This is a review of:
Andrews-Hanna, J. C., M. T. Zuber, et al. (2010). "Early Mars hydrology: Meridiani playa deposits and the sedimentary record of Arabia Terra." Journal of Geophysical Research (Planets) 115: 06002.
This paper outlines a scenario in which groundwater movement on Mars results in massive sedimentation throughout Arabia Terra and elsewhere. The author uses geophysical groundwater modeling to constrain the possibilities of groundwater flow on Mars - and many of the areas the model predicts to accumulate sediment are observed to have large deposits of sediment. The correlation between the model prediction for Meridiani being one of the areas of high groundwater flow and the existence of the Meridiani sediments has convinced many people that a groundwater origin is the most likely explanation for the Meridiani sediments.
The work assumes a very arid Mars (drier than the Atacama) with very long groundwater residence times. It assumes higher rainfall in the equatorial region which then serves as the source of groundwater to the rest of Mars.
Overall I really enjoyed this paper. It deals with big topics and addresses them in a process oriented manner setting forth a clear testable hypothesis. However, I do not agree with the conclusions and feel that the paper misses many important points.
The paper assumes a Mars similar to today except warmer, with temperatures above freezing in the equatorial regions which . This fails to account for the fact that the obliquity of modern Mars is rare. In fact ancient Mars was more frequently at higher obliquity (>30deg) than not. A more realistic model should take into account the average obliquity of Mars during the late Noachian/early Hesperian. Of course, if one does take this into account, the equatorial regions no longer are the warmest spots on the planet and would be unlikely to be source regions for rainfall. Instead these regions would be cold and icy.
The paper argues that recharge of valley networks is necessary to explain their origin and thus precipitation and melting must have occurred in the low latitude region. However, it is clear that valley network formation has occurred in more recent times (most likely during cold, dry conditions) but did not result in massive sedimentation. Also it is unclear that the volume of water needed to carve the valley networks compares with the volume needed to deposit the hundreds of thousands of km3 of sedimentary material in Arabia and Meridiani.
There are obvious problems with the assumptions made in the paper about permeability, porosity, and the lack of any regional groundwater barriers. There are also problems with the fact that some of the intracrater layered deposits are higher than the rims of the craters that they lie in the middle of. The paper argues that so much sediment accumulated that it filled the available basins, and the water table rose high enough to start sedimentation in the plains around the craters. But I want to focus on four other specific things here.
Firstly, the lack of playas. There are no discernable or identifiable playas anywhere to be found in this region. Given the huge amount of evaporite deposition, it would seem likely that in some areas playas would form. There are none to be found so far, and the lack of these playas means there is no direct evidence for evaporative environments.
Secondly, the intracrater sediment mounds are almost universally sitting in the very middle of their respective craters. They have been differentially eroded around the edges of the crater leaving the central mound intact. If the water cementing these deposits was sourced from below, one would imagine that the areas of strongest cementation would be near the water source at the bottom of the crater and around the edges. Thus the sediment would grown from the outside inward. The reverse is true here, and it is unclear why the central portions of the basin should be so much more erosionally resistant than the outer edges.
Thirdly, the dips of these materials are too steep to account for deposition via liquid water. Instead layers frequently drape topography. This would be consistent with eolian reworked material - but then it would require all of the material to be reworked by eolian processes. This subverts the main mechanism for sedimentation argued for in the paper where it suggests that eolian materials are cemented in place by rising groundwater. If they are subsequently reworked by eolian processes, then it is not clear why they would remain in the same place and not be redistributed at the next shift in martian climate.
Fourthly, it seems that erosion is neglected in this model during the crucial time period in which the sediments are being deposited. Calculated deposition rates neglect any loss via erosion. If eolian reworking is indeed a part of the model, then it seems very likely that erosion could have been significant. Estimates of erosion at Meridiani at this time are around 0.8x10-5 m/yr (Golombek et al. 2006, Hynek and Phillips, 2001). This is on the same order magnitude as the deposition rates especially later in the model, and might be sufficient to reduce the deposition rate to zero.
The first problem here is the source of sulfur. If primitive martian rocks contain on average 1800 ppm sulfur, then you need to completely leach 100 times the volume of rock to get the equivalent volume of sulfate-rich material (20% Sulfate). It seems at such low concentrations, that in order to get enough sulfate, the amount of basalt weathering needed would create a super-alkaline solution or at least buffer the pH at neutral values. The salinity values used in this model assume Earth-like weathering processes to collect and concentrate salts. Without large sedimentary evaporite deposits from ancient oceans it seems unlikely that a brine with 80% of the salinity of sea water could be formed simply by interaction with basalt - no matter how long that interaction took place. Also, if it did, then it wouldn't be a sulfur-rich brine. In fact Na-Cl should play a much more dominant role.
The second problem is the bottom-up style of weathering. If the sediments were laid down by a rising water table, it is essential that sediments previously deposited would be saturated with water as the water table rose. Thus the most soluble components would be redissolved and redeposited on the surface. This repeated process would result in stratification of the deposit by solubility with the topmost portions being entirely composed of the most soluble minerals. This is the opposite of what we see at Meridiani, and it also is not indicated by remote sensing results from Arabia. Instead it seems that the sediments are well mixed with clastic material and the most soluble minerals are intermixed with the less soluble ones.
The third problem is acidity - since the basaltic aquifer should not yield acidic solutions - This is dealt with in the previous blog post - along with the Fe-oxidation theory published by Hurowitz et al..
Finally, the age of the Meridiani deposits place their formation at the end of the Noachian into the early Hesperian (Hynek and Phillips, 2008). Thus sedimentation ceased near 3.5 Ga. This is substantially after the dynamo shut down and loss of the magnetic field which was likely ~4.3 Ga. Atmospheric loss processes were well underway, and climate models have had a difficult time simulating any martian climate that is warm enough to host liquid water on it surface at any time (see Tian et al. 2010). This period also postdates the late heavy bombardment which likely sloughed off much of the martian atmosphere. Thus it becomes very difficult to argue that conditions on Mars were warm and wet for any prolonged period during this time.
The best argument for warm/wet conditions come from the existence of valley networks in the region, and evidence for surface flow of water within the Meridiani bedrock. However, there have been multiple arguments made that suggest valley networks do not require warm/wet conditions - and there are several examples of late Hesperian and early Amazonian valley networks that most likely did not form in a warm/wet climate.