Have you ever seen the rain? Observing a record summer cloudburst with multiple radars and opportunistic sensors

National-scale precipitation observations in Sweden have traditionally relied on a combination of weather stations and C-band weather radar networks. These observations provide good spatiotemporal coverage and accurate quantification of most stratiform precipitation events across large areas. In the urban context, however, their resolution may be insufficient to capture critical rainfall variations. This limitation is particularly evident for convective rainfall, which is often highly localized (e.g., cloudbursts), and capable of causing severe damage to infrastructure. In light of this, the Swedish Meteorological and Hydrological Institute (SMHI) is exploring complementary ways to monitor rainfall in urban environments.

This study evaluates data from an X-band weather radar (XWR), Commercial Microwave Links (CML), and Private Weather Stations (PWS) to observe a cloudburst event that hit the Bjärehalvön peninsula in in southwestern Sweden in August 2022. The observations are bench-marked with the official monitoring network: SMHI’s weather stations and a C-band radar composite. A maximum volume of 75 mm in 1 hour was reported by a weather station operated by Båstad municipality. This station showed good agreement with long-term observations (2 years) from the nearest SMHI gauge (9 km away) and matched well with the XWR’s measurements of the event. High-resolution (sub-km and 1-minute) XWR data were used to evaluate precipitation variations along a 4.5 km long CML reach, suggesting new potential for correction of CML observations. Additionally, we propose methods for pre-processing of CML and PWS data to ensure consistent precipitation estimates and facilitate cross-referencing with the other sensors.

The results suggest that complementary sensors can add important data on rainfall intensity and volume, enhancing SMHI’s ability to monitor localized heavy rainfall events. However, the opportunistic sensors (CML and PWS) appear to have reached a maximum detectable intensity of rainfall during the event, which likely caused an underestimation of the total volume. Further, the findings highlight the challenge of estimating return periods of convective storms, as the return period varies significantly depending on which sensor that is chosen as the ground truth for the event.