slideshow 1 slideshow 2 slideshow 3 slideshow 4 slideshow 5 slideshow 6

You are here

Natalia Artemieva

2011 Research Report

1.  Ejecta layer at K-Pg boundary and extrapolation to Archean spherules (PI, 4 year of PGG grant. Collaboration with Imperial College, London, GB; University of Edinburgh, GB; University of Muenster, Germany; Oberlin University).


The K-P boundary is widely recognized as a global ejecta layer formed by a large meteorite impact 65 million years ago. The discovery of an iridium anomaly, shocked quartz grains, and the Chicxulub crater provided the strongest confirmation of the impact hypothesis. However, a lot of questions are still opened. The main results of my numerical models are as follows: 1) the K-Pg spherules represent mainly the projectile material, not the crystalline basement material (as it was assumed for many years); 2) shock quartz grains from the basement were distributed globally non-ballistically, i.e. by impact-generated winds; 3) ejecta re-entry may heat the atmosphere and initiate local fires up to a distance of 5000 km; ejecta itself may be molten and partially vaporized during the re-entry.


Results of numerical modeling – ejecta dispersion (Artemieva and Morgan 2011). Up to distances of 2000 km from the crater, ejecta are deposited rapidly without substantial deviation from ballistic trajectories. At larger distances, heating by re-entering ejecta creates strong atmospheric flows in the upper atmosphere that can disperse small fragments (molten spherules and shocked quartz grains of < 1 mm in diameter) preferentially downrange for large distances. According to the model, two-three hours after the impact, small ejecta fragments may have traveled up to a few thousand km beyond their nominal re-entry site based solely on ballistic calculations. After that, their final deposition through the dense lower atmosphere or ocean may take days or weeks. This mechanism is similar in some ways to horizontal dispersal of volcanic aerosols in stable atmospheric winds, but it is much more intense. A more analogous mechanism (atmospheric skidding) has been suggested to explain re-entering ejecta after the Shoemaker-Levy 9 Comet impact on Jupiter.


Results of numerical modeling – re-heating and re-melting of spherules. The temperature of an isolated particle entering the atmosphere is defined by the Whipple law and depends on particle velocity and local atmospheric density. Usually, mm-sized fragments reach the surface without any substantial heating whereas cm-sized particles may be partially molten (Australasian tektites represent these partially molten fragments). If ejecta particles re-entry the atmosphere in massive numbers, much higher temperatures may be attained. Calculations for the K-Pg impact indicate the ejecta did not generate temperatures high enough to initiate wildfires, but within a distance of 5000 km, temperatures may have been high enough to cause re-melting and partial vaporization of silicate particles and thermal decomposition of calcite during initial re-entry. This result shows that to understand the spherules that dominate distal ejecta layers, we have to take into account reactions between different components of ejecta not only within an impact plume (which expands and cools within a minute), but also during the re-entry (where particles maintain higher temperature for tens of min).


Application to wildfires at K-Pg boundary (Morgan et al. 2012). The discovery of large amounts of soot in K-Pg boundary clays led to the hypothesis that major wildfires were a consequence of the K-Pg impact. Subsequently, several lines of evidence, including the lack of charcoal in North American sites, were used to argue against global wildfires. New analysis shows that the K-Pg record appears to be consistent with the occurrence of some local fires in North America. Fires could be ignited locally by the impact fireball and globally through the re-entry of hypervelocity ejecta. Using the results of the Chicxulub ejecta modeling, we estimate the thermal pulse delivered to the Earth’s surface when ejecta re-enters the atmosphere. Our models predict thermal pulses of over 35 kW/m2 in the downrange direction, up to a few thousand kilometers from Chicxulub, but are less severe in other directions and for distal sites.   A fire propagation apparatus has been used to conduct experiments on three leaf types, an angiosperm, fern and conifer, to explore what thermal radiation levels are required for ignition.


Experimental evidences (Deutsch et al. 2011, Deutsch and Artemieva 2012). According to numerical models, over 500 km³ of the CM2 projectile, mainly Si, Fe, and Mg should have been deposited in the K-Pg event bed outside the continuous ejecta blanket. So far,  ~1.5 % of the projectile have been detected in the K-Pg boundary layer; i.e., PGE as well as other siderophiles which always form a spike at the very top of the ejecta layer. Microrpobe analyses of K-Pg spherules (University of Muenster, Germany) suggest that Rear Earth elements (REE) patterns are flat and chondritic or sub-chondritic (= 0.1 x CI), rarely exceeding the CI abundances. Mixing calculations limit the contribution of Upper Continental Crust material to the REE budget to (much) less than 2%. The flat REE patterns substantiate that the siliceous host material (= spherules) consists solely of projectile matter. Depending on the sampling site, these spherules amount to between 10 and ~70 vol% of the event bed. The widespread occurrence of condensed projectile matter in the K-Pg event bed reconciles observations with impact models.


