Machine learning-based evidence and attribution mapping of 100,000 climate impact studies

Abstract

An ever-growing body of evidence suggests that climate change is already impacting human and natural systems around the world. Global environmental assessments assessing this evidence, for example by the Intergovernmental Panel on Climate Change (IPCC)1, face increasing challenges to appraise an exponentially growing literature2 and diverse approaches to climate change attribution. Here we use the language representation model BERT to identify and classify studies on observed climate impacts, producing a machine-learning-assisted evidence map which provides the most comprehensive picture of the literature to date. We identify 100,724 (62,950 − 162,838) publications covering a broad range of impacts in human and natural systems across all continents. By combining our spatially resolved database with human-attributable changes in temperature and precipitation on the grid cell level, we infer that attributable climate change impacts may be occurring in regions encompassing 85% (80%) of the world's population (land area). Our results also reveal a substantial 'attribution gap' as robust evidence for attributable impacts is twice as prevalent in high income compared to low income countries. While substantial gaps remain on confidently establishing attributable climate impacts at the regional and sectoral level, our unique database illustrates the broad extent to which anthropogenic climate change may already be impacting natural systems and societies across the globe.

Designing and describing climate change impact attribution studies: a guide to common approaches

“…Mortality-temperature relationships are among the top phenomena documented in this literature. These studies are likely a significant fraction of the broader evidence base for climate change impacts, which Callaghan et al (2021) document in an analysis of over 100,000 studies (presumably including some focused on human health outcomes); they show these are closely correlated with the geography of attributable changes in the climate. Future work might bridge the two approaches, and directly show that evidence of excess heat mortality is clustered in the fastest warming parts of the world.…”

“…Table 2. Examples of how literature-based attribution frameworks could be applied to heat and mortality (Berrang-Ford et al, 2021;Callaghan et al, 2021;Ebi et al, 2020;Yiou et al, 2020).…”

“…Variously called "joint" or "synthesis" attribution (Rosenzweig et al, 2008;Stone et al, 2009), attribution mapping (Callaghan et al, 2021), consistency analysis (Bannister-Tyrrell et al, 2015), or impact fingerprint analysis (Parmesan & Yohe, 2003), what we term the synthesis approach extends the descriptive approach to a much broader scale (though the two approaches obviously exist on a continuum). Synthesis studies draw on hundreds or thousands of studies (increasingly with the help of machine learning), and examine either (1) correspondence between observed impacts and expected impacts of human-caused climate change (e.g., are species' ranges non-randomly shifting towards the poles?…”

“…Climate change is having a marked impact on humans and ecosystems, ranging from shifting burdens of disease to exacerbating global economic inequality and biodiversity loss (Callaghan et al, 2021;Carleton & Hsiang, 2016;Hans-Otto Pörtner et al, 2021. By 2021, over 100,000 studies had reported potential climate-driven changes in human and natural systems (Callaghan et al, 2021).…”

“…Climate change is having a marked impact on humans and ecosystems, ranging from shifting burdens of disease to exacerbating global economic inequality and biodiversity loss (Callaghan et al, 2021;Carleton & Hsiang, 2016;Hans-Otto Pörtner et al, 2021. By 2021, over 100,000 studies had reported potential climate-driven changes in human and natural systems (Callaghan et al, 2021). Quantifying these impacts-and tracing them back to specific sources of human influence on the climate system-is a key step towards building the scientific evidence base to spur climate action, including investment in adaptation and reparative justice.…”