A1449
Title: Assessing urban water loss patterns: A comparative time series study between Brazil and Portugal
Authors: Elisa Henning - Universidade do Estado de Santa Catarina (Brazil) [presenting]
Andreza Kalbusch - Universidade do Estado de Santa Catarina (Brazil)
Amilcar Oliveira - Center of Statistics and Applications of University of Lisbon and Universidade Aberta (Portugal)
Teresa Oliveira - Universidade Aberta (Portugal)
Abstract: Water losses in urban supply systems constitute a significant challenge in the context of climate change and increasing pressure on water resources. Understanding their temporal dynamics, key drivers, and cross-country differences is essential to support effective mitigation strategies. Statistical modeling provides a useful framework for this analysis. A comparative assessment of water losses in urban supply systems in Brazil and Portugal is conducted based on time series methodologies. Time series regression models were employed to characterize trends, identify structural differences in water loss behavior, and examine potential associations with per capita water consumption. Additionally, the Grey Model was applied for short-term forecasting, given its suitability for limited or incomplete data series. The results indicate divergent trajectories between the two countries, with a decreasing trend in water losses observed in Portugal and an increasing trend in Brazil. In contrast, per capita water consumption exhibits an upward trend in Portugal and relative stability in Brazil. These findings suggest that variations in water loss patterns are not directly attributable to per capita consumption levels, but are more likely driven by context-specific factors, such as infrastructure conditions, management practices, and regulatory frameworks. Furthermore, the Grey Model demonstrated high predictive accuracy, with forecast errors below 2\%.
