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 literature 2 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 con dently 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. MainThere is overwhelming evidence that the impacts of climate change are already being observed in human and natural systems 3 . These effects are emerging in a range of different systems and at different scales, covering a broad range of research elds from glaciology to agricultural science, and marine biology to migration and con ict research 1 . The evidence base for observed climate impacts is expanding 4 , and the wider climate literature is growing exponentially 5,6 . Systematic reviews and systematic maps offer structured ways to collectively identify and describe this evidence while maintaining transparency, attempting to ensure comprehensiveness and reduce bias 7 . However, their scope is often con ned to very speci c questions covering no more than dozens to hundreds of studies.In the climate science community, evidence-based assessments of observed climate change impacts are performed by the Intergovernmental Panel on Climate Change (IPCC) 1 . Since the rst Assessment Report (AR) of the IPCC in 1990, we estimate that the number of studies relevant to observed climate impacts published per year has increased by more than two orders of magnitude (Fig. 1a). Since the third AR, published in 2001, the number has increased ten-fold. This exponential growth in peer-reviewed scienti c publications on climate change 5,6 is already pushing manual expert assessments to their limits. To address this issue, recent work has investigated ways to handle big literature in sustainability science by scaling systematic review and map methods to large bodies ...

Abstract. The degree of trust placed in climate model projections is commensurate to how well their uncertainty can be quantified, particularly at timescales relevant to climate policy makers. On interannual to decadal timescales, model uncertainty due to internal variability dominates and is imperative to understanding near-term and seasonal climate events, but hard to quantify owing to the computational constraints on producing large ensembles. To this extent, emulators are valuable tools for approximating climate model runs, allowing for exploration of the model uncertainty space surrounding select climate variables at a significantly reduced computational cost. Most emulators, however, operate at annual to seasonal timescales, leaving out monthly information that may be essential to assessing climate impacts. This study extends the framework of an existing spatially resolved, annual-scale Earth System Model (ESM) emulator (MESMER, Beusch et al. 2020) by a monthly downscaling module (MESMER-M), thus providing local monthly temperatures from local yearly temperatures. We first linearly represent the mean response of the monthly temperature cycle to yearly temperatures using a simple harmonic model, thus maintaining month to month correlations and capturing changes in intra-annual variability. We then construct a month-specific local variability module which generates spatio-temporally correlated residuals with month and yearly temperature dependent skewness incorporated within. The performance of the resulting emulator is demonstrated on 38 different ESMs from the 6th phase of the Coupled Model Intercomparison Project (CMIP6). The emulator is furthermore benchmarked using a simple Gradient Boosting Regressor based, physical model trained on biophysical information. We find that while regional-scale, biophysical feedbacks may induce non-uniformities in the yearly to monthly temperature downscaling relationship, statistical emulation of regional effects shows comparable skill to approaches with physical representation. Thus, MESMER-M is able to generate ESM-like, large initial-condition ensembles of spatially explicit monthly temperature fields, thereby providing monthly temperature probability distributions which are of critical value to impact assessments.

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.