There is a pressing need to gather detailed information on the cities of the world that can be used by the urban climate science community (UC) to design observations, run models, compare urban effects, design policy and transfer knowledge. Much of the recent progress in UC has come about by studying the links between aspects of urban form (such as street height to width ratio and impervious surface cover) and the climate outcome. However, in the absence of information on cities generally, this understanding is of limited practical value for addressing urban issues globally. WUDAPT has been conceived as a project for the acquisition, storage and dissemination of climate relevant data on the physical geographies of cities worldwide. The project takes a hierarchal approach to data gathering. At the lowest level (0), cities are decomposed into Local Climate Zones (LCZ), which are culturally neutral descriptions of neighbourhoods. Data at this level is derived using locally-based urban experts and satellite data. Once completed, the LCZ map is used as a sampling frame to gather more detailed (levels 1 and upwards) information on aspects of the form and functions of cities. The data that is gathered will be stored in a geographically referenced database and can be downloaded to support a variety of urban studies. The WUDAPT protocol will enable researchers to contribute to an international project of great relevance for addressing climate and climate change issues. This paper will outline the scope of WUDAPT and progress to this point. It will formally launch the WUDAPT project and invite colleagues to participate in creating a community database that will be of immense value.

In addition to recent progress in the delimitation of global high resolution urban land cover masks from multispectral optical and SAR imagery, the internal differentiation of urban structures and morphologies from the same data is still a challenging task. Further, the required level of detail and the boundaries between the classes depend on the application. In the urban climatology community the Local Climate Zones (LCZ) scheme has recently gained acceptance as a standard typology for the classification of local scale urban landscapes. The method was originally developed for meta-data communication of observational urban heat island studies, but meanwhile it has been successfully applied to mapping studies as well. An especially promising approach is based on multi-temporal multi-spectral and thermal remote sensing data and modern machine learning methods. Due to the culturally neutral nature of the LCZ scheme, the individual classes have different spectral properties in different parts of the world, which makes local training data unavoidable. Therefore, a universal, simple and objective LCZ mapping method based on free data and free software was designed. The method allows local experts to conduct and validate LCZ-classifications for their respective cities and thus contribute to the generation of the Level 0 product for the worldwide database on urban form and materials, WUDAPT.

In this paper we present conceptual considerations for the development of a common methodology to derive LCZ from remote sensing data. We discuss the appropriateness of LCZ mapping, the requirements, as well as limitations. Further, the results from an expert workshop in Dublin are presented where 16 cities in Africa, Asia Europe, as well as North and South America were classified by local experts according to the proposed method. The resulting LCZ atlas for the 16 cities is seen as a proof-of-concept as well as a major contribution to WUDAPT.

The LCZ framework provides a range of values for a number of different parameters needed as inputs to climate and weather models. These parameters are associated with different characterizations of urban land cover (eg. buildings, roads, other pervious surfaces), building and road geometry (e.g. building heights and footprints, canyon widths), building materials, and urban function and activity (e.g. temperature settings, presence of air conditioning). Level 1 and 2 data provide more refined estimates of these parameters at point locations, which can then be aggregated to the resolution of a particular climate or weather prediction model. For the collection of Level 1 data, LCZs can be used as a sampling frame to estimate average parameter values (with the variation) across different LCZs while Level 2 refers to wall-to-wall or more comprehensive data collection. This can either be collected systematically if there are sufficient resources or ancillary databases (e.g. building footprint files) can be used to extract some Level 2 data variables automatically. Different methods of data collection are proposed. For urban land cover, the Geo-Wiki crowdsourcing tool and Google Earth imagery are used for collection of the data. The results of initial experimentation with the collection of Level 2 urban land cover data for the city of Dublin is presented, in particular to determine the optimal sampling strategy for other cities, and to examine the tradeoffs between Level 1 and Level 2 data collection. Other data collection methods will be presented, e.g. use of Google StreetView for collection of data on building and road geometry and building materials, while data collection on the ground using mobile devices will be needed to complete the full suite of parameters. This paper will present the range of data collection options available, which will be at various stages of implementation.

