The effect of using multiple rain gauge stations and spatially X-band radar images as input into a calibrated hydrodynamic stormwater model : A case study in Åstorp municipality in southern Sweden
(2025) VVAM05 20242Chemical Engineering (M.Sc.Eng.)
- Abstract
- The number of flooding incidents in our urban cities and communities has increased drastically as a result of the increased effect of global warming in recent years. The probability of occurrence for heavy rainfalls and cloudburst events is higher, increasing the risks of pluvial flooding in urban areas. As a result of more frequent flooding, the hydraulic capacity of existing stormwater pipe systems is often overloaded and in need of maintenance and repair. The hydrological model provided by NSVA included the Åstorp municipality in southern Sweden and was calibrated using rainfall from the three rain gauge stations installed in the study area. In addition, the provided rainfall records from the three rain gauge stations were between... (More)
- The number of flooding incidents in our urban cities and communities has increased drastically as a result of the increased effect of global warming in recent years. The probability of occurrence for heavy rainfalls and cloudburst events is higher, increasing the risks of pluvial flooding in urban areas. As a result of more frequent flooding, the hydraulic capacity of existing stormwater pipe systems is often overloaded and in need of maintenance and repair. The hydrological model provided by NSVA included the Åstorp municipality in southern Sweden and was calibrated using rainfall from the three rain gauge stations installed in the study area. In addition, the provided rainfall records from the three rain gauge stations were between 06-16-2023 and 09-01-2023. It was found that August had the most accumulated rainfall depths and consisted of rain events of varying rainfall intensities and durations. Consequently, two measured rainfall periods were selected and used as the model simulation periods: 08-06 to 08-12 and 08-25 to 08-31. The main objective of this thesis project was to investigate the effect of using multiple rain gauges and spatially X-band radar images as input in a calibrated Mike Plus hydrodynamic stormwater model. The measured discharge was used as the reference flow and was obtained from five flow devices placed across the Åstorp model. This thesis project aims to answer three following stated research questions thoroughly:
1. To what extent is the modelled discharge affected when rainfall from rain gauge stations is used as input into a calibrated urban hydrological model?
2. To what extent does the X-band weather radar affect the modelled discharge when used as rainfall input into a calibrated urban hydrological model, compared to rainfall from rain gauge stations? How do the findings from the thesis project compare to other major findings from the literature review?
3. To what extent do different rainfall types affect the model simulation run time when used as rainfall input into a calibrated Åstorp hydrological model?
Four rainfall scenarios were investigated and defined using different spatial extent options in the Mike+ software. The first three scenarios (1 to 3) used rainfall data from the implemented rain gauge stations, and the last scenario (4) used high-resolution radar data from an existing X-band radar station in Helsingborg, a coastal city in southern Sweden. Objective functions and error indicators were used to evaluate the model performance.
The simulation results were shown to be inconsistent using data from multiple gauge stations and spatially X-band radar images. The finding from this thesis project has further shed light on the importance of filtering X-band radar data before applying it as rainfall input in the model.
Based on the final simulation results, a fundamental question needed to be answered as to how to define the calibrated runoff model and how spatially distributed rainfall data should be used concerning the calibration procedures. In conclusion, the study has demonstrated that more elaborate studies are needed in order to understand how spatially distributed rain data should be used in connection with any existing simulation software. However, any advanced radar merging techniques were outside the scope of this thesis and two future recommended techniques were recommended in order to obtain better simulation results, namely filtering of X-band radar rainfall and using bias correction factors to obtain an adjusted radar rainfall. (Less) - Popular Abstract (Swedish)
- I takt med ökad nederbörd och urbanisering ökar risken för pluviala översvämningar i våra urbana miljöer som följd av den ökade effekten av klimatförändringar. En allvarlig konsekvens som följd av alltmer kraftiga nederbörd och skyfall är att den hydrauliska kapaciteten i ett befintligt dagvattenledningssystem är överbelastad, vilket kan leda till översvämning.
Detta arbete syftade att belysa hur regn från tre regnmätarna och X-band väderradar påverka en kalibrerad urban hydrologisk modell (avrinningsmodell) där effektiviteten utvärderades utifrån prestationsmåtten Nash-Sutcliffe efficiency index (NSE) baserade på flödesdata från fem mätstationerna. Det kalibrerade hydrologiska modellen omfattade Åstorp kommun i Skåne, södra Sverige och... (More) - I takt med ökad nederbörd och urbanisering ökar risken för pluviala översvämningar i våra urbana miljöer som följd av den ökade effekten av klimatförändringar. En allvarlig konsekvens som följd av alltmer kraftiga nederbörd och skyfall är att den hydrauliska kapaciteten i ett befintligt dagvattenledningssystem är överbelastad, vilket kan leda till översvämning.
Detta arbete syftade att belysa hur regn från tre regnmätarna och X-band väderradar påverka en kalibrerad urban hydrologisk modell (avrinningsmodell) där effektiviteten utvärderades utifrån prestationsmåtten Nash-Sutcliffe efficiency index (NSE) baserade på flödesdata från fem mätstationerna. Det kalibrerade hydrologiska modellen omfattade Åstorp kommun i Skåne, södra Sverige och testades med rumsligt distribuerade regndata från perioder som inte är kalibreringsperiod. Två typer av nederbördsdatauppsättningar användes som indata i modellen, nämligen nederbörd från regnmätarna och X-band radarbilder.
