Sensitivity analyses and validation of an ENVI-met microclimate model for a greened urban study area in Cologne Südstadt under various typical weather conditions

Increasing heat exposure and heat stress due to global warming is a major problem in cities, and adaptation strategies to improve the urban microclimate are urgently needed. Successful climate change adaptation requires data, models and scenario analyses to evaluate the potentials and thermal effects of measures like green infrastructures, facade and roof greenings, shadings or light surfaces. In order to be able to estimate the effectiveness of different adaptation strategies in agglomeration areas, a physically-based 3D ENVI-met model was setup to simulate the urban microclimate of a 16 ha study area in Cologne Südstadt which is characterized by dense development in the east as well as an urban park area in the western part. In order to assess and test the model outputs, various sensitivity analyses were implemented, especially with regard to the effects of different meteorological forcing data. We compared the output data i) when using measured radiation (incoming direct and diffuse shortwave radiation and incoming longwave irradiance), and ii) when using measurements of the relative cloud cover of low-level, medium-level and high-level clouds as radiative forcing of the model. Radiation is measured with a diffusometer consisting of two pyranometers with and without a shadow ring and a pyrgeometer. The instruments are installed at our meteorological station (Campbell Scientific) in the urban park. The relative cloud cover for the different cloud levels is derived from the measurements of a close-by ceilometer. In addition to the sensitivity analyses, a model validation is carried out based on reference data of a densely distributed weather station network consisting of 33 NETATMO weather sensors and ultrasonic anemometers in the study area, which have previously been checked for long-term stability and consistency in the field and under laboratory conditions. On the basis of air temperature, relative humidity, wind speed and wind direction measurements, the quality of the ENVI-met model outputs is evaluated for statistical goodness of fit, and it is checked how accurate the ENVI-met model can represent and reproduce the actual measurements at the heterogeneous NETATMO test sites. The model validations and sensitivity analyses are accomplished for various typical weather conditions in order to show how the model quality differs between low-exchange high-pressure conditions and mixed weather conditions with different dominant wind directions. Further research will use this validated ENVI-met model to simulate and analyse the effects and potentials of various hypothetical adaptation measures taking into account climate change projections until 2099, and different scenarios will be compared in order to mitigate future heat stress and improve the urban microclimate.