Validering av nederbördsdata från VViS : Prestandaanalys ur ett larmsystemperspektiv
Year: 2025
Type: dissertation
Abstract: To reduce the impact of heavy precipitation on the road network, the Swedish Transport Administration is developing a weather alarm system that will use precipitation data from the weather information system VViS, among other sources. To increase the understanding of how precipitation data from VViS can be used in the weather alarm system, the VViS precipitation sensor MS7 was evaluated, both generally and under specific weather conditions, by comparing it with data from nearby SMHI stations and radar images. 38 stations were evaluated, divided into two groups, which were analyzed differently due to differences in available reference data. VViS underestimated the total amount of precipitation compared to SMHI. In the two groups, the VViS stations measured on average 34% and 31% less total precipitation compared to the SMHI stations. Eight VViS stations measured larger amounts of snow than SMHI, which was likely due to the SMHI stations being affected by wind. This was supported by the fact that VViS often measured more snow than SMHI during strong winds. During precipitation events with a maximum intensity of over 4 mm/h, the measurements at some station pairs agreed well, while others showed large differences. When precipitation events from all stations located adjacent to roads were analyzed together, R² = 0.34 for rain and R² = 0.36 for snow. Through comparison of radar images, it was concluded that the differences in precipitation amounts could not be explained by spatial variability. The differences in performance between the station pairs suggest that external factors, such as dirt or obstructions, affect some stations more than others. To increase the accuracy of the VViS measurements, extended management of these factors is recommended. One way to compensate for the general underestimation is to adjust the threshold values for the weather alarms, which should be set by considering the desired system sensitivity.
Author: Eriksson, Linnea
Language: English
Subfield: Atmospheric Science
Field: Earth and Planetary Sciences
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7082312164
