NASA Technical Reports Server (NTRS) 19940017209: The Martian valley networks: Origin by niveo-fluvial processes

2 Pages·0.23 MB·English

by  NASA Technical Reports Server (NTRS)

22
Workshop onEarl),Mars: How Warm and How Wet?
spheric lifetimes, will need to take into account the rate at which "-Some problems with the sapping model will be discussed below.
they may have been supplied tothe atmosphere. Schemes analogous FirSt, the measurement of junction an#es between individual in-
to that presented for CO_, will have to be explored in order to assess terse_zling tributaries of the valley networks does not provide evi-
the absolute contribution of any potential greenhouse gas on early dence-\o refute the view that the networks were formed by rainfall/
Mars. snowmelt-fed erosion. Stream junction angles are controlled by
References: [1]W_nkeH.etal.(1992)LPSXXlll, 1489-1490. slope, stfijcture, lithology, and basin development stage, not pre-
[2] Pien D. C. (1980) Science, 210, 895-897. [3] Gulick V, C. and cipitation-J6]. Sapping requires that zones of low hydraulic head
Baker V. R. (1993) LPSXIV. 587-588. [4] Squyres S. W. (1989) somehow be established to support the gradients needed to allow
Fourth Intl. Co1_ on Mars. [5] Sagan C. and Mullen G. (1972) groundwater flow, and that zones of high hydraulic head be re-
Science, 177. 52-56. [6] Toon O. B. etal. (1980) Icarus, 44, 552- charged, presumably by precipitation. Additionally, some of the
607. [7] Pollack J. B.et al. (1987) Icarus, 71,203-224. [8] Kasting valley networl(s whose channels originate on crater-rim crests indi-
J. (1991 )lcarus, 94, 1-13. [9] Kasting J. F. (1992) Workshop on the cate that the loca[water table must have intersected the surface high
Evolution of the Martian Atmosphere. [10] Postawko S. E. and on the crater wall _ sapping was involved [3]. This would mean that
Fanale F. P. (1992) Workshop oll the Evolution of the Martian the crater was oncefilled with water, but there isno evidence, such
At_nosphere. 0 as inflowing channels, to support this condition. It should also be
noted that all the valley networks have been modified by mass
wasting processes such as gelifluction and thermal erosion.
/ In order to more full_yunderstand niveo-fluvial systems on Mars
THE MARTIAN VALLEY NETWORKS: ORIGIN BY one should study terrest_al periglacial regions such as the North-
NIVEO-FLUVIAL PROCESSES. J.W. Rice Jr., Department west Territories in the Cafi_dian High Arctic. Itis proposed that the
of Geography, Arizona State University, Tempe AZ 85287, USA. following geomorphic processes and resulting landfornas of
snowmelt-fed rivers beused to explain the dendritic valley networks
The valley networks may hold the key to unlocking the on Mars.
: paleoclimatic history of Mars. These enigmatic ]andforms may be The Mecham River near Resolute, Northwest Territories, pro-
regarded as the martian equivalent of the Rosetta Stone. Therefore, vides an excellent example of s_ream action and valley development
a more thorough understanding of their origin and evolution is inthe periglacial realm. The are a_isunderlain bycontinuous perma-
required. However, there is still no consensus anaong investigators frost and mean monthly air temperatures are below zero for 9-
regarding the formation (runoff vs. sapping) of these features. 10months ayear. The Mecham Riyer has 80-90% of itsannual flow
Recent climatic modeling [lr precludes warm (0°C) globally concentrated ina 10-day period. Tl_s is typical for periglacial rivers
averaged surface temperatures prior to 2b.y. when solar luminosity in the High Arctic. During dais brief period of concentrated flow
was 25-30% less than prt,sent levels. This paper advocates snowmeh extensive movement of bedload occurs, sometimes with peak ve-
as the dominant process responsible for the fornaation of the den- locities up to 4m/s, causing the whole bed to be in motion [7]. This
drilic valley networks. Evidence for martian snowfall and_subse- pattern of intense activity has far gre_iter erosive and transporting
quent melt has been discussed in previous studies.It_has been potential than a regime in which rivet: flow is evenly distributed
suggested [21 that Mars has undergone periods of very high obliq- throughout the year. The dominance of _dload movement inArctic
uities, up to 45% thus allowing snow accumulations, several terls of streams helps explain the distinctive flal-bottomed forna of many
meters thick, at low latitudes as a result of sublimation from the periglacial stream valleys [8]. Thermal el:gsion and the subsequent
poles. Clow investigated the conditions under which sn_ow could collapse of river banks provides material for bedload transport and
