Spatiotemporal responses of water quality to riverine floods: Insights from a global dataset

Floods alter water quality, yet their long-term temporal dynamics and recovery trajectories remain poorly understood at the global scale. To address this, we integrated the latest global river water quality records with remote-sensing flood data spanning 845 flood events over 2000-2015 to examine water quality responses across various flood stages. Using a strict three-step site-screening approach, 395 stations worldwide were retained for analysis. Results show that Total Suspended Solids (TSS) increased by 24.3% during the flood period. Dissolved oxygen (DO) shows more pronounced changes in North America than South Asia, attributable to the denser monitoring network in North America that captured a greater proportion of sites with notable DO variations. Stage separation enables in-depth analysis of the deterioration and recovery processes in water quality. For instance, pH slightly decreased during flooding but roughly recovered to baseline within one month and slightly elevated thereafter (ΔpH>0.2). In contrast, nutrients including total nitrogen (TN) and total phosphorus (TP) exhibited minimal change during flooding but showed a sustained or modest increase within one month post-flood. Although these findings are still limited by data coverage and methodological resolution, long-term monitoring proves essential for capturing delayed post-flood risks in the environment and ecology that are frequently overlooked by conventional assessments.