Page 2: Regional- to global-scale hydrologic variability
couple of centuries, can occur (Figs. 4 and Figs.9; Street-Perrott and Perrott, 1990; Forman et al., 1992, 1995; Knox, 1993; Ely et al., 1993; Gasse and Van Campo, 1994; Hodell et al., 1995; Muhs and Holliday, 1995; Lamb et al., 1995; Laird et al., 1996; Overpeck, 1996). PAGES/CLIVAR needs to take the lead in mapping out the time-space patterns of these unprecedented changes, and in understanding their causes.

The PAGES/CLIVAR working group on hydrologic variability identified several research foci in which a paleoclimatic perspective would contribute significantly to the goals of CLIVAR. These include:
  • Time-space reconstructions of past climate and hydrologic variability

The compilation and integration of seasonally-to-annually resolved paleoclimatic time series into broad-scale networks suitable for synoptic analysis has been identified as a priority for PAGES, and would provide the time series needed to extend CLIVAR's record back beyond the limited period of instrumental data coverage. A well-organized, documented, easy-to-access, public domain database should include raw paleoclimatic data, climatic reconstructions based on these data, and time series of hypothesized climatic forcing (e.g., trace-gas, solar, and volcanic). Both time-series and spatial reconstructions should be made available for joint investigations within the PAGES/CLIVAR framework.

  • Investigating hypothesized climate system interactions and forcing

Many of the available data discussed above have already been used to examine aspects of climate variability of relevance to CLIVAR. For example, spatial networks of tree-ring chronologies have been used to investigate the past interaction of the tropical Pacific variability and climate change in the Americas, Africa, and China (Cleaveland et al., 1992; Diaz and Pulwarty, 1994). The urgency for investigating the paleoclimatic record of climate and hydrologic variability is highlighted by the greater range of variability observed in some paleoclimatic time series than is observed in instrumental records. For example, paleoclimatic reconstructions point increasingly to the pre-instrumental period occurrence of extreme multidecadal drought in California (Graumlich, 1993; Stine, 1994; Hughes and Graumlich, 1995). PAGES/CLIVAR collaboration should focus on understanding the full range of past climatic and hydrologic variability, including any possible linkages to oceanic, trace-gas, solar, or volcanic forcing.

  • Evaluating predictive climate and hydrologic models

The need for an expanded paleoclimatic model evaluation framework is highlighted in Section 2.5, but this need clearly extends to understanding how well predictive models can simulate the full range of observed climatic and hydrologic variability. Priority efforts should include networks of well-dated time series from periods characterized by modes of variability unlike today's. These include periods within the current Holocene interglacial, but also during the last glacial maximum.



Hydrologic Variability.
Extensive networks of trees and other proxy sources provide the unique opportunity for annually-dated spatial reconstructions of past hydrologic variability extending back centuries. PAGES-CLIVAR collaboration will tap this opportunity to help determine the extent to which the full range of hydrologic variability (e.g., droughts and floods) is predictable.