The importance of input data quality and quantity in climate field reconstructions – results from a Kalman filter based paleodata assimilation method

Abstract. Differences between paleoclimatic reconstructions are caused by two main factors, the method and the input data. While many studies compare methods, we will focus in this study on the consequences of the input data choice in a state-of-the-art paleo data assimilation approach. We evaluate reconstruction quality based on three collections of tree-ring records: (1) 54 of the best temperature sensitive tree-ring chronologies chosen by experts; (2) 415 temperature sensitive tree-ring records chosen less strictly by regional working groups and statistical screening; (3) 2287 tree-ring series that are not screened for climate sensitivity. The three data sets cover the range from small sample size, small spatial coverage and strict screening for temperature sensitivity to large sample size and spatial coverage but no screening. Additionally, we explore a combination of these data sets plus screening methods to improve the reconstruction quality. Neither a large, unscreened collection of proxy data nor the small expert selection leads to the best possible climate field reconstruction. A large collection of unscreened data leads to a poor reconstruction skill. The few best temperature proxies allow for a skillful high latitude temperature reconstruction but fail to provide improved reconstructions for other regions and other variables. We achieve the best reconstruction skill across all variables and regions by combing all available input data but rejecting records with small, insignificant information and removing duplicate records. In case of assimilating tree ring proxies, it appeared to be important to use a tree-ring proxy system model that includes both major growth limitations, temperature and moisture.