Tracking adaptation to climate change using big data

Adaptation has become a core element of climate change policy and research, and figures prominently in the UN Paris Agreement (Lesnikowski et al., 2017; Magnan and Ribera, 2016). Global funds to support adaptation in low and middle income nations have begun to be disbursed and adaptation financing is expected to significantly increase by 2020, with $100bn pledged (Donner et al., 2016). National governments in high income nations have also identified the importance of adaptation, including across Europe, and have begun to invest in specific actions (Biesbroek et al., 2010; Jude et al., 2017). Strategic allocation of adaptation funds and assessment of adaptation progress will require measurement of whether pronouncements on the need for adaptation are translating into action (Ford et al., 2015a; Magnan, 2016; Magnan and Ribera, 2016). The importance of developing systematic and standardized means of assessing adaptation from which future progress can be monitored and evaluated, including the creation of baselines and indices, has been identified by the UN, national governments, and the private sector. Longitudinal assessment in particular, is critical for assessing national investments in adaptation, facilitating policy learning and the sharing of best practices between nations, promoting accountability and transparency of adaptation financing, and for guiding national adaptation planning (Arnott et al., 2016; Lesnikowski et al., 2016). Despite this, there has been little consideration of how to track adaptation systematically across nations, and we thus have limited and fragmented evidence on adaptation progress globally (Berrang-Ford et al., 2014; Ford and Berrang-Ford, 2016).

In response to this challenge, the Tracking Adaptation to Climate Change Collaboration (TRAC3) was created to facilitate new and innovative research that improves our understanding of adaptation to climate change around the world (www.trac3.ca). A key focus of TRAC3 is the development of novel approaches and indicators for assessing adaptation progress across nations globally. First generation work has used national reporting on adaptation action as a basis for creating a global adaptation index for nations globally, but is constrained by limited and biased reporting, and an absence of dataset on adaptation actions (Araos et al., 2016; Berrang-Ford et al., 2014; Epule et al., 2017; Ford et al., 2015b; Ford et al., 2015c; Lesnikowski et al., 2016; Lesnikowski et al., 2013; Lesnikowski et al., 2015; Lesnikowski et al., 2011; Panic and Ford, 2013). Herein we are seeking a student to help pioneer the use of methods rooted in computer science/big data to identify, document, and characterize what nations, regions, and sectors are doing on adaptation as a basis for creation a second generation global adaptation index. The student will work with the supervisors bringing and further developing skills such as latent semantic analysis (word2vec), topic modelling (LDA), web scraping methods, and supervised machine learning to document, retrieve, and analyze data on adaptation policies contained in official government documents, primarily laws, and ministry or executive actions at the national level. The work will directly feed into the global climate policy stocktaking process as part of the Paris Agreement, as well as regional efforts to examine adaptation policy (e.g. through European Environment Agency, World Bank, UNDP).