The impact of proxy selection strategies on a millennium-long ensemble of hydroclimatic records in Monsoon Asia

摘要: Abstract Large-scale palaeoclimate reconstructions can be very sensitive to the proxy records they are based on, and hence criteria used select records. Data selection rarely follows objective that applicable all types of proxies, including both low- high-resolution Thus, there is a need for uniform transparent approach assess suitability input data reconstruction. Here, we develop classification multiple evaluate different strategies using network 62 millennium-long terrestrial hydroclimate from Monsoon Asia. Our results reveal robust evidence coherent climate signal high dating accuracy important benchmarking each record. We determine these by reviewing literature record (rather than screening against instrumental data). show proposed yield with stronger common signal. By evaluating uncertainty centennial variability composite reconstructions, differently selected subsets network, it appears beneficial use suitable proxies stemming archives, as well having dense sites. suggest future large-scale might improved networks according universal categories presented here and, if indicated, removing less This will strengthen in final reconstruction, allowing more precise inferences about past comparisons model simulations.