Simuleringsresultat var inte entydiga med avseende på NSE prestationsmått när olika rumsligt distribuerade regndata användes som modellinput. Studien visade dock tydligt det inte bara är regndata som har betydelse för att förklara dagvattenavrinning över tid men även andra faktorer som påverkat simuleringsresultat. Det grundläggande avrinningsmodellen är välkänt men simuleringsprogrammet beskriver en komplex struktur och involverar intrikata kalibrerings-proceduren som även är tidsberoende. En grundläggande fråga som kvarstår i sammanhanget är begreppet kalibrerad hydrologisk modell och är utanför ramen för denna studie att utveckla detta i detalj, men har visat det är uppenbart att ytterligare studier behövs dock för att bättre förstå hur rumsligt distribuerade regndata bör användas i samband som indata i befintliga simuleringspaket. Dessutom även procedurer och dataförberedelser som används för att beskriva distribuerade regndata behöver utvecklas minst för regndata baserade på X-band väderradar.
Avslutningsvis går det att konstateras att simuleringstiden var avsevärt längre när X-band radarbilder användes som input till simuleringsprogrammet i jämförelse med regnmätarna. Detta tydde på ett högre krav med avseende på datorminnet och lagring för en kalibrerad avrinningsmodell att simulera regnhändelserna. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/9186640
- author
- Wang, Christopher LU
- supervisor
- organization
- course
- VVAM05 20242
- year
- 2025
- type
- H3 - Professional qualifications (4 Years - )
- subject
- keywords
- Water engineering, Hydrological urban stormwater modelling, Stormwater system, Urban drainage, X-band weather radar rainfall, Rain gauge rainfall, Calibrated hydrological model, Spatially distributed rainfall, model simulation time
- report number
- 2025-01
- language
- English
- id
- 9186640
- date added to LUP
- 2025-03-25 11:04:17
- date last changed
- 2025-03-25 11:04:17
@misc{9186640,
abstract = {{The number of flooding incidents in our urban cities and communities has increased drastically as a result of the increased effect of global warming in recent years. The probability of occurrence for heavy rainfalls and cloudburst events is higher, increasing the risks of pluvial flooding in urban areas. As a result of more frequent flooding, the hydraulic capacity of existing stormwater pipe systems is often overloaded and in need of maintenance and repair. The hydrological model provided by NSVA included the Åstorp municipality in southern Sweden and was calibrated using rainfall from the three rain gauge stations installed in the study area. In addition, the provided rainfall records from the three rain gauge stations were between 06-16-2023 and 09-01-2023. It was found that August had the most accumulated rainfall depths and consisted of rain events of varying rainfall intensities and durations. Consequently, two measured rainfall periods were selected and used as the model simulation periods: 08-06 to 08-12 and 08-25 to 08-31. The main objective of this thesis project was to investigate the effect of using multiple rain gauges and spatially X-band radar images as input in a calibrated Mike Plus hydrodynamic stormwater model. The measured discharge was used as the reference flow and was obtained from five flow devices placed across the Åstorp model. This thesis project aims to answer three following stated research questions thoroughly:
1. To what extent is the modelled discharge affected when rainfall from rain gauge stations is used as input into a calibrated urban hydrological model?
2. To what extent does the X-band weather radar affect the modelled discharge when used as rainfall input into a calibrated urban hydrological model, compared to rainfall from rain gauge stations? How do the findings from the thesis project compare to other major findings from the literature review?
3. To what extent do different rainfall types affect the model simulation run time when used as rainfall input into a calibrated Åstorp hydrological model?
Four rainfall scenarios were investigated and defined using different spatial extent options in the Mike+ software. The first three scenarios (1 to 3) used rainfall data from the implemented rain gauge stations, and the last scenario (4) used high-resolution radar data from an existing X-band radar station in Helsingborg, a coastal city in southern Sweden. Objective functions and error indicators were used to evaluate the model performance.
The simulation results were shown to be inconsistent using data from multiple gauge stations and spatially X-band radar images. The finding from this thesis project has further shed light on the importance of filtering X-band radar data before applying it as rainfall input in the model.
Based on the final simulation results, a fundamental question needed to be answered as to how to define the calibrated runoff model and how spatially distributed rainfall data should be used concerning the calibration procedures. In conclusion, the study has demonstrated that more elaborate studies are needed in order to understand how spatially distributed rain data should be used in connection with any existing simulation software. However, any advanced radar merging techniques were outside the scope of this thesis and two future recommended techniques were recommended in order to obtain better simulation results, namely filtering of X-band radar rainfall and using bias correction factors to obtain an adjusted radar rainfall.}},
author = {{Wang, Christopher}},
language = {{eng}},
note = {{Student Paper}},
title = {{The effect of using multiple rain gauge stations and spatially X-band radar images as input into a calibrated hydrodynamic stormwater model : A case study in Åstorp municipality in southern Sweden}},
year = {{2025}},
}