Understanding the urban atmosphere through conceptual modelling and opportunistic sensing
This thesis explores the urban climate, with a focus on urban wind speed, from two distinct approaches. The first approach is the use of crowdsourcing, or opportunistic sensing techniques, to collect and process vast quantities of urban meteorological data. The second approach is to use (conceptual) physical modelling to gain a better understanding of wind differences between city and countryside.Chapter 2 expands upon a previously developed technique to derive urban air temperatures from smartphone battery temperatures, studying finer spatial and temporal scales than ever before. Over 10 million smartphone temperature records for the city of São Paulo, Brazil, are combined to derive daily and even hourly averaged urban air temperatures. Optimal results are achieved for 700 or more retrievals aggregated into daily or hourly temperature values. Daily temperature estimates are good (coefficient of determination of 86 \%), and temperature differences between Local Climate Zones (LCZs) can be distinguished at this scale. Hourly estimations of air temperature require a correction through a diurnally varying parameter in the used heat transfer model. The results show the value of smartphones as a measuring platform when routine observations are lacking.Chapter 3 makes use of Personal Weather Station (PWS) data to investigate their use for urban wind research. While the potential of PWS data for rain and temperature has been established, PWS wind data remained unused because of its high risk of error. We use crowdsourced wind speed observations from 60 Netatmo-brand PWSs in Amsterdam, the Netherlands, to analyse wind speed distributions of different LCZs. In a field test against a reference station, the Netatmo PWS anemometer appears to systematically underestimate the wind speed, and episodes with rain or a high relative humidity deteriorate the measurement quality. Therefore, we developed a quality assurance (QA) protocol to correct PWS measurements for these errors. The applied QA protocol strongly improves PWS data to a point where they can be used to infer the probability density distribution of wind speed of a city or a neighbourhood. The PWS performs poorly for periods with very low wind-speed; however, results for a year-long climatology of wind speed are satisfactory.In Chapter 4 the aforementioned crowdsourcing techniques are combined, along with Commercial Microwave Link (CML) data, to study meteorological phenomena. During a 17-day summer period in Amsterdam the passage of a cold front, followed by a warm episode, is monitored using the opportunistic sensing techniques. Measurements of temperature (from smartphone and PWSs); wind speed (PWSs); precipitation (PWSs and CML); air pressure (smartphone and PWS); radiation (smartphones) and humidity (PWSs) are compared to an established reference network. While the opportunistic sensing data have large uncertainties, both the cold front and the Urban Heat Island (UHI) during the hot period are successfully captured by these innovative techniques, showcasing their combined potential for urban (hydro-)meteorological monitoring.Chapter 5 introduces the Urban Wind Island (UWI): a positive wind difference of around 0.5 m/s between city and countryside. We research urban--rural wind differences with mixed-layer model, a bulk representation of the urban and rural boundary layers. In the model, these two distinctly surfaces are not connected, but experience the same large-scale atmospheric influences. The UWI appears to be caused by a combination of differences in boundary-layer dynamics between city and countryside. Oscillation of the wind around the geostrophic equilibrium can cause these UWI episodes. Sensitivity of the UWI to urban morphology is researched by implementing the 10 urban LCZs. The ideal circumstances for UWI formation seem to be moderate wind speeds (around 5 m/s), low building heights (up to 12 metres) and a deeper initial urban boundary layer.The UWI exploration is expanded upon in Chapter 6, which uses the mesoscale model WRF at 500m resolution to investigate the characteristics of the UWI over 2 consecutive summers in Amsterdam, in a more realistic model setting than the previous chapter. Large scale influences that could influence the wind field, such as frontal passages and strong precipitation events, are filtered out to focus on surface-induced wind changes. In this more realistic modelling environment, the data show that the UWI is present roughly half of the time, with a very similar magnitude as previously found (in the order of 0.5 m/s). The formation of the daytime UWI proceeds very similar to that of the theoretical concept, but a nocturnal UWI also forms, as a result of a delayed transitions towards a stable nocturnal urban boundary layer. While the relation between UWI and the initial conditions of the boundary layer (depth, geostrophic wind speed and temperature) are less clear, the UWI is consistently present in the data with a similar magnitude and timing as the conceptual model.The main findings of this thesis are that opportunistic sensing and crowdsourced data form a valuable addition to urban meteorological measurements, though care needs to be taken to account for the large errors present in these data and associated techniques. Secondly, that average wind in urban areas can be faster than over the countryside under certain conditions: this UWI can form both during the day, through different boundary-layer growth dynamics; and during night-time, related to the delayed collapse of the convective boundary layer. Both unconventional observation techniques as well as conceptual modelling can contribute to understanding the urban atmosphere.
