Technical Comments

Response to Comment on "Athabasca Valles, Mars: A Lava-Draped Channel System"

See allHide authors and affiliations

Science  20 Jun 2008:
Vol. 320, Issue 5883, pp. 1588
DOI: 10.1126/science.1155124


The recent geologic history of Athabasca Valles, Mars, is controversial. Some studies report ice-rich sediment in its channels, whereas others find only lava. Data from the High-Resolution Imaging Science Experiment camera now confirm that, although certain features exhibit a superficial similarity to ice-related landforms, solidified lava coats the entire channel system.

Jaeger et al. (1) reported that in geologically recent times a voluminous lava flow inundated and then drained out of Mars' Athabasca Valles channel system, leaving it coated by a thin layer of residual lava that hardened to rock. Some earlier studies interpreted this lava carapace as the frozen dregs of the aqueous flood(s) that carved Athabasca Valles (24). However, data from the High-Resolution Imaging Science Experiment (HiRISE) camera onboard the Mars Reconnaissance Orbiter (MRO) spacecraft refute the ice hypothesis. The HiRISE images, which now cover >10% of the channel system, show crosscutting and kinematic relationships among ring-mound landforms (RMLs) in Athabasca Valles that indicate they formed via deep-seated, fixed-source steam explosions through the thin, solid, and rafting crust of an active lava flow (1). In other words, the RMLs are hydrovolcanic constructs—they are a suite of crater- and cone-shaped landforms produced by the explosive interaction of surface lava and underlying groundwater—and they are intrinsic to the lava flow in which they occur.

Page (5) claims that Jaeger et al. (1) mistook secondary features that postdate the flow “by many millions of years” for primary volcanic landforms. He contends that two types of features (RMLs and polygons) overprint impact craters that postdate the flow. We dispute both of his arguments, though for different reasons. The claim that polygons overprint impact craters is dubious because the supporting evidence comes from an entirely different region of Mars located >2000 km from Athabasca Valles in seemingly unrelated terrain (Fig. 1). It is unclear what, if any, relationship the features depicted in figure 1, F to I, of (5) have to those discussed by Jaeger et al. (1). Moreover, the polygons invoked by Page are fundamentally different from those in Athabasca Valles. The former have raised edges and low centers, whereas the latter have low edges and high centers. The two are not analogous structures, and the comparison is unjustified.

Fig. 1.

Viking Mars Digital Image Model mosaic of Elysium and Amazonis planitiae, Mars. The giant shield volcanoes Elysium Mons and Olympus Mons are labeled, as is the Athabasca Valles channel system. The study area depicted in figure 1 of Jaeger et al. (1) is outlined by a black box. The locations of MOC images R23-00683, R14-01144, and S09-02331 [and, therefore, of figure 1, F to I, of Page (5)] are indicated by a black star.

We submit that the remaining argument raised by Page is due to a misunderstanding of impact dynamics and stratigraphic (crosscutting) relationships. In figure 1 (A to E) and figure 2 in (5), Page shows examples of small, noncircular impact craters in Athabasca Valles. In one or two locations, the rim of an impact crater deflects around an RML. It follows that either the impact was not energetic enough to fully destroy the RML or the RML postdates the impact. Page (5) favors the latter interpretation, claiming that the impact should have “obliterated these landforms on contact.” However, such destructive power is only true of high-velocity primary impacts, not secondaries, which may have much lower velocities (68). Secondaries form by the fallback of ejecta thrown from primary craters. As such, they are relatively low-energy events that commonly produce noncircular craters. This is particularly true of proximal secondaries like those in Athabasca Valles, which are mainly associated with the nearby primary crater Zunil (9). One only need look ∼50 m west of the region shown in figures 1A and 2A of (5) to see that these craters have irregular shapes regardless of their proximity to RMLs, as there is a comparably noncircular impact crater outside the swath of RMLs (Fig. 2).

Fig. 2.

Subsection of HiRISE image PSP_002292_1875 showing two impact craters (A and B) and several RMLs. The gray box indicates the location of figure 1A of Page (5). The irregular shape of these craters is not indicative of postimpact modification. Rather, it is a characteristic of secondaries formed by the relatively low-velocity, low-energy impact of ejecta from a nearby larger crater, probably Zunil (9).

It is also worth noting that the impact craters in question are similar in scale to those produced by conventional aerial bombs. The U.S. Army and U.S. Air Force conducted a series of experiments between the years 1935 and 1976 in which they bombed active and inactive lava flows on the Mauna Loa shield volcano in Hawaii (10). Using 900-kg Mk84 bombs, they produced craters up to 30 m in diameter with visible damage extending as much as 50 m from the craters. [For comparison, the majority of Zunil secondaries are 10 to 50 m in diameter (9).] The aerial bombing experiments showed that 100-m-scale volcanic spatter cones could withstand even multiple explosions of this magnitude (10). Thus, we conclude that RMLs too can survive small, relatively low-energy impacts and influence the shapes of the resultant craters.

In summary, the “lava or ice” question cannot be answered by the ambiguous stratigraphic relationships presented by Page (5). Instead, definitive evidence for a lava lithology can be found in the annular moat sequence and the rafted wakes and chains of overlapping RMLs shown in figures 3 and 4, respectively, in (1). These observations demonstrate that the RMLs are intrinsic to the lava and formed while it was actively flowing; they are not intrusive ice-related landforms that postdate it. Additionally, several other data sets contraindicate ice: (i) gamma ray spectrometer data show the upper ∼1 m of material in this region to be exceedingly hydrogen poor (11), (ii) ice-stability models show that near-surface ice is highly unstable at these latitudes (12), (iii) data from the SHARAD (shallow radar) instrument onboard the MRO spacecraft suggest a lava composition rather than ice (13), and (iv) our preliminary analysis of data from the Compact Reconnaissance Imaging Spectrometer, which is also on board MRO, found no spectral features indicative of water or hydrated minerals in the region, even at sites where impact craters penetrate the flow. Thus, the conclusion that Athabasca Valles is coated by lava rather than ice appears to be robust.

References and Notes

View Abstract


Navigate This Article