A major obstacle to producing reliable predictions of climate change and its impacts is a lack of data on time scales longer than the short instrumental record. Recently initiated climate observation programs will need to be continuously operated for at least 50 years before they begin to provide information that is relevant to this problem. In contrast, natural archives of past climate variability can provide relevant information now. Unfortunately, some of the most valuable paleoclimate archives are being rapidly destroyed, largely as a result of human influences (see the related News of the Week article in this issue by Koenig). We cannot afford such an irreversible loss. The Past Global Changes (PAGES) program of the International Geosphere-Biosphere Programme therefore calls for scientists, funding agencies, and institutional partners to establish immediately a coordinated international Global Paleoclimate Observing System (GPOS) to complement the Global Climate, Terrestrial, and Ocean Observing Systems (GCOS, GTOS, and GOOS, respectively) that focus only on contemporary observations. An example of the loss of paleoarchives is the rapid retreat of alpine glaciers in the tropics and temperate latitudes. Ice cores from such glaciers have been used to reconstruct temperature, precipitation, and atmospheric dust levels, and to provide records of changes in the strength of the Asian monsoon and El Nino-Southern Oscillation ([1][1]). As shown in the figure, the total area of the summit glacier on Mt. Kilimanjaro decreased by 82% between 1912 and 2000. Soon, the only information left from the Kilimanjaro ice will be what is contained in the cores extracted last year and stored in freezers at Ohio State University. The situation on Kilimanjaro is not unique. Tropical warming is causing the rapid retreat of ice caps and glaciers at high elevations in the tropics and subtropics around the world ([2][2]). A second example of paleoarchives that are being lost is the widespread damage to tropical corals. Measurements in corals have been successfully used to reconstruct sea surface temperature, salinity, and the surface circulation of the tropical oceans for the past several hundred years, and for isolated windows in the more distant past ([3][3]). Large living corals (more than 100 to 200 years old) suitable for climate reconstruction purposes are relatively rare in most reef areas of the world. These corals are under intense pressure from a myriad of localized stresses related to coastal development and population pressure. In addition, widespread bleaching and mortality of corals are occurring with increasing frequency as the consequences of rising temperatures. Furthermore, studies indicate that corals and other calcifying marine organisms are subject to geochemical stresses from rising CO2 concentrations in seawater ([4][4]). Another biological source of paleoclimate records are tree rings, which have the potential to yield information on many aspects of tropical climate, from the Asian monsoon and El Nino to the factors controlling the storage of carbon in tropical forests ([5][5]). The use of tree rings from tropical trees is relatively new, but is moving into the phase where continental-scale collections must be made and analyzed. Massive felling of the commercially valuable timber of old-growth timber such as teak raises the possibility that, by the time scientists are able to sample them, many of the old trees containing the most valuable information will already have been sent to the sawmills. Paleoarchives provide a wealth of information about past variability of the climate system relevant to future concerns. Thus, we call for an internationally coordinated effort designed to rescue endangered natural archives of past environmental variability and initiate large-scale observational and experimental campaigns to investigate the processes recorded in these natural archives. 1. [↵][6]1. L. G. Thompson , Quat. Sci. Rev. 19, 19 (2000). [OpenUrl][7][CrossRef][8] 2. [↵][9]An overview of the status of glaciers around the world is available from the world glacier monitoring service at . 3. [↵][10]1. M. K. Gagan 2. et al. , Quat. Sci. Rev. 19, 45 (2000). [OpenUrl][11][CrossRef][12] 4. [↵][13]An overview of regions susceptible to bleaching can be found at: . 5. [↵][14]1. R. D'Arrigo , PAGES News 6, 14 (1998). [OpenUrl][15] [1]: #ref-1 [2]: #ref-2 [3]: #ref-3 [4]: #ref-4 [5]: #ref-5 [6]: #xref-ref-1-1 View reference 1 in text [7]: {openurl}?query=rft.jtitle%253DQuat.%2BSci.%2BRev.%26rft.volume%253D19%26rft.spage%253D19%26rft.atitle%253DQUAT%2BSCI%2BREV%26rft_id%253Dinfo%253Adoi%252F10.1016%252FS0277-3791%252899%252900052-9%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [8]: /lookup/external-ref?access_num=10.1016/S0277-3791(99)00052-9&link_type=DOI [9]: #xref-ref-2-1 View reference 2 in text [10]: #xref-ref-3-1 View reference 3 in text [11]: {openurl}?query=rft.jtitle%253DQuat.%2BSci.%2BRev.%26rft.volume%253D19%26rft.spage%253D45%26rft.atitle%253DQUAT%2BSCI%2BREV%26rft_id%253Dinfo%253Adoi%252F10.1016%252FS0277-3791%252899%252900054-2%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [12]: /lookup/external-ref?access_num=10.1016/S0277-3791(99)00054-2&link_type=DOI [13]: #xref-ref-4-1 View reference 4 in text [14]: #xref-ref-5-1 View reference 5 in text [15]: {openurl}?query=rft.jtitle%253DPAGES%2BNews%26rft.volume%253D6%26rft.spage%253D14%26rft.atitle%253DPAGES%2BNEWS%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx