Abstract. This work provides a hypothesis of the links between the multi-millennia scale recurrent solar and tidal influences and Earth's climate lagged responses, associated with the oceanic transport mechanisms with a variable modulation. As a part of this hypothesis, empirical and simple, non-linear lagged models are proposed for five of the most representative Earth's climate variables (a continental tropical temperature, an Antarctic temperature [at James Ross Island], the Greenland temperature, the global temperature and the southeast asian monsoon) with multi-millennia records to account for the lagged responses to solar forcing. The proposed models implicitely include a well-known oceanic heat transport mechanism: the Ocean Conveyor Belt. This oceanic mechanism appears to generate a climate modulation through the intensity of the ocean/atmosphere circulation, and a heat and mass transport, with a consequent climate lag of several thousands of years. Tidal forcing is also considered for global temperature modelling and forecast. The consequent millennia-scale global forecasts, after being integrated/verified with an accumulated ocean travelled distance from the tropical East Pacific, and with a double evaluation of the tidal influences based on similarities and on the NASA’s solar system astronomical dynamics, indicates a cooling for the next century, and gentle oscillations over the next millennia. Our preliminary results that strongly suggest that millennial scale changes in solar activity induce circulation and thermal global impacts, also suggest that the Younger Dryas event, may be influenced by the lagged outcomes of solar driven changes in the tropical Pacific, and by tidal influences. The detected Earth's climate delayed responses, that have been working in the past and present climates, and will be working in the future climates, must be, as soon as possible, independently verified and theoretically sustained, before to be fully included in a multi-scale climate models as a scientific theory. A final example for the global temperature record over the last 170 years demonstrates with experimental results for the twenty first century evolution the convenience of a multi-scale climate modelling with contrasting lower values compared with the IPCC global temperature scenarios.