Historical Roofs as a Resource: Towards an automated roof cadastre for Lower Saxony’s heritage
Roofs strongly shape the image of our cities and settlements. Historical roofs contribute to the value of numerous buildings as testimonies to our culture and history. However, it can be observed that historical roof coverings and roof structures are subject to considerable change and are often lost during renovations or conversions. Currently, Germany’s heritage protection laws are being revised so that solar roofs can largely be approved on listed buildings, and, in addition, cities and municipalities are about to revoke or change their preservation statutes as well. It can be assumed that the use of roofs for energy generation through photovoltaic (PV) and solar thermal systems will be an essential factor in the change of listed buildings in the future. This represents a major challenge for the heritage and requires extensive knowledge of the building's roofing materiality, construction and cultural significance. Therefore, a central database with information about all roofs of listed buildings would be of great interest to assess and oversee future developments. The question arises as to how the roofs can be systematically researched and analysed, especially within large inventories.
The Lower Saxony State Office for Monument Preservation (NLD) is leading an interdisciplinary research project with the Institute for Geodesy and Photogrammetry at the Technical University of Braunschweig (IGP) to develop a monument roof cadastre for Lower Saxony. The suitability of monument roofs for solar systems will be the focus of the project. The goal is to develop an ArcGIS-based tool that analyses and evaluates roofs, not only based on their solar potential but also on their roof material, geometry and visibility.
The application will be used for an expanded solar cadastre, which will include qualified heritage data for the first time. The map will provide information on the impact of PV and solar thermal systems on listed roofs and informs which roof areas are less suitable for solar panels. Finally, the results will be integrated into Lower Saxony's geographic information systems and made publicly accessible through publication in the online monument atlas of Lower Saxony. This roof cadastre will benefit local preservation authorities and those involved in planning. The qualified data allows a better planning of solar systems, since the visibility and optical limitation of the monument value can be clarified in advance. The research also allows for the identification of historical material on non-listed objects. Finally, the roof cadastre provides information about the distribution of roofing materials. This creates insights into the historical development. In addition, the data can provide information about the longevity of certain roofing materials. The results associated with the project shall not only improve the handling of the objects in practice but also contribute to a greater knowledge of the inventory.
Modern geographic information technology is key to the project. Analysis methods of geoinformatics and criteria from monument preservation and construction history are combined. A 5.6 km² study area in the city centre of Hanover was selected for the development of a prototype of the roof cadastre. It contains various building types from different epochs with different roof shapes and materials. The study area is shown in Figure 1. The data basis for the study area was provided by the Lower Saxony State Office for Geoinformation and State Surveying (LGLN). It consists of 3D building models in LoD2, TrueDOP, DSM, DTM and ALS data. The data basis is described in more detail in Wichmann et al. (2023).
The roof cadastre is the overall result of several analysis processes. The analysis processes are divided into several automated work packages (Figure 2). Each process results in either a new data set or generates the parameter values relating to a roof area. The final process combines all these parameters into the overall result.
As shown in Figure 2, the first step is to calculate and evaluate the solar potential of the listed buildings. For this step, various existing approaches were tested and compared (e.g. Nelson, 2020; Agugiaro, 2012; Fu, 2000). The 3D building models and the DSM were used as input data. The result is a new data set of listed buildings with high solar potential.
The second step, as shown in Figure 2, is a detailed roof analysis. The detailed roof analysis consists of the visibility analysis, the analysis of the construction features and the classification of the roof material. For the visibility analysis, existing approaches such as that of Wissim et al. (2011) are used. Only the roofs selected in the first step are analysed in terms of their visibility from public areas. The public areas were defined by the LGLN and provided as a layer. As a result, each selected roof receives a parameter value for its visibility.
For the classification of roof materials, a deep learning model is trained with labelled image data. The training dataset contains 13 different roof coverings. The approach of Wyard et al. (2023) is also being tested, in which spectral information of the roofing material is used in addition to the image data. For this purpose, the spectral library for building materials in Karlsruhe, Germany, named KLUM is used (Ilehag, 2019).
So the project deals with georeferenced representation, analysis of material, geometry and visibility of monument roofs. This leads to interdisciplinary collaboration between geoinformatics and heritage preservation. Through this approach the information inherent in aerial photographs, laser scan data and 3-D models can be analysed with regard to the values of listed buildings and converted into maps helpful to the requirements of heritage preservation.
References (see seperate file attached)