[Revision of Röck et al. 2023, "A Global Database on Whole Life Carbon, Energy and Material Intensity of Buildings (CarbEnMats-Buildings)"] Data descriptor article (Preprint): https://zenodo.org/doi/10.5281/zenodo.8378938 (all versions) Data records / database: https://zenodo.org/doi/10.5281/zenodo.8363894 (all versions) Latest version: Available via https://github.com/mroeck/carbenmats-buildings (main) Abstract Globally, interest in understanding the whole life cycle related greenhouse gas (GHG) emissions of buildings is increasing. Robust data is required for benchmarking and analysis of parameters driving so-called ‘whole life carbon’ (WLC) emissions, but also the energy and material flows. Here the seed for a global buildings database (GBDB) on whole life carbon, energy performance, and material intensity of buildings is presented. GBDB compiles information on more than 1,200 building case studies and includes 155 attributes in its core version and more than 330 attributes overall. It provides insights on context and site, building design, assessment methods, energy performance and material intensity, as well as GHG emissions across the different building life cycle stages. GBDB contains more than 130,000 specified building data points, representing more than 5,000,000 square meters of floor space. The data was collected through various meta-studies, combining both data from scientific literature as well as industry sources, using dedicated data collection templates and Python scripts for data processing, all of which are available alongside this descriptor. Background & Summary The need for radically reducing greenhouse gas (GHG) emissions globally demands defining and implementing thorough assessments and performance classes for both operational and embodied GHG emissions of buildings to provide relevant guidance for policymakers and practitioners1,2. The topic of so-called whole life carbon (WLC) of buildings is gaining increasing attention among decision-makers concerned with climate and industrial policy, as well as building procurement, design, and operation3,4. However, most open buildings datasets published thus far have been focusing on building’s operational energy consumption and related parameters5–8. Recent years furthermore brought large-scale datasets on building geometry (footprint, height)9,10 as well as the publication of some datasets on building construction systems and material intensity11–15. Heeren and Fishman’s database seed on material intensity (MI) of buildings13, an essential inspiration for this work, was a first step towards an open data repository on material-related environmental impacts of buildings. In their 2019 descriptor, the authors present data on the material coefficients of more than 300 building cases intended for use in studies applying material flow analysis (MFA), input-output (IO) or life cycle assessment (LCA) methods. Furthermore, Yang et al.15 established an MI database for Chinese urban buildings with 813 samples. Sprecher et al. 14 presented MI data for Dutch buildings, compiling data on 61 large scale demolition projects. Guven et al.12 elaborated on these efforts by publishing a construction classification system database for understanding resource use in building construction. And most recently, Fishman et al. 11 published RASMI, which provides global ranges of building material intensities differentiated by region, structure, and function. However, there is still a lack of publicly available data that combines material intensity, energy performance as well as assesses life cycle-related environmental impacts, such as life cycle-related GHG emissions, i.e. building’s whole life carbon. Here we present a global buildings database seed on whole life carbon emissions, energy performance, and material intensity (GBDB CarbEnMats). This GBDB dataset provides information on more than 1,200 buildings worldwide, representing more than 5,000,000 square meters of floor space. The database contains more than 130,000 specified building data points. The dataset includes attributes on geographical context and site, main building design characteristics, LCA-based assessment methods, as well as information on energy and material use, and related WLC emissions with a focus on embodied carbon (EC) emissions. The dataset compiles data obtained through a systematic review of the scientific literature as well as systematic data collection from both literature sources and industry partners. By applying a uniform data collection template (DCT) and related automated procedures for systematic data collection and compilation, we facilitate the processing, analysis and visualization along predefined categories and attributes, and support the consistency of data types and units. The descriptor includes specifications related to the DCT spreadsheet form used for obtaining these data as well as explanations of the data processing and feature engineering steps undertaken to clean and harmonise the data records. The validation focuses on describing the composition of the dataset and values observed for attributes related to whole life carbon, energy performance and material intensity. The data published with this descriptor offers the largest open compilation of data on whole life cycle emissions, energy performance and material intensity of buildings published to date. This open dataset is expected to be valuable for research applications in the context of approaches like material flow analyses (MFA), input-output analyses (I/O) and building LCA modelling. It offers a unique data source for benchmarking whole life carbon in context of buildings’ energy performance and material intensity to inform policy and decision-making decarbonization of building construction and operation as well as for informing sustainable real estate practices. The data are particularly suitable for gaining insights on the scale and proportions of the corresponding values to be able to carry out benchmarking and plausibility checks in future studies. At the same time, the analysis of the case study sources used provides insights into the status of the type, scope, level of detail and completeness of the description of the objects of assessment reported. This reveals considerable gaps that can lead to uncertainties in the evaluation. This article includes suggestions for an extensive list of attributes that should be used for the description of building case studies. It is also intended as a call to improve transparency and traceability in the documentation of building LCA studies in both research and industry practice. This database seed is intended as a starting point for further and more comprehensive publications on whole life carbon building data in the future. Files All files related to this database descriptor are available on a public GitHub repository (https://github.com/mroeck/carbenmats-buildings) and the related release via Zenodo (https://doi.org/10.5281/zenodo.13221978). The repository contains the following files: README.MD is a text file with instructions on how to use the files and documents. gbdb_data_core.CSV is the CORE building dataset in CSV text format (separator = ‘;’). gbdb_data_full.CSV is the FULL building dataset in CSV text format (separator = ‘;’). gbdb_data.XLSX is the building database in MS Excel format, containing both the CORE and FULL datasets in separate sheets, with CORE in the first sheet. gbdb_attributes.XLSX is a table with the complete attribute description. gbdb_materials.XLSX is a table for mapping the original material attributes with different other common types of material classifications used in this database. gbdb_py_benchmarking.IPYNB is a Python Jupyter Notebook including code to define data subsets for creating custom benchmark data tables for carbon and materials. gbdb_py_visualizations.IPYNB is a Python Jupyter Notebook for visualization of key attributes and properties, some of which are presented in this descriptor. Authors and contact Authors: Martin Röck1,2,3*, Andreas Sørensen3, Maria Balouktsi4,5, Marcella Ruschi Mendes Saade2, Freja Nygaard Rasmussen6, Harpa Birgisdottir4, Rolf Frischknecht7, Thomas Lützkendorf5, Hoxha Endrit4, Guillaume Habert8, Daniel Satola6, Barbara Truger2, Buket Tozan4, Matti Kuittinen9, Alaux Nicolas2, Karen Allacker1, Alexander Passer2 Affiliations: 1. KU Leuven, BE; 2. TU Graz, AT; 3. Ramboll, DK; 4. AAU BUILD, DK; 5. KIT Karlsruhe, DE; 6. NTNU, NO; 7. Treeze Ltd., CH; 8. ETH Zurich, CH; 9. Aalto University, FI. *Corresponding author: Martin Röck (martin.roeck@kuleuven.be)