Identifying the drivers of flood generation mechanisms and their seasonal variabilities.

Extreme rainfalls and floods have caused severe socio-economic and environmental losses in most parts of the world and are predicted to exacerbate due to the changing climate. Highly saturated soil, extreme rainfall, and heavy snowmelt are the most common flood triggers. However, the relative contributions of extreme rainfall, excess soil moisture, and snowmelt and how they are vary with time and change from catchment to catchment are not fully understood. This information is critical for a better understanding of flood generation mechanisms and can improve flood risk management plans and strategies. We examined precipitation, streamflow, and soil moisture at daily time scale from more than 671 hydrological stations across the country. Our main objectives were creating flood driving mechanisms according to hydrometeorological characteristics, identifying the contribution of independent variables (excess soil moisture, snowmelt, extreme rainfall) and understanding the spatial and temporal variability of mutual information.

The relative importance of each variable and its respective flood-generating processes were identified by combining the multilinear regression and ANOVA approaches. we confirm that most of peak flows are strongly associated with both the extreme rainfall (67%) and excess soil moisture (26%) conditions. There is a clear difference in flood magnitude and their respective generating mechanisms between regions, and regions with an expected decrease in soil moisture into the future were highly statistically correlated with a decrease in annual average peak flood magnitude. The role of extreme rainfall is the most dominant factor across the UK; however, seasonal total rainfall is not a strong influencing factor of peak floods in the southern UK. Extreme rainfall and peak floods are positively corrected with catchment drainage area. This linkage between drainage area and the most common flood generation mechanisms is crucial to quantifying the magnitude and level of flood risk in ungauged catchments.