Have plumes been detected seismologically? Maeve O’Shea University of Durham October 2004

Introduction Historically, volcanic activity has been explained by the Plate Tectonic theory. Historically, volcanic activity has been explained by the Plate Tectonic theory. Hotspots do not fit this model. Hotspots do not fit this model. In 1971, Morgan produced the Plume theory In 1971, Morgan produced the Plume theory → proposed that beneath hotspots are narrow plumes which originate from CMB. Model been refined over past 3 decades. Model been refined over past 3 decades. Popular, little opposition. Popular, little opposition. Recent debate as to whether plumes actually exist. Recent debate as to whether plumes actually exist. One of most popular ways to investigate the presence of plumes is to track seismic waves through the mantle. One of most popular ways to investigate the presence of plumes is to track seismic waves through the mantle.

Seismological Methods A number of different techniques have been used to detect plumes, each with limitations: 1. Seismic tomography (i) Teleseismic (ii) Whole-mantle (iii) Surface-wave 2. Receiver functions 3. Multiple ScS 4. Plume waves

Recognising a “plume” What seismological signals do we look for when searching for plumes in the mantle? Large areas of low velocity Large areas of low velocity Shape: bulbous head with thin stem. Shape: bulbous head with thin stem. Must be careful when interpreting low velocity zones as plume structures because a number of factors affect the speed of waves in the mantle: Temperature Temperature Pressure Pressure Rock composition Rock composition Melting Melting Anelasticity Anelasticity Anisotropy. Anisotropy.

Case studies of three hotspot locations

Evidence for the presence of deep mantle plumes Hawaii: Low velocity zone detected, but ~1000km SE of Hawaii- explained by “mantle wind” (Russell et al., 1998). Low velocity zone detected, but ~1000km SE of Hawaii- explained by “mantle wind” (Russell et al., 1998). However, location of anomaly disputed (Bréger and Romanowicz, 1998). However, location of anomaly disputed (Bréger and Romanowicz, 1998). Regional imaging can be difficult due to remote location. Regional imaging can be difficult due to remote location. Different method of receiver functions used: thinned transition zone detected (Li et al. 1999). Different method of receiver functions used: thinned transition zone detected (Li et al. 1999). Global tomographic imaging of the Hawaiian hotspot (Zhao, 2001): Global tomographic imaging of the Hawaiian hotspot (Zhao, 2001):

Deep mantle plume theory is supported by new method of global tomography: finite- frequency (Montelli et al., 2004). Deep mantle plume theory is supported by new method of global tomography: finite- frequency (Montelli et al., 2004). Alternative theory proposed: Tear in the crust (Foulger). Alternative theory proposed: Tear in the crust (Foulger). Other deep plumes: Ascension, Azores, Canary, Easter, Samoa and Tahiti (Montelli et al., 2004)

Evidence for the presence of shallow plumes Iceland: First imaged as a plume originating from CMB (Bijwaard & Spakman, 1999) : First imaged as a plume originating from CMB (Bijwaard & Spakman, 1999) : Foulger et al. (2000) started seriously questioning the belief that all plumes originated from lower mantle Foulger et al. (2000) started seriously questioning the belief that all plumes originated from lower mantle Used teleseismic images Found problems with past models of the Iceland plume. Ritsema et al. (1999)- whole mantle tomography as evidence for NO deep plume. Ritsema et al. (1999)- whole mantle tomography as evidence for NO deep plume.

More recently, others have come to a similar conclusion that Iceland originated from the upper mantle (Allen et al., 2002; Montelli et al., 2004). More recently, others have come to a similar conclusion that Iceland originated from the upper mantle (Allen et al., 2002; Montelli et al., 2004). Other shallow plumes: Bowie, Eastern Australia, Eifel, Etna, Cocos-Keeling, Galapagos and Juan de fuca/Cobb (Montelli et al. 2004).

Absent Plumes Yellowstone: First thought to be a typical deep plume hotspot. First thought to be a typical deep plume hotspot. Scientists have since realised that it did not fit deep-mantle plume model: Scientists have since realised that it did not fit deep-mantle plume model: Teleseismic tomography: shallow Teleseismic tomography: shallow plume, extends only as far as 200 km depth Receiver functions: do not show Receiver functions: do not show any thinning of transition zones (Christiansen et al., 2002) Whole-mantle tomography: no Whole-mantle tomography: no evidence for lower mantle influence (Ritsema et al., 1999). Montelli et al. (2004)- no evidence for a substantial plume underneath Yellowstone. Montelli et al. (2004)- no evidence for a substantial plume underneath Yellowstone.

Conclusion Low velocity anomalies been detected under many hotspots. However, still dispute over whether to call them plumes. Low velocity anomalies been detected under many hotspots. However, still dispute over whether to call them plumes. Difficult to accept since not all hotspots fit the deep-mantle plume model. Difficult to accept since not all hotspots fit the deep-mantle plume model. Limitations in imaging Earth’s interior, therefore new data and more effective analysis techniques are needed. Limitations in imaging Earth’s interior, therefore new data and more effective analysis techniques are needed. Combine seismic evidence with geochemical observations in order to get a clearer picture of what exactly is happening underneath hotspots. Combine seismic evidence with geochemical observations in order to get a clearer picture of what exactly is happening underneath hotspots. The current research project on Hawaii will provide greater insight into the debate of whether plumes really do exist. The current research project on Hawaii will provide greater insight into the debate of whether plumes really do exist.

Additional material Methods: Receiver functions Receiver functions