Spatio-temporal changes in urban water consumption during 2 years of the COVID-19 pandemic in southern Brazil

This study investigated the changes that occurred during the pandemic of COVID-19 in urban water consumption in residential, commercial, industrial, and public agencies in the city of Sao Leopoldo, southern Brazil, which has about 55,000 consumers and over 200,000 inhabitants. Overall, the city increased water consumption by 5.6% during the 2-year pandemic, with 5.9% in 2020 and 5.5% in 2021. Residential and industrial consumption increased by 6.77 and 9.92% in the first year, and by 5.47 and 14.45% in the second year, respectively. On the other hand, commercial and public sector consumption decreased, respectively, 5.48 and 46.26%, in the first year, and also 1.83 and 40.99% in the second year. In the first months of the pandemic, there was a sharp increase in residential water consumption at the same time as a reduction in consumption in the other categories. In contrast, there was a slight return to previous water consumption patterns in the following months. Overall, we can affirm that the more central neighborhoods presented higher changes in water consumption than the peripheral neighborhoods. In addition, the water consumption during the pandemic and pre-pandemic periods was statistically different for residential, industrial, and public consumers.

Influence of exogenous factors on water demand forecasting models during the COVID-19 period

Water scarcity has urged the need for adequate water demand forecasting to facilitate efficient planning of municipal infrastructure. However, the development of water consumption models is challenged by the rapid environmental and socio-economic changes, particularly during unforeseen events like the COVID-19 pandemic. This study investigated the impact of COVID-19 on the efficiency of water demand prediction models, considering the lockdown measures and various exogenous features, such as previous consumption (PC) and socio-demographic (SDF), seasonal (SF), and climatic (CF) factors. Multiple ensemble models, gradient-boosting machines (GBM), extreme-gradient-boosting (XGB), light-gradient-boosting, random forest (RF), and stack regressor (STK) were examined, compared to other machine-learning techniques, multiple -linear regression (MLR), decision trees, and neural networks. The models were tested using 3-year metering records for 128,000 consumers in Dubai. The feature importance analysis indicated that PC and SDF had a significant impact on consumption rates with correlation coefficients of 0.95 and 0.74, respectively, as opposed to SF and CF, which had negligible effect. The results showed that, before COVID, RF and STK outperformed other models with a coefficient-of-determination (R2) and root-mean-squared-error (RMSE) of 0.928 and 0.039, followed by XGB at 0.923 and 0.041, respectively. However, MLR achieved the highest prediction accuracy amid COVID with R2 and RMSE of 0.90 and 0.05, followed by GBM and XGB equally at 0.83 and 0.06, respectively. An ensemble-based error prediction model was applied, resulting in up to 9.2% improvement in predictions. Overall, this research emphasized the efficiency of ensemble models in handling fluctuating data with a high degree of nonlinearity.

Estimation of household water consumption pattern in a metropolitan area taking the impact of the COVID-19 pandemic.

A new approach for estimating the household water consumption pattern was developed by taking the impact of the COVID-19 pandemic using geographical data. Water consumption data for two years before and a year after the outbreak of the pandemic were analyzed to recognize the consumption pattern on annual and bi-monthly time scales as well as in different spatial classes. Following the recognition of the pattern, the spatiotemporal distribution of household water consumption was estimated based on the discovered connections between consumption and geographical variables. Once a regression relationship between consumption and population density was observed, an idea was developed to investigate the linear equations and their coefficient of parameters in water consumption groups from very low to very high classes using the training data. The coefficients were then adjusted to account for the pandemic's impact on the consumption pattern. Results showed that the highest increases in consumption were 11% for May-July due to the impact of the pandemic while the impact was from decreasing type during lockdowns. A pandemic-induced decline in the mean of consumption was linked to temporary migration by high-income families, whereas the water consumption of others faced an increase. The impact has also increased the slope of the linear relationship between the annual water consumption and population density increased by 3.5%. The proposed model estimated the annual water consumption with the accuracy of %3.77, %1.82, and %1.85 for two years before, one year before and one year after the pandemic, respectively.