Meteorologische Zeitschrift (Nov 2017)
Intra and inter ‘local climate zone’ variability of air temperature as observed by crowdsourced citizen weather stations in Berlin, Germany
Abstract
A one-year data set for the year 2015 of near-surface air temperature (T$T$), crowdsourced from ‘Netatmo’ citizen weather stations (CWS) in Berlin, Germany, and surroundings was analysed. The CWS data set, which has been quality-checked and filtered in a previous study, consists of T$T$ measurements from several hundred CWS. It was investigated (1) how CWS are distributed among urban and rural environments, as represented by ‘local climate zones’ (LCZ), (2) how LCZ are characterised in T$T$ along the annual cycle and concerning intra-LCZ T$T$ variability, and (3) if significant T$T$ differences between LCZ (ΔT$\Delta T$) can be detected with CWS data. Further, it was investigated how the results from CWS compare to reference data from standard meteorological measurement stations. It can be shown that all ‘urban’ LCZ are covered by CWS, but only few CWS are located in ‘natural’ LCZ (e.g. forests or urban parks). CWS data along the annual cycle show generally good agreement to reference data, though for some LCZ monthly means between both data sets differ up to 1 K. Intra-LCZ T$T$ variability is particularly large during night-time. Statistically significant ΔT$\Delta T$ can be detected with CWS data between various LCZ pairs, particularly for structurally dissimilar LCZ, and the results are in agreement with existing literature on LCZ or the urban heat island. Furthermore, annual mean ΔT$\Delta T$ in CWS data agree well with reference data, thus showing the potential of CWS data for long-term studies. Several challenges related to crowdsourced CWS data need further investigation, namely missing meta data, the non-standard measurement locations, the imbalanced availability in time and space, and potentials to combine CWS and reference data to benefit from the main advantages of both, i.e., the large number of stations and the high quality of data, respectively.
Keywords