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Doctoral thesis biblioteca |
| Language: | English |
| Published: |
Wageningen University
|
| Subjects: | Life Science, |
| Online Access: | https://research.wur.nl/en/publications/understanding-the-urban-atmosphere-through-conceptual-modelling-a |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This thesis explores the urban climate, with a focus on urban wind speed, from two distinct approaches. The first approach is the use of crowdsourcing, or opportunistic sensing techniques, to collect and process vast quantities of urban meteorological data. The second approach is to use (conceptual) physical modelling to gain a better understanding of wind differences between city and countryside.Chapter 2 expands upon a previously developed technique to derive urban air temperatures from smartphone battery temperatures, studying finer spatial and temporal scales than ever before. Over 10 million smartphone temperature records for the city of São Paulo, Brazil, are combined to derive daily and even hourly averaged urban air temperatures. Optimal results are achieved for 700 or more retrievals aggregated into daily or hourly temperature values. Daily temperature estimates are good (coefficient of determination of 86 \%), and temperature differences between Local Climate Zones (LCZs) can be distinguished at this scale. Hourly estimations of air temperature require a correction through a diurnally varying parameter in the used heat transfer model. The results show the value of smartphones as a measuring platform when routine observations are lacking.Chapter 3 makes use of Personal Weather Station (PWS) data to investigate their use for urban wind research. While the potential of PWS data for rain and temperature has been established, PWS wind data remained unused because of its high risk of error. We use crowdsourced wind speed observations from 60 Netatmo-brand PWSs in Amsterdam, the Netherlands, to analyse wind speed distributions of different LCZs. In a field test against a reference station, the Netatmo PWS anemometer appears to systematically underestimate the wind speed, and episodes with rain or a high relative humidity deteriorate the measurement quality. Therefore, we developed a quality assurance (QA) protocol to correct PWS measurements for these errors. The applied QA protocol strongly improves PWS data to a point where they can be used to infer the probability density distribution of wind speed of a city or a neighbourhood. The PWS performs poorly for periods with very low wind-speed; however, results for a year-long climatology of wind speed are satisfactory.In Chapter 4 the aforementioned crowdsourcing techniques are combined, along with Commercial Microwave Link (CML) data, to study meteorological phenomena. During a 17-day summer period in Amsterdam the passage of a cold front, followed by a warm episode, is monitored using the opportunistic sensing techniques. Measurements of temperature (from smartphone and PWSs); wind speed (PWSs); precipitation (PWSs and CML); air pressure (smartphone and PWS); radiation (smartphones) and humidity (PWSs) are compared to an established reference network. While the opportunistic sensing data have large uncertainties, both the cold front and the Urban Heat Island (UHI) during the hot period are successfully captured by these innovative techniques, showcasing their combined potential for urban (hydro-)meteorological monitoring.Chapter 5 introduces the Urban Wind Island (UWI): a positive wind difference of around 0.5 m/s between city and countryside. We research urban--rural wind differences with mixed-layer model, a bulk representation of the urban and rural boundary layers. In the model, these two distinctly surfaces are not connected, but experience the same large-scale atmospheric influences. The UWI appears to be caused by a combination of differences in boundary-layer dynamics between city and countryside. Oscillation of the wind around the geostrophic equilibrium can cause these UWI episodes. Sensitivity of the UWI to urban morphology is researched by implementing the 10 urban LCZs. The ideal circumstances for UWI formation seem to be moderate wind speeds (around 5 m/s), low building heights (up to 12 metres) and a deeper initial urban boundary layer.The UWI exploration is expanded upon in Chapter 6, which uses the mesoscale model WRF at 500m resolution to investigate the characteristics of the UWI over 2 consecutive summers in Amsterdam, in a more realistic model setting than the previous chapter. Large scale influences that could influence the wind field, such as frontal passages and strong precipitation events, are filtered out to focus on surface-induced wind changes. In this more realistic modelling environment, the data show that the UWI is present roughly half of the time, with a very similar magnitude as previously found (in the order of 0.5 m/s). The formation of the daytime UWI proceeds very similar to that of the theoretical concept, but a nocturnal UWI also forms, as a result of a delayed transitions towards a stable nocturnal urban boundary layer. While the relation between UWI and the initial conditions of the boundary layer (depth, geostrophic wind speed and temperature) are less clear, the UWI is consistently present in the data with a similar magnitude and timing as the conceptual model.The main findings of this thesis are that opportunistic sensing and crowdsourced data form a valuable addition to urban meteorological measurements, though care needs to be taken to account for the large errors present in these data and associated techniques. Secondly, that average wind in urban areas can be faster than over the countryside under certain conditions: this UWI can form both during the day, through different boundary-layer growth dynamics; and during night-time, related to the delayed collapse of the convective boundary layer. Both unconventional observation techniques as well as conceptual modelling can contribute to understanding the urban atmosphere. |
|---|