Spatio-temporal interpolation of paleoclimate data based on δ18O climate variability observations

Abstract

In this study we present a method that allows to compute paleoclimate datasets for the last 1.2 mio. years in 1ky temporal resolution. The computational model is based on a spatio-temporal interpolation along δ18O measures, to provide a high resolution record of paleoclimate variability, between modeled paleoclimate and current observed climate data. In particular we interpolate temperature and precipitation monthly values between time steps. The interpolated results are evaluated, using different techniques and approaches, against modeled and observed published paleoclimate data sets, based on qualitative and quantitative evaluation methods. The interpolated climate data, that are 36 layers per interpolated time step of monthly precipitation, minimum temperature and maximum temperature, will be published as open data, to enable and foster reuse of the data for example in the context of paleoenvironmental modeling applications. The presentation will detail mostly on the development of the interpolation technique and the evaluation of the datasets. It will also give advice for what kind of applications, in terms of scale and scope, this data is suitable and for what kind of applications this data is unsuitable.