In urban areas, dense atmospheric observational networks with high-quality data are still a challenge due to high costs for installation and maintenance over time. Citizen weather stations (CWS) could be one answer to that issue. Since more and more owners of CWS share their measurement data publicly, crowdsourcing, i.e., the automated collection of large amounts of data from an undefined crowd of citizens, opens new pathways for atmospheric research. However, the most critical issue is found to be the quality of data from such networks. In this study, a statistically-based quality control (QC) is developed to identify suspicious air temperature (T) measurements from crowdsourced data sets. The newly developed QC exploits the combined knowledge of the dense network of CWS to statistically identify implausible measurements, independent of external reference data. The evaluation of the QC is performed using data from Netatmo CWS in Toulouse, France, and Berlin, Germany, over a 1-year period (July 2016 to June 2017), comparing the quality-controlled data with data from two networks of reference stations. The new QC efficiently identifies erroneous data due to solar exposition and siting issues, which are common error sources of CWS. Estimation of T is improved when averaging data from a group of stations within a restricted area rather than relying on data of individual CWS. However, a positive deviation in CWS data compared to reference data is identified, particularly for daily minimum T. To illustrate the transferability of the newly developed QC and the applicability of CWS data, a mapping of T is performed over the city of Paris, France, where spatial density of CWS is especially high.

In recent years, the collection and utilisation of crowdsourced data has gained attention in atmospheric sciences and citizen weather stations (CWS), i.e., privately-owned weather stations whose owners share their data publicly via the internet, have become increasingly popular. This is particularly the case for cities, where traditional measurement networks are sparse. Rigorous quality control (QC) of CWS data is essential prior to any application. In this study, we present the QC package “CrowdQC+,” which identifies and removes faulty air-temperature (ta) data from crowdsourced CWS data sets, i.e., data from several tens to thousands of CWS. The package is a further development of the existing package “CrowdQC.” While QC levels and functionalities of the predecessor are kept, CrowdQC+ extends it to increase QC performance, enhance applicability, and increase user-friendliness. Firstly, two new QC levels are introduced. The first implements a spatial QC that mainly addresses radiation errors, the second a temporal correction of the data regarding sensor-response time. Secondly, new functionalities aim at making the package more flexible to apply to data sets of different lengths and sizes, enabling also near-real time application. Thirdly, additional helper functions increase user-friendliness of the package. As its predecessor, CrowdQC+ does not require reference meteorological data. The performance of the new package is tested with two 1-year data sets of CWS data from hundreds of “Netatmo” CWS in the cities of Amsterdam, Netherlands, and Toulouse, France. Quality-controlled data are compared with data from networks of professionally-operated weather stations (PRWS). Results show that the new package effectively removes faulty data from both data sets, leading to lower deviations between CWS and PRWS compared to its predecessor. It is further shown that CrowdQC+ leads to robust results for CWS networks of different sizes/densities. Further development of the package could include testing the suitability of CrowdQC+ for other variables than ta, such as air pressure or specific humidity, testing it on data sets from other background climates such as tropical or desert cities, and to incorporate added filter functionalities for further improvement. Overall, CrowdQC+ could lead the way to utilise CWS data in world-wide urban climate applications.

Air pollution is the number one environmental cause of premature deaths in Europe. Despite extensive regulations, air pollution remains a challenge, especially in urban areas. For studying summertime air quality in the Berlin–Brandenburg region of Germany, the Weather Research and Forecasting Model with Chemistry (WRF-Chem) is set up and evaluated against meteorological and air quality observations from monitoring stations as well as from a field campaign conducted in 2014. The objective is to assess which resolution and level of detail in the input data is needed for simulating urban background air pollutant concentrations and their spatial distribution in the Berlin–Brandenburg area. The model setup includes three nested domains with horizontal resolutions of 15, 3 and 1 km and anthropogenic emissions from the TNO-MACC III inventory. We use RADM2 chemistry and the MADE/SORGAM aerosol scheme. Three sensitivity simulations are conducted updating input parameters to the single-layer urban canopy model based on structural data for Berlin, specifying land use classes on a sub-grid scale (mosaic option) and downscaling the original emissions to a resolution of ca. 1 km × 1 km for Berlin based on proxy data including traffic density and population density. The results show that the model simulates meteorology well, though urban 2 m temperature and urban wind speeds are biased high and nighttime mixing layer height is biased low in the base run with the settings described above. We show that the simulation of urban meteorology can be improved when specifying the input parameters to the urban model, and to a lesser extent when using the mosaic option. On average, ozone is simulated reasonably well, but maximum daily 8 h mean concentrations are underestimated, which is consistent with the results from previous modelling studies using the RADM2 chemical mechanism. Particulate matter is underestimated, which is partly due to an underestimation of secondary organic aerosols. NO_x (NO + NO₂) concentrations are simulated reasonably well on average, but nighttime concentrations are overestimated due to the model's underestimation of the mixing layer height, and urban daytime concentrations are underestimated. The daytime underestimation is improved when using downscaled, and thus locally higher emissions, suggesting that part of this bias is due to deficiencies in the emission input data and their resolution. The results further demonstrate that a horizontal resolution of 3 km improves the results and spatial representativeness of the model compared to a horizontal resolution of 15 km. With the input data (land use classes, emissions) at the level of detail of the base run of this study, we find that a horizontal resolution of 1 km does not improve the results compared to a resolution of 3 km. However, our results suggest that a 1 km horizontal model resolution could enable a detailed simulation of local pollution patterns in the Berlin–Brandenburg region if the urban land use classes, together with the respective input parameters to the urban canopy model, are specified with a higher level of detail and if urban emissions of higher spatial resolution are used.