Developing a global spatiotemporal database of urban adaptation actions

Climate adaptation is needed as cities around the world face climate impacts increasing in intensity and frequency. Cities have taken many different adaptation approaches, but efforts have largely fallen short. Tracking adaptation progress is needed to ensure adequate implementation, especially in the most vulnerable and under-resourced areas. Despite the importance of tracking progress, a systematic mapping of adaptation at the local level has been hampered by the lack of structured data, given the variety of languages and formats of municipal documents. Most notably, there remains a critical gap in conducting multi-lingual syntheses and, as a result, countries in the Global South tend to be disproportionately underrepresented.

Here, we aim to develop a global spatiotemporal database on urban adaptation plans and actions. More specifically, we want to investigate (a) which measures have been proposed for which sectors and climate impacts, (b) to which extent these plans have been implemented and outcomes recorded, including comparisons of actions mentioned in previous studies, (c) who are the key actors involved in these plans and which voices are “unheard”. We will focus on a range of globally representative cities for which climate change adaptation plans and hazard-specific plans (eg. flood, drought, wildfire) have been developed. Relevant documents will be collected from previous studies and complemented by adding new data sources including self-disclosed adaptation actions from the Carbon Disclosure Project and other adaptation-specific databases. The identified documents will then be analysed using a range of natural language processing (NLP) methods, including neural network pre-trained models (e.g. XLM-RoBERTa) with state-of-the-art transformers architecture. By developing such a multilingual text classifier, we expect to be more inclusive of the data assessed. The expected results will add to the current understanding of adaptation progress globally, filling in data gaps and identifying key information for further research to support implementation.