There are a number of regional to global scale community-based modeling system for simulating urban climate and weather e.g., the Community Earth System Models (CESM) Community Land Model – Urban component (CLM-U), several Weather Research and Forecasting (WRF) urban components, the Town Energy balance Model (TEB), the Local Scale Meteorological Parameterization Scheme (LUMPS), and similarly configured community systems for air quality, e.g., the Community Multi-scale Air Quality (CMAQ) are powerful state-of-science based systems. These models provide a modeling framework to provide guidance towards meeting the challenges of population growth, climate changes, air quality, urban sustainability, livability, and human comfort confronting decision makers and society. WUDAPT is designed to provide the database infrastructure so these models can be applied to perform and provide urban climate simulations at from city to global scales. This presentation will provide examples of weather, climate, energy balance and air quality simulation results using the requisite gridded urban morphological data for a variety of cities and to illustrate their utility for providing policy relevant assessment and planning guidance.

Societal guidance, insights and assurances are needed in this period when the world is becoming increasingly more urbanized and confronted with life-impacting climate and pressing environmental issues. WUDAPT is engaged at this juncture, by its support toward facilitating the utilization of advanced science-based modeling tools towards the anticipated myriad of model applications requiring specialized data for all our current and planned urban centers on urban form and human activity. Recognizing that the science in models is rapidly advancing and embodied with increasing sophistication, WUDAPT, through its innovative data collection approach will both provide the requisite urban morphology and activity data, needed in short order, science consistent, fit-for-purpose and on a worldwide basis. Moreover, the Portal and the suite of tools are being designed to provide stakeholder communities with user friendly targeted capabilities to facilitate a host of urban applications. This presentation will describe unique features of this Portal and provide some example applications. Subsequently, we review and summarize the overall progress and outline detailed plans, describe our websites, opportunities for community involvement, country coordination, and next steps towards achieving the goal of a fully implemented WUDAPT

The Local Climate Zone (LCZ) classification is an outstanding concept for the climate-related classification of urban areas in global scale. One of the most important advantages is the possibility to use these zones for the input of different climate or weather models in order to better represent urban areas. The use of this concept in these models is advantageous because this classification is based on the thermal characteristics of the urban areas, and it is connected to the most obvious alteration of the climate in urban areas, the urban heat island.

The LCZ system was initially designed for the classification of urban measurement sites, but meanwhile several methods for LCZ mapping have been proposed. The aim of this study is to present and compare two different LCZ mapping methods. The first approach is based on free multi-temporal remote sensing data and modern machine learning methods using classifiers like random forest (Bechtel-method). The entire workflow was implemented in the open source GIS SAGA. The second method is a GIS based automatic software tool (Lelovics-Gál-method). As an input it uses different parameters of the urban structure (like building height, sky view factor, fraction of buildings, vegetation, built up areas, albedo) acquired from different sources (e.g. satellite and aerial images, building databases, CORINE land cover dataset, road databases and maps). The basic elements of this GIS method are the building block and the lot area polygon around it. The approach consists of a fuzzy preliminary classification and a post-processing scheme. Initially, all lot area polygons are assigned to a most similar and a second most similar LCZ using the parameter ranges given by the LCZ fact sheets. Consequently, the polygons are aggregated to achieve at least the minimal size of 500 m x 500 m for a single LCZs using similarity rules.

The study area of this comparison is Szeged, Hungary, because in this city all of the needed input parameters are available for both methods. As a part of this comparison we analyze which are the most problematic built up types, and also we try to find the advantage of the methods.

Finally, we aim the integration of both approaches combining the respective advantages. Therefore, we conduct the initial classification using the Bechtel-method, since it needs only few and globally available input data. As the second step, the aggregation of the Lelovics-Gál-method was implemented in a JAVA tool, in order to create LCZs of sufficient size.