Damaged Cave Deposits Record 200, 000 Years of Paleoseismicity: Dead Sea Transform Region
by
Elisa Joy Kagan
Institute of Earth Sciences, The Hebrew University of Jerusalem, 91904, Israel and Geological Survey of Israel, 30 Malkhe Israel St., Jerusalem, 95501, Israel
Coauthors: Amotz Agnon, Institute of Earth Sciences, The Hebrew University of Jerusalem, 91904, Israel; Miryam Bar-Matthews and Avner Ayalon, Geological Survey of Israel, 30 Malkhe Israel St., Jerusalem, 95501, Israel

Geological research of past earthquakes, typically retrieving records from soft sediment deformations, can benefit from the study of rockfalls and damaged deposits in caves. Dating of damaged speleothems and deposits overgrowing rockfalls constrains the dates of the damaging earthquakes. We have compiled a long-term (200 kyr) paleoseismic record at the Soreq and Har-Tuv caves, near Jerusalem, Israel. The study caves, located 60 km west of the Dead Sea Transform (DST), record earthquake damage from DST ruptures (and possibly, smaller local intraplate events on faults that have not reached the surface or have yet to be observed). The study caves are outside of the rift zone and most likely record stronger and less frequent earthquakes.

Non-seismic sources of collapse, such as ice-movements, ground subsidence, and cave-bears, cited often as alternative origins of cave damage, were considered and refuted. Neither ice cover, nor perma-frost, have occurred in this region during the investigated period. Ground subsidence does not pose a problem since the cave floors are solid carbonate rock. The caves have only non-natural openings, and therefore cave-bears have not entered. The study caves offer an excellent opportunity for paleoseismic research as they contain a large amount of fallen cave deposits of all types and sizes, such as stalactites, stalagmites, soda-straw speleothems, and pillars, as well as other forms of damage. The two study caves present the opportunity to correlate between two nearby sites.

Comprehensive maps of the Soreq and Har-Tuv caves were prepared and demonstrate dominant EW and NW-SE orientation of fractures, and dominant EW and NS orientation of collapsed speleothems. The prevailing orientations of collapsed speleothems are parallel or perpendicular to the trend of the DST. These preferential orientations of collapse strongly support a seismic source of collapse. We identified "new generations" of speleothem growth on top of collapses. This post-collapse precipitation constrains ages of collapse. Unconformities between the collapses and the in-situ regrowth were recognized, and termed paleoseismic "contacts". Laminae above and below each unconformity were separated and dated by the 230Th/234U mass spectrometry and other methods. The pre- and post-seismic dates of a collapse bracket the period within which the earthquake occurred. The closer in age the pre- and post-seismic deposits are, the better constrained the earthquakes age is. When dating post-seismic regrowth on collapsed bedrock (as opposed to collapsed speleothem), only the post-seismic age is available. We also drilled cores into the flowstone floor and discovered laminae that embed fallen small stalactites (known in the literature as soda-straw formations). We dated the laminae that embed the fallen stalactites, which give the age of the seismic event. We also compared the oxygen stable isotopic record (d18O) of the laminae adjacent to the tectonic unconformities with the extensive well-dated stable isotope record of Soreq Cave speleothems, as was reconstructed for the last 185 kyr by Bar-Matthews et al. (2000) and Ayalon et al. (2002), for paleoclimate purposes. This stable isotope technique comparison improves and corroborates the U/Th ages.

Thirty-eight collapses were sampled and dated of which at least 13 (up to 18) separate events were dated. Dating of simultaneous collapses at different areas of the same cave and in the two different caves also strongly supports a seismic source of damage. The one Holocene event observed in the cave correlates with lacustrine seismites dated in cores from the Dead Sea and with an archeologically recorded earthquake. An event dated to ~ 13 ka is recorded by two collapses, for which there is no known dated contemporaneous soft sediment record in the region to correlate with. Most Dead Sea sediments recovered so far show a prominent hiatus at the Pleistocene-Holocene transition. For the period between 75 to 20 ky, we identified 4 events, 2 events (at 51.0-52.0 and 35.5-40.5 ka) correlate with seismites in the Lisan record (at 52 and 39 ka). Another event (at 46.5-46.7 ka) coincides with a hiatus in the lacustrine sections from 49 to 44 ka. An event at 70.2-72.8 ka correlates with a cluster of seismites in the pre-Lisan section. Twelve cave events older than 75 ka are at present the only dated paleoseismic record for this period in the region. Future work on the middle to lower Pleistocene Amora formation can be potentially correlated with our cave record. Overall, the karst paleoseismic record supports the lacustrine seismite evidence.

The long dating range of calcite cave deposits and their potential for recording seismic events can vastly increase the length of the seismic record. The long Late Pleistocene cave record is useful for correlation with other records and for substantiation of the method, while the Holocene events are valuable for seismic hazard assessment. The unique opportunity to compare the stable isotope profiles of the speleothem seismites with the extensive paleoclimate record of the caves improves the accuracy of the seismite ages.

Date received: June 29, 2005

Copyright © 2005 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # caqy-12.