Extrapolation to Archean spherules (Artemieva 2011, Artemieva and Simonson 2012). Spherule-rich layers from a minimum of 4 large impacts that happened between ~2.63 and 2.49 Ga have been identified in Western Australia, 3 of which have been correlated to layers in South Africa believed to have been formed by the same impacts. The textures of spherules in APB and Early Archean ejecta seem to differ statistically from those in younger layers. This suggests a systematic difference in the chemical composition and/or crystallization history of spherules formed in the Archean vs. later in Earth history. One contributing factor is that terrestrial target rocks were probably more mafic on average during the Archean than later in Earth history. The numerical models suggest a second factor. The fact that large impacts generate much greater numbers of ejecta particles than small impacts means that the spherules from large impacts will experience much more heating during re-entry. This in turn suggests spherules from larger impacts will be more extensively crystallized via thermal devitrification and/or wholesale melting during re-entry than spherules from smaller impacts. Several lines of evidence suggest Archean impacts were larger on average than younger ones. The crystallized rims that are typically found on APB spherules but rare on younger spherules could perhaps represent a case of heating intermediate between larger (on average) impacts in the Early Archean and smaller impacts later in Earth history.


2.  Impact ejecta on the Moon and Ar-Ar method (PI of the LASER project, 1-2 year. Collaboration with Natural History Museum in Berlin).

The 40Ar/39Ar method is widely used for the acquisition of impact related cooling ages of a variety of lunar rocks. Several of the Apollo, Luna and lunar meteorite samples dated (e.g., [1-3]) have demonstrated that there is a decoupling between K-Ar reset ages and their shock related petrographic features: partial to total resetting of the K-Ar system occurs even at low shock pressures when post-shock temperature increase is low (~100ºC). Therefore, the 40Ar/39Ar data suggest that the Apollo samples, after an impact event, were kept in a warm environment for a certain amount of time which permitted the partial to total resetting of the K-Ar clock. The most likely feature for this warm environment is the impact ejecta, which in the case of the impact basins formed during the period from ~4.5 to ~3.9 Ga (e.g. Imbrium), correspond to large volumes of material being displaced globally.

Numerical model and results. We model impact cratering on the Moon with the 3D hydrocode SOVA complemented by the ANEOS equation of state for geological materials. We use two thermal profiles within the target: cold-present-Moon with a thermal gradient of 2 K/km, and hot-ancient-Moon with a thermal gradient of 15 K/km. To define ejecta distribution on the surface, we use a method of ballistic continuation on a sphere. The evolution of post-depositional temperature profile within the ejecta blanket is estimated via one-dimensional thermal conductivity equation.

At the present-cold-Moon, antipodal ejecta (at 5460 km from impact) have a temperature range of 400-600 K, while proximal ejecta have only slightly elevated T of 250-300 K. If the early Moon was substantially hotter (thermal gradient of 15 K/km during the first 0.2-0.5 Gyr), an average ejecta temperature reaches 600 K at distances > 2400 km from a crater center, and is above 400 K at smaller distances. Immediately after deposition, the ejecta blanket is a mixture of materials with a variety of temperatures – from cold unshocked fragments to high-temperature melt lumps. These temperature-spikes within the blanket are equilibrated quickly due to the heat exchange between hot and cold fragments. Then, the whole layer cools slowly with time. Ejecta cooling time τ after its deposition depends strongly on the ejecta blanket thickness H (τ~ H2/α, α is a thermal diffusivity, α≈10-6 m2/s), which, in turn, increases quickly with the projectile size Dpr increase (H~Dpr3). Thus, gas losses due to an elevated post-deposition temperature is not as important for small impacts (with fast cooling rate), however it may lead to a partial or total resetting of the K-Ar system in large, basin-forming impacts.


Imbrium (~3.85 Ga) ejecta blanketed the area extending for more than 800 km outward. Many of the Apollo lunar samples are believed to be contaminated by or are Imbrium ejecta. For this basin, a 769 km transient cavity diameter was estimated. Although it is quite difficult to estimate a projectile diameter for this large basin, it would be in the range of 130-200 km, depending on the impact angle, impact velocity, and thermal gradient within the Moon at the moment of the impact. Nonetheless it is plausible to suggest that during this event temperatures within the ejecta blanket could have been sufficient to have cause the K-Ar resseting of unshocked (and cold) rock material that was included within it.


