Abstract ID: 3.11709 | Accepted as Talk | Talk/Oral | TBA | TBA

Ecosystems and inhabitants of the Andes mountains are vulnerable to warming and precipitation changes, requiring urgent adaptation measures. The standard procedure to design adaptation strategies relies on the use of climate projections and impact models. However, most projections are based on the outputs of coarse global climate models where the Andes cordillera is barely represented. Dynamical and statistical downscaling methods are used to overcome this issue around the world. However, the Andean region requires special analysis because on top of the complex orography, factors such as the observational scarcity, the vast latitudinal range and the diversity of climate regions influence the performance of these methods. To this end, we systematically reviewed downscaling applications covering the Andes mountains. More than 200 scientific articles were included as well as tens of grey literature documents mainly from National Communications. The downscaling journey began in the early 2000s, when regional models with a spatial resolution of around 80km were used. Progressively, dynamical methods have evolved, and spatial resolution has increased to allow convection-permitting simulations that investigate the impact of climate change. In parallel, the statistical downscaling community has grown and moved from multiple linear regression models to state-of-the-art machine learning methods and weather generators. Our review highlights promising advances in the downscaling community. The performance of regional models has improved thanks to optimized parametrizations. Unfortunately, evidence also suggests a disconnection between the methods used for process understanding and the methods used in national projections that are eventually considered for adaptation planning. Furthermore, even when the added value of dynamical simulations has been proven over the Andes, most impact models still rely on simpler bias correction techniques. Another identified barrier is the absence of a consistent evaluation over tropical Andean regions. We have come a long way in the downscaling pathway, but it might be time to consolidate the community, integrate as a region and aim to use the most reliable methods not only for evaluation studies but also for climate change projections and the design of adaptation strategies to improve the resilience of our mountainous countries.