

West: Dakar

In Dakar (Elhadj Mbaye Diop School in Ouakam), daily concentrations generally hovered around 15 to 25 µg/m³ at the beginning of the month, then peaked at 50 µg/m³ on 17 February. At the second site observed in Dakar (Groupe Scolaire Asselar), the peak also occurred on 17 February, but with a daily average value exceeding 60 µg/m³.

The increase is clear and synchronised with the central-western regions.

South: Ziguinchor

In Ziguinchor (Lycée Djignabo), we observed a peak of 58 µg/m³ on 17 February, while at Cap Skiring Airport, we observed a peak of 50 µg/m³ on the same day.

Unlike the northern regions, the Ziguinchor area recorded its highest concentration levels between 17 and 18 February, before beginning to decline on 19 February. This time lag suggests that the dust moved from the centre to the south of the country. This hypothesis seems to be confirmed by the episode on 1 March: while the peak occurred on 27 and 28 February in northern regions such as Richard-Toll and Saint-Louis, it did not reach other localities until later.

Consistency with the ANACIM dust Bulletin

On 14 February 2026, the ANACIM published a Special Weather Bulletin announcing the arrival of a dense layer of dust that would gradually affect the territory from 14 February to 17 February at 12 noon. The bulletin specified that northern localities would be affected first, before the dust spread to the centre, west and south. The measurements taken by the Kaikai micro-sensor network are perfectly consistent with the official forecasts. The correlation between the weather alerts and the real-time measurements taken by the Kaikai network reinforces the credibility of the micro-sensor network in highlighting pollution phenomena at the national level.

Conclusion

The dust episode observed between 14 and 17 February 2026 is a case study illustrating how a meteorological event translates into measurable concentrations of fine particles at the local level.

Data from the micro-sensor network deployed by Kaikai show consistent dynamics from north to south, with peaks reaching approximately 70 µg/m³ locally and a gradual decline after 17 February.

This region-by-region analysis highlights the value of a distributed network in complementing our understanding of atmospheric dynamics and strengthening the culture of environmental data.