3.  Ries crater study and genesis of suevite (in collaboration the Natural History Museum in Berlin and with the Ries crater Museum, Noerdlingen)

A recent hypothesis (Artemieva et al. 2012) suggests that suevite formation took place by (post-)impact interaction of water with an intra-crater melt pool. To model this scenario, we start with a 250-m thick and 6-km-radius flat layer (possible size of the Ries melt pool), in which melt is thoroughly mixed with water and solid fragments. Initial temperature of the mixture T0 is 800 K; water content varies between 2 and 10 wt%. We also assume that water is instantaneously vaporized by heat exchange with melt, reaches an equilibrium temperature, but has no time/space to expand. It means that the vapor pressure is about 0.4-0.8 GPa, the fuel-coolant (FC) explosion occurs and initiate a flow which is similar to ignimbrite flow in volcanic explosions. Particle temperature within the flow decreases slowly and, at the end, is ~10-15% below T0. This final temperature approximately corresponds to the suevite temperature (>575ºC) after deposition, as derived from geophysical studies. The flow base is 10-100 times denser than the atmosphere, i.e., it is a gravitationally driven flow. After 90 s, the basal layer density decreases below atmospheric density. It means that the uppermost layer of suevite may be layered and graded.

The thickness of deposits from the FC-driven flow is consistent with observations. High water content of 5 or 10 wt% results in a rampart shape of ejecta deposits (thickening at some distance from the crater rim). As the flow is gravity-driven, any non-flat features on the surface may lead to non-homogeneous deposits (thicker in valleys, thinner on hills). Although particle sizes vary from 10 µm up to 10 cm (melt bombs), final deposit is non-graded except for the upperost 50 cm.



Artemieva N. 2011. Distal impact ejecta - an efficient tool to study ancient impacts? (invited). American Geophysical Union Fall Meeting, San Francisco, 5-10 December 2011.

Artemieva N., Morgan J. 2011. Global ejecta from Chicxulub: spherules, shocked quartz, and more. LPSC-42. Abstr. #1180.

Artemieva N. and Simonson B. 2012. Elucidating the formation of Archean-Proterozoic boundary spherule layers. LPSC-2012, submitted.

Artemieva N., K.Wünnemann, D.Stöffler3, and W.U. Reimold. 2012. Ries suevite – plume ejecta, melt flow or something else? LPSC-2012, submitted.

Deutsch A., Artemieva N., Schulte P., Berndt J. 2011. Heureka! Projectile matter in the Cretaceous-Paleogene (K-Pg) boundary event bed – n the benefit of  REE (invited). GSA Fall meeting, session T21.

Deutsch A., Artemieva N. 2012. Tracking down tracers of the Chicxulub projectile in K-Pg boundary deposits. LPSC-2012, submitted.

Morgan Joanna, Natalia Artemieva2, Claire Belcher, Tamara Goldin, Elisabetta Pierazzo, Guillermo Rein, Rory Hadden (in preparation) Revisiting wildfires at the K-Pg boundary, in prep.





  1. Artemieva N. and Pierazzo E. (2011) The Canyon Diablo impact event: 2. Projectile fate and target melting upon impact. M&PS 46, 805-829.
  2. 2.     Meyer C. and 11 co-authors (2011). Shock experiments in support of the Lithopanspermia theory. The influence of host rock composition, temperature, and shock pressure on the survival rate of endolithic and epilithic microorganisms. M&PS 46, 701-718.


Conference abstracts

Artemieva N., Morgan J. 2011. Global ejecta from Chicxulub: spherules, shocked quartz, and more. LPSC-42. Abstr. #1180. 

Artemieva N., Morgan J. 2011. Modeling the formation of the global K-Pg layer. 74th Annual meteoritical Society Meeting. Abstr. #5065. 

Artemieva N., Fernandes V. 2011. Impact Ejecta Temperature Profile on the Moon — What are the Effects on the Ar-Ar Dating Method? 74th Annual meteoritical Society Meeting Abstr. #5137.

Morgan J., Artemieva N., Belcher C., Goldin T., and Pierazzo E. 2011. Revisiting fires at the K-Pg boundary. 74th Annual meteoritical Society Meeting. Abstr. #5163. Artemieva N. 2011. Modeling the Chicxulub ejecta (invited). Fragile Earth. Geological Processes from Global to Local Scales, Associated Hazards & Resources. Munich, 4-7 September 2011.  

Deutsch A., Artemieva N., Schulte P., Berndt J. 2011. Heureka! Projectile matter in the Cretaceous-Paleogene (K-Pg) boundary event bed – n the benefit of  REE (invited). GSA Fall meeting, session T210.  

Artemieva N. 2011. Distal impact ejecta - an efficient tool to study ancient impacts? (invited). American Geophysical Union Fall Meeting, San Francisco, 5-10 December 2011.


Awards and Honors 

Berlin-Brandenburg Academy of Science Prize


Research Year: 
Page maintained by
gmn [at] (G. Nelson)

PSI, a Nonprofit Corporation 501(c)(3), and an Equal Opportunity/M/F/Vet/Disabled/Affirmative Action Employer.
Corporate Headquarters: 1700 East Fort Lowell, Suite 106 * Tucson, AZ 85719-2395 * 520-622-6300 * FAX: 520-622-8060
Copyright © 2019 . All Rights Reserved.