Accessing New Cities for Urban Climate Modelling and Study

Currently, over one-half of the planet’s population (7 billion) live in cities and this proportion is expected to increase to 65% by 2050. Most of this increase will occur in the newly emerging cities in the global south. The role of cities as key drivers of environmental change at local, regional and global scales is now widely recognised, especially in the area of climate change. Furthermore, during the COP 21 in Paris, cities were identified as a major component of future engagement in the effort to tackle climate change. One of the first steps in this matter is to produce reliable information for urban decisions makers for climatic mitigation through weather and climate forecasting. Yet a lot of cities, especially in developing countries, lack the resources to perform formal case studies.

Although the urban effect on local climate has been studied during the past 100 years, most of our knowledge has been acquired over the last three decades. Urban climate science has progressed significantly and there are now sophisticated models that link the properties of the urban landscape to atmospheric effects. These properties include details on urban form (buildings, vegetation, etc.) and on urban functions (traffic, industry, etc.). However, we still know very little about the physical characteristics of most cities. Indeed, the poor availability of detailed urban data around the globe was identified by the International Panel on Climate Change (IPCC) as one of the major obstacles to scientific progress. This knowledge and data gap conducted the World Urban Database and Access Portal Tool (WUDAPT) team to find new pathways to obtain data for climate related investigations. Through a supervised classification of Landsat imagery, the technique developed by Bechtel et al. (2015) freely and rapidly offers maps of Local Climate Zones (Stewart and Oke, 2012) for climatological studies. Those Local Climate Zones (LCZ) are described by ranges of physical and morphological values, dividing its land use properties into ten climatic representative neighbourhoods.

For the first time, this new land use classification was implemented in the Weather Research Forecasting (WRF) model, used under the Building Effect Parameterization coupled with the Building Energy Model (BEP-BEM) scheme. The study was done over Madrid, Spain, for a week of typical summer and winter case to test the potential of such data. After a comparison with results obtained under the CORINE classification, the viability of the data was discussed, proving that the WUDAPT technique is definitely reliable, but that further studies need to be done to formalize and improve it. A comparable study on Vienna, Austria, demonstrated that the WUDAPT technique for classifying urban areas was at least as effective as using high-resolution GIS data as input data for WRF.

Considering these studies and the new tools that are being developed by the WUDAPT team, the goal of this research project is to ensure the viability of the LCZ maps for typical developing cities with the objective of modelling their local climate at a mesoscale level. In fact, research groups are already showing enthusiasm regarding this tool for India and Africa. However, the procedure still needs explicit endorsements to be used. Thus, a double sensitivity test will be done over Madrid to evaluate critical parameters for urban climatological models (i.e., BEP-BEM):

1)      As a first step, we will determine how the selection of different Training Areas (zones on which the classification relies), and hence maps, influences the final results within the model. Indeed, since the classification is supervised, the possible impact of human error has to be evaluated.

2)      Secondly, changing morphological and physical values would allow a better delineation of the representative mid-ranges, for each LCZs, used in modelling developing cities. Because this form of data would be the only one available, it needs to be calibrated.

Overall, this study will propose a framework for applying the WUDAPT technique for land use classification to urban climate in developing cities. In particular, the study will focus on the sensitivity analysis of WRF in response to alterations in the input data provided by the WUDAPT method.