Enhancing hydrological model performance during severe droughts: insights from the Po River basin

Water resource management has become an increasingly pressing challenge due to the rising frequency and severity of extreme events, such as droughts and floods, related to climate change. In this context, hydrological models are fundamental tools for understanding and predicting the behavior of watersheds and supporting adaptation and mitigation strategies. However, the ability of these models to accurately simulate flows during extreme events is often limited, especially in areas heavily influenced by anthropogenic activities. A new article examines the performance of the Continuum hydrological model in simulating the runoff of the Po River during droughts of varying severity, providing important insights for improving hydrological modeling in an era of climate change.?

The article titled "Hydrological model skills change with drought severity; insights from multi-variable evaluation" was published in the Journal of Hydrology (Volume 634, 2024) and is authored by our researchers Francesco Avanzi , Lorenzo Alfieri , Andrea Libertino , Simone Gabellani , together with Giulia Bruno and Doris Duethmann from the Leibniz Institute of Freshwater Ecology and Inland Fisheries (IGB) . It analyzes the ability of the Continuum hydrological model to simulate the runoff of the Po River, in Italy, during droughts of varying severity over the past thirteen years, including the severe event of 2022. The authors evaluated the model by comparing simulations of runoff (Q) with ground-based observations, as well as of evapotranspiration (ET) and terrestrial water storage (TWS) with benchmark data based on remote sensing observations.?

Methodology?

The Continuum model was calibrated using historical Q data from the study period and evaluated applying performance indicators such as the Kling-Gupta Efficiency (KGE) and Pearson Correlation Coefficient (r). The variables analyzed included:?

• Q: evaluated using the KGE, achieving a value of 0.81 for the basin outlet.?

• ET: obtained via remote sensing, with a correlation coefficient of 0.94.?

• TWS: also derived from remote sensing, with an r of 0.76.?

Results

The results showed that while the Continuum model adequately simulated runoff and the variables ET and TWS during wet periods and moderate droughts, its accuracy significantly decreased during the severe drought of 2022. Including moderate drought data in the calibration period did not significantly improve the model's performance during the severe drought of 2022.?

This finding suggests that hydrological conditions during extreme droughts may have unique characteristics not adequately represented by less extreme historical data. Additionally, the neglection of irrigation flows was identified as one of the main causes of the model's performance deterioration, suggesting the need to incorporate more detailed land use and agricultural practices data in future simulations. ?

This study provides valuable insights for improving hydrological modeling and offers a significant contribution to understand hydrological processes in human-influenced contexts, thus supporting management and adaptation strategies for new climate conditions.?

Find out more in the article: