Geodaysit 2023

Laura Magnabosco


Sessions

06-12
17:30
15min
Evolution of the surface waters of the Po river in the 2020-22 period - a quantitative analysis of the drought effects with Sentinel-2 images
Carlo Masetto, Niccolo' Tolio, Eleonora Cagliero, Benedetta Gori, Umberto Trivelloni, Alessandra Amoroso, Laura Magnabosco

For Veneto Region, the year 2022 was affected by anomalies in terms of average temperature and recorded rainfall, compared to the climatic average of the last thirty years. The occurrence of these climatic conditions highlighted the inevitable negative effects for the environment; in particular, water resources were affected by this combination of climatic criticalities, both in terms of rivers’ flow rate and underground water flows, also due to the poor snow accumulations recorded in the Alps during the winter period.
Critical profiles in terms of surface runoff have been highlighted on various rivers and streams, particularly in the summer period. The main Italian river, the Po, was also affected by a water flow decrease in 2022, making evident the increase in the surfaces characterized by the presence of sand islands, visible along the river path.
This work, carried out by the Territorial Planning Department of the Veneto Region, aimed at a quantitative analysis of the surface covered by sand islands and surface water in a sufficiently representative area of the Po basin.
For the purposes of this study, the area included between the municipalities of Occhiobello (RO) and Ferrara was analysed (border area between Veneto and Emilia Romagna regions, where Po river flows). The analysis was carried out for the month of July 2022, comparing the data obtained with those relating to previous years (2020 and 2021).
In order to identify the islands of sand, multispectral Sentinel-2 satellite images were analysed, taking into consideration the wavelengths of the visible (B02, B03 and B04 bands), and that of the near infrared (B08 band). The area was then classified using supervised classification with the Random Forest classification algorithm, a methodology that allows to obtain high-precision classifications.
Considering the pixels’ size and the limits of the supervised classification, the precision of the analysis performed an accuracy higher than 95%. The analysis remarks relevance for monitoring the negative effects caused by the drought on the Po river. In the area under examination, a constant decrease in the surface of surface water was observed, and a corresponding increase of natural sand islands. From the analysis obtained, thanks to the use of the classifier, it is evident that the year 2022 was a year in which drought contributed to worsen the water stress for the Po river, with evident consequences on the environment in terms of availability of the water resource and the rise of the salt wedge near the river mouth. Moreover, the study presented here confirms the importance of using satellite data and classification tools for monitoring water bodies.

AIT Contribution
Sala Videoconferenza @ PoliBa