On the ubiquitous unstable relation among climate modes and their local effects

Meteorological data recorded in logbooks form the sailing era has received increased attention in the last decades. The daily and even sub-daily data kept by seamen, allows extending back in time the time series of wind and other relevant variables along the historically busiest naval routes. Different projects have integrated these data in the ICOADS open database, and the availability of this massive data collection in its raw format has inspired the development of a new class of climatic indices based on the wind direction variability known as directional indices. They have the advantage that wind direction is the only variable needed, and little or no metadata are required to convert historical observations to current standards. This minimizes the possibility of appearance of data-induced inhomogeneities and non-climatic biases in the resulting series. On these bases, during the last decade, our team has worked in the development of a suite of instrumental circulation indices for remote areas and for periods of time significantly longer than those attained by any other methodology based uniquely on in-situ meteorological observations.Directional indices have been applied to the construction of NAO-like records, to the evaluation of the intensity of several monsoons, to the examination of the variability of local wind systems and even to the analysis of the long-term variability of upwelling favorable winds for several Eastern Boundary Upwelling Systems. The common factor of all these investigations is the lack of stationarity in the relation of the structures analyzed and the classical modes of global climate variability such as El Nio / Southern Oscillation, the Atlantic Multidecadal Oscillation, both Annular Modes, etc. Our combined results provide direct instrumental evidence of the lack of stationarity between global variability modes and their associated local impacts at worldwide scale, challenging the implicit assumption of stationarity that underlies the use of proxies for climate reconstruction.