A Spatiotemporal Radar-Based Precipitation Model for Water Level Prediction and Flood Forecasting

TL;DR

This work introduces STRPMr, a spatiotemporal radar-based precipitation model that predicts river water levels using RADOLAN radar data without upstream hydrological inputs. The model employs a (2+1)D CNN to extract spatiotemporal precipitation features and LSTM layers to forecast residual water-level changes $\Delta h_t$, reconstructing $\hat{h}_t = h_{t-\\Delta t} + \Delta h_t$. Validated on Goslar and transferred to Göttingen, STRPMr shows superior accuracy across 2–12 hour horizons (e.g., high BP and NSE, IoA near unity) and demonstrates strong capability to capture extreme flood events, highlighting its potential for regional, radar-based flood forecasting in data-scarce settings. By leveraging residual modeling and radar precipitation, the approach reduces dependency on upstream inputs and enables transfer learning to additional regions, with practical implications for timely flood warnings and risk management.

Abstract

Study Region: Goslar and Göttingen, Lower Saxony, Germany. Study Focus: In July 2017, the cities of Goslar and Göttingen experienced severe flood events characterized by short warning time of only 20 minutes, resulting in extensive regional flooding and significant damage. This highlights the critical need for a more reliable and timely flood forecasting system. This paper presents a comprehensive study on the impact of radar-based precipitation data on forecasting river water levels in Goslar. Additionally, the study examines how precipitation influences water level forecasts in Göttingen. The analysis integrates radar-derived spatiotemporal precipitation patterns with hydrological sensor data obtained from ground stations to evaluate the effectiveness of this approach in improving flood prediction capabilities. New Hydrological Insights for the Region: A key innovation in this paper is the use of residual-based modeling to address the non-linearity between precipitation images and water levels, leading to a Spatiotemporal Radar-based Precipitation Model with residuals (STRPMr). Unlike traditional hydrological models, our approach does not rely on upstream data, making it independent of additional hydrological inputs. This independence enhances its adaptability and allows for broader applicability in other regions with RADOLAN precipitation. The deep learning architecture integrates (2+1)D convolutional neural networks for spatial and temporal feature extraction with LSTM for timeseries forecasting. The results demonstrate the potential of the STRPMr for capturing extreme events and more accurate flood forecasting.

Figures (11)

• Figure 1: Location overview of Goslar in Lower Saxony, Germany GoslarSVG.
• Figure 2: Top: A sample radar precipitation image over Germany from extreme flood event on 26th July 2017 03:00 with a resolution of 1km $\times$ 1km. The zoomed-in box shows the data for the Goslar region. Below: Hourly precipitation intensity maps for Goslar, blue indicating lower and brown indicating higher precipitation intensities.
• Figure 3: Water levels (cm) captured by a hydrological sensor at Sennhuette (until red line: training dataset, until green line: validation dataset and last part: testing dataset.
• Figure 4: Water levels at Göttingen sensor station from 2003 to 2018.
• Figure 5: (2+1)D convolution architecture for a sample precipitation image from flood event on 26th July 2017 in Goslar..