An updated model for generating historic precipitation isotope time series from machine learning applied in Europe

<p>Hydrogen and oxygen isotope values of precipitation are critically important quantities for applications in Earth, environmental, and biological sciences. However, direct measurements are not available at every location and time, and existing precipitation isotope models are often not sufficiently accurate for examining features such as long-term trends or interannual variability. This can limit applications that seek to use these values to identify the source history of water or to understand the hydrological or meteorological processes that determine these values. We developed a framework using gradient boosted regression tree-based machine learning, which we used to implement a procedure for calculating isotope time series at monthly resolution using available climate and location data. Here we present two new updates to our model, Piso.AI, one of which applies the original approach to new climate predictor data to extend the time series to the 1950-2020 time interval, and the second of which uses a restricted set of predictors to allow time series to be generated that span the range from 1901-2020 with slightly reduced accuracy compared to the original model. Both new products can be applied over most of Europe, and were trained on the historic archive of precipitation isotope data available from the Global Network of Isotopes in Precipitation. These model products facilitate simple, user-friendly predictions of precipitation isotope time series that can be generated on demand and are accurate enough to be used for exploration of interannual and long-term variability in both hydrogen and oxygen isotopic systems. These predictions provide important isotope input variables for ecological and hydrological applications, as well as powerful targets for paleoclimate proxy calibration, and they can serve as resources for probing historic patterns in the isotopic composition of precipitation with a high level of meteorological accuracy. Predictions from our modelling framework are available at https://isotope.bot.unibas.ch/PisoAI/.</p>