Improving Flood Resilience through Data-Driven Safety Nets in Chad
Focus on research
Pierre Biscaye,
Chaire de Professeur Junior,
CERDI-UCA-CNRS-IRD
Chad is one of the most climate-vulnerable nations in the world. As temperatures climb and rainfall becomes more variable, the frequency of devastating floods has turned a seasonal threat into a recurring national crisis. This report analyzes fifteen years of flooding data to understand where the risk is greatest and where populations are facing growing dangers. By merging data on flooding from satellite imagery and household surveys, the research identifies critical gaps in current flood detection and adds to the evidence on the adverse household impacts of flood exposure. The findings are being used by the Chadian government and partners such as the World Bank to refine their national safety net programs, ensuring that emergency relief reaches the most vulnerable communities faster and more effectively.
A growing threat in a fragile context
Chad has long faced environmental challenges, but recent flooding marks a dangerous shift. Since 1980, nearly three-quarters of reported floods occurred in the last 20 years. What was once an infrequent threat is now an almost annual national disaster. The 2022 and 2024 floods were particularly catastrophic, displacing millions and causing massive economic losses.
While heavy local rainfall causes flooding in dry areas with low absorption capacity, the primary threat is fluvial floods. These floods occur as upstream precipitation causes rivers to overflow downstream, proving more damaging due to greater depths and high population density near river networks. Although rainfall is not increasing, climate change is making it more erratic, with frequent concentrated bursts intensifying flood risk.
But vulnerability is not just a matter of geography. While high-flood risk areas will only grow 6% by 2050, population growth will more than double those exposed—from 4.2 to 9.9 million people (Rogers et al. 2025). Chad’s young population is increasingly concentrated in urban areas like N’Djamena or southern agricultural regions. New settlements often sit in high-risk zones with little protection. Combined with limited drainage infrastructure and a reliance on subsistence farming, minor water increases can trigger disaster. In 2024, around 2 million people—around 10% of Chad’s population—were affected, a reminder that the "worst-case scenario" is the new normal.
Identifying flooding hotspots
One of the core challenges in managing flood risk is the fundamental need to understand exactly where water is and who it hits. Identifying "hotspots" of recurring floods is essential for targeting long-term social protection and infrastructure programs toward the most vulnerable regions.
On the ground, government emergency relief units and hydrometeorological agencies do collect detailed local data. However, these institutions face significant resource constraints in monitoring flooding across Chad's vast territory. Furthermore, they typically do not publish granular, high-frequency data on flood incidence, making it difficult to use these sources for real-time response planning.
Bridging the data gap with remote sensing
To bridge this information gap, this study utilizes the NOAA/George Mason University VIIRS Flood Mapping (VFM) archive, which hosts daily high-resolution global maps based on satellite imagery. Remote sensing enables granular, low-cost, and high-frequency flood detection nationwide. While powerful, satellite measures have limitations: optical imagery cannot see through clouds, small floods may be missed by vegetation, and algorithms can struggle to distinguish seasonal water from destructive flooding. Despite these constraints, satellite data effectively monitors areas where traditional reports are unavailable. In 2024, satellites detected flooding across roughly 6% of Chad’s land—the highest on record.
We complement the remotely-sensed flood measures with an analysis of flood reports in geolocated household surveys. These reports indicate where households were affected by floods, rather than just where flooding occurred. Analysis using panel data shows that flood exposure increases household non-farm enterprise activity—a common coping strategy—and suggests increased food insecurity.
In 2024 nearly half of all rural households across 80% of sample communities reported suffering from flood-related shocks. This emphasizes population concentration in high-risk areas and reveals where remote sensing may miss local impacts. On the other hand, satellite detection can be very precise: we find that analyzing the density of detected flooding in the surrounding area can identify over 80% of communities where households reported flood shocks in surveys.
Building an adaptive safety net
The ultimate goal of this research is to turn data into action. The findings are directly informing the World Bank’s Adaptive and Productive Safety Nets Project in Chad, part of the broader Sahel Adaptive Social Protection Program. This project, a partnerships with the Government of Chad, aims to provide rapid emergency cash transfers to households immediately after a flood is detected. Because remote sensing can provide immediate evidence of a disaster before detailed ground assessments are complete, it offers a powerful mechanism to "trigger" emergency support.
By identifying "hotspots" of recurring risk and methods for detecting flooding in near real-time, the report provides critical inputs to modernizing social protection, moving from reactive aid to a proactive, adaptive safety net. As climate change continues to affect Chad’s weather patterns and the population in areas at high risk of flooding grows rapidly, the ability to predict and respond to floods with precision could be the difference between a manageable shock and a humanitarian catastrophe.