Comparison and integration of LCZ classification methods based remote sensing and GIS

Evaluating the effect of adaptation and mitigation measures is important for urban development strategies. This can be achieved using high resolution numerical models. However, they are computationally expensive, thus simulating a 30-year climate period is challenging. An approach can be to simulate only a subset of days from the 30 years. Identifying the number of days which are sufficient to represent the urban climate is the aim of this presentation. The presented statistical dynamical downscaling method is applied to simulate the urban climate of Hamburg. It utilises 30-year time series from 27 weather stations in Northern Germany and The Netherlands. For some meteorological quantities measured at these stations, the frequency distributions have been analysed. These are compared with artificial frequency distributions built with bootstrapping and a lower number of days. For comparing these distributions, a skill score following Perkins et al. (2007) is further developed, now taking into account the relationship between the quantities. The results of this statistical dynamical downscaling method indicate that the statistics of the urban climate of Hamburg can be simulated with a much lower number of days than the 30-year time series. Perkins, S. A., A. J. Pitman, N. J. Holbrook, J. McAneney (2007): Evaluation of the AR4 climate models simulated daily maximum temperature, minimum temperature and precipitation over Australia using probability density functions, Journal of climate, 20, 4356-437