|
Abstract:
|
Understanding the dynamics of climate change in its full richness requires the knowledge of long temperature time series, which is however not available upon direct observation. We propose a Bayesian hierarchical model to reconstruct past temperatures. This model enables the reconstruction to integrate information from different sources, such as proxies with different temporal resolution and forcings acting as the external drivers of temperature revolution. On the other hand, this model allows us to study the role of different sources in the reconstruction. Via synthetic climate data from global climate models, the skill of each type of proxies, the role of forcings and the effect of noise contained in proxies are carefully investigated, and the strength of combining different sources is well demonstrated.
|
