AquaCast: Urban water dynamics forecasting with precipitation-informed multi-input transformer

This work addresses the challenge of forecasting urban water dynamics by developing a multi-input, multi-output deep learning model that incorporates both endogenous variables (e.g., water height or discharge) and exogenous factors (e.g., precipitation history and forecast reports). Unlike conventional forecasting, the proposed model, AquaCast, captures both inter-variable and temporal dependencies across all inputs, while focusing forecast solely on endogenous variables. Exogenous inputs are fused via an embedding layer, eliminating the need to forecast them and enabling the model to attend to their short-term influences more effectively. We evaluate our approach on the LausanneCity dataset, which includes measurements from four urban drainage sensors, and achieve up to 60.6 % reduction in MSE and 30.4 % improvement in MAE compared to state-of-the-art baseline for one-day-ahead forecasting when exogenous variables are included. To assess generalization and scalability, we additionally test the model on three large-scale synthesized datasets, generated from MeteoSwiss records, the Lorenz Attractors model, and the Random Fields model, each representing a different level of temporal complexity across 100 nodes. The results confirm that our model consistently outperforms existing baselines and maintains a robust and accurate forecast across both real and synthetic datasets. • Multi-variable forecasting of urban water dynamics. • Real-life urban drainage data from Lausanne city is analyzed and structured. • Transformer that fuses precipitation history with water dynamic time series. • Enhanced predictions by integrating the precipitation forecast. • Scalability evaluated on synthetic datasets with varying structural complexity.