City Scale Modeling of Ultrafine Particles in Urban Areas with Special Focus on Passenger Ferryboat Emission Impact

Abstract

Air pollution by aerosol particles is mainly monitored as mass concentrations of particulate matter, such as PM₁₀ and PM_2.5. However, mass-based measurements are hardly representative for ultrafine particles (UFP), which can only be monitored adequately by particle number (PN) concentrations and are considered particularly harmful to human health. This study examines the dispersion of UFP in Hamburg city center and, in particular, the impact of passenger ferryboats by modeling PN concentrations and compares concentrations to measured values. To this end, emissions inventories and emission size spectra for different emission sectors influencing concentrations in the city center were created, explicitly considering passenger ferryboat traffic as an additional emission source. The city-scale chemical transport model EPISODE-CityChem is applied for the first time to simulate PN concentrations and additionally, observations of total particle number counts are taken at four different sampling sites in the city. Modeled UFP concentrations are in the range of 1.5-3 × 10⁴ cm^-3 at ferryboat piers and at the road traffic locations with particle sizes predominantly below 50 nm. Urban background concentrations are at 0.4-1.2 × 10⁴ cm^-3 with a predominant particle size in the range 50-100 nm. Ferryboat traffic is a significant source of emissions near the shore along the regular ferry routes. Modeled concentrations show slight differences to measured data, but the model is capable of reproducing the observed spatial variation of UFP concentrations. UFP show strong variations in both space and time, with day-to-day variations mainly controlled by differences in air temperature, wind speed and wind direction. Further model simulations should focus on longer periods of time to better understand the influence of meteorological conditions on UFP dynamics.

Keywords: chemistry transport model; city scale modeling; in-land ferryboat emissions; particle number size distribution; ultrafine particles; urban air quality.

Figure A1

Variation of meteorological conditions on the selected days of this study. Plots for each day show hourly data of wind speed (blue lines), wind direction (green lines), temperature (black lines), and relative humidity (orange lines). Meteorological data from the ground weather station Finkenwerder West.

Figure A2

Map showing the locations of the virtual weather stations. The meteorological data extracted from the DWD forecast at the eight virtual weather stations were used as input to the meteorological pre-processor MCWIND.

Figure A3

Comparison of (a) wind speed (m/s) and (b) wind direction (°) used in the model simulation against measurements at weather station Finkenwerder West in Hamburg during February 2021. Missing model data on 9 February 2021. Measurement data obtained from the air quality monitoring network of the city administration (Luftmessnetz Hamburg (HaLM); https://luft.hamburg.de/clp/meteorologie/clp1/, accessed on 8 September 2021). The transparent orange bars highlight the selected days that were investigated in this study.

Figure 1

Emission size spectra of different emission sectors (black): ferryboats and road traffic (a), ocean-going ships (b), residential heating (c), and background concentration (BSH station Wedel) (d) based on the respective size distributions reported by other studies, provided as grey lines. Solid black lines show the representation of the size distribution in the EPISODE-CityChem model.

Figure 2

Overview of the model processing chain. Blue: data sources; green: input files; orange/red: output sources; yellow: concentration values.

Figure 3

Map of Hamburg with overview of measurement locations. Dark blue: suburban background (Neugraben, 52NG); cyan: urban background (Sternschanze, 13ST); purple: road traffic (Max-Brauer Allee, 70MB); magenta: ferryboat pier (Övelgönne, 90OE); white: background/shipping traffic (location of the measurement data for the background concentration; Wedel, 15WE). The motorway visible on the map right next to the pier is tunneled. The marking in Wedel (15WE) shows the location of constant PN monitoring by BSH that was used for the definition of the constant background concentration.

Figure 4

Modelled average particle number concentration (in cm⁻³) in the Hamburg urban area on five days in February; area: 30 × 30 km², resolution: 1 km²; blue: Water and harbor area. Circles indicate the locations of stations of the Hamburg air monitoring network. A list of the station acronyms is given in Table A1. The color of the circle outline indicates the station category that the station belongs to (dark blue: suburban; cyan: urban background; purple: road traffic; magenta: close to shore/pier).

Figure 5

Modeled PN size distributions (on Day 5; 22 February 2021) according to the determined size classes at selected HaLM measurement stations sorted by station category. Top row shows the stations that were selected as PN measurements sites during the campaign. Stations were grouped into four categories: suburban, urban, road traffic and close to shore/pier. A list of the station acronyms is given in Table A1.

Figure 6

Daily profiles of hourly modeled PN concentrations at the four measurement sites for the five measurement days and the measured PN concentration values at respective daytime. The variation in PN concentrations during the day is mainly controlled by the diurnal variation of emissions and the changes in the meteorological conditions between daytime and nighttime. Red dashed line: model. Blue line segments: measurements. Stations: 13 ST (Sternschanze, urban background); 52 NG (Neugraben; suburban background); 70MB (Max-Brauer Allee, road traffic); 90OE (Övelgönne, close to shore/pier).

Figure 7

Modeled daily PN concentrations at the four measurement sites for the days 1–5 indicating the UFP share (turquoise) in total particle number counts (red).

Figure 8

Comparison of modeled daily PN concentrations at the four measurement sites for days 1–5 with ferryboat emissions included (dark blue) and excluded (light blue).

Figure 9

Hourly extract of the PN concentration map for Hamburg on Day 4 (19 February 2021) (a) with ferryboat emissions and (b) without ferryboat emissions including the Elbe River near the city center as well as the shores. The scale for PN is manipulated for better visualization (see color scale) of the dispersion of ferryboat emissions.

Figure 10

Map of the 500 hPa-geopotential height (in gpdm) and surface pressure (in hPa) for Day 1 (8 February 2021) when an easterly flow of clean polar air led to low PN concentrations in the area of interest. A relatively weak low-pressure system centered over Southeastern Europe and a strong high pressure over Greenland dominated the flow in Northern Europe. (Source: German Weather Central, www.wetterzentrale.de, CFS reanalysis; https://www.wetterzentrale.de/reanalysis.php?jaar=2021&maand=2&dag=8&uur=1200&var=1&map=1&model=cfsr, accessed on 26 November 2021). The red star denotes the area of Hamburg.