have mehed by solar radiation by using an optical-theaZmal model deposition downstream. This process also _,ids in the development
developed for dusty snowpacks [3]. Itwas found that_die low thermal of the broad flat-floored valleys. The pernm fl:ost also favors the flat-
conductivity of snow and its partial transparency to solar radiation floored valley profiles because it provides/l near-surface limit to
can result in subsurface melting despite surface temperatures ,,veil downward percolation of water, thereby promoting runoff [9]. An-
below freezing. Melting and subsequent runoff can occur at atmo- other interesting feature of these periglacial rivers is that they lack
spheric pressures as low as 30-100 mbar [3]. Can" showed that if a pronounced channel on their floors. This holds true for valleys
streams 2mdeep or larger can be initiated and sustained, then flows eroded into either bedrock or unconsolidated d_bris.
up to afew hundred kilometers long can be established, even under Other work [10] indicates that fluvial proces_s have often been
present-day climatic conditions [4],Therefore, based on the above- underestimated inperiglacial regions. Budei illusi_rates this point in
mentioned work, it seems logical to the author that snowfall and Spitsbergen, where he pointed out that ground ice breaks apart the
subsequent snowmelt has many advantages to other explanations rocks and prepares them for fluvial action. Periglacial rivers do not
for the formation of the valley networks. need tocarry out new erosive action but need only n_elt the eisrinde
Ithas been argued that the valley networks were formed prima- and transport the shattered debris. The eisrinde is cdla_posed of the
rily by groundwaterseepage. This is based on the measurement of upper frozen and highly shattered layer of the perm:afrost. Rivers
junction angles between intersecting tributaries and on morphologic operating under this regime can deepen their beds rapidly; down-
characteristics that appear to suggest headward extension through cutting rates on the order of 1-3 nv'1000 yr over the lhst 10,000 yr
basal sapping [5]. The evidence for sapping is insome cases con- have been estimated for Spitsbergen [10].
vincing (i.e., Nirgal Vallis). but it does not explain many of the References: [1] Kasting J. F. (1991) Icarus, 94, 1-13.
dendritic valley systems, e.g., those located in the Margaritifer [2] Jakosky B. M. and Cart M. H. (1985) Nature, 315, 559-561.
Sinus region. [3] Clow G. D. (1987) Icarus, 72.95-127, [4] Cart M. H. (1983)
LPI Technical Report 93-03. Part I 23
Icarus, 56, 476-495. [5] Pieri D. C. (1980) Science, 210, 895-897. The most promising nonclimatic evidence for main sequence
L
[6] Schunma S. A. (1956) GSA, 67, 597-646. [7] Cook F. A. (1967) mass loss from the early Sun isthe direct observation of similar mass
Geo. Bull., 9, 262-268. [8] French H. M. (1976) The Periglaeial loss from young main-sequence G stars. Detection of stellar mass
Environment. 309. [9] Washburn A. L. (1980) Geocryology, 406. loss f.rom late dwarfs aithe predicted rate (less than-I 0-I° Mo yr-I)
by optical techniques is generally not possible. However, in one
[10] Budei _41_7t_) Ct_imatiel_4jeo_orp_olo _. _4.
JL !, unique ease where itcould be measured, an outflow 1000x that of
the present Sun was found in a K2V dwarf [11]. Recently, huge
winds have been reported from several M dwarfs [12]. This tech-
MARS AND THE EARLY SUN./'D. P.Whitmire l,L. R.Doyle'-,
R. T. Reynolds 3,and P. G. Whitm_tn LIUniversity of Southwestern nique involves detections over awide range of radio and nlillimeter
Louisiana, Lafayette LA 70504-4210, USA, -'SETIInstitute, NASA wavelengths and the fact that free-free emission from an optically
Ames Research Center, Moffett Field CA 94035, USA, 3Theoretical thick wind has acharacteristic spectrum inwhich the flux ispropor-
Studies, NASA Ames Research Center, Moffett Field CA 94035, tional to the 2/3 power of the frequency. As afirst step in extending
USA. this technique to solar-type stars we have recently used the VLA to
obtain the radio emission at2 and 6 cm infour nearby young G-type
Global meantemperatures near 273 Kon early Mars are difficult stars.
to explain in the context of standard solar evolution models. Even In the second model-(ecliptic focusing) we assume standard solar
assuming maximum CO, greenhouse warming, the required flux is evolution. Young G starsare often observed to be heavily spotted
- 15% too low.[ 1]. Here we consider two astrophysical models that (10-50%). In contrast tomature Gstars like the Sun, which typically
could increase the flux by thisamount. The first model is anonstand- have only amaximunl covef'age of-0.1%, the net effect of star spots
ard solar model inwhich the early Sun had amass somewhat greater on young G stars is to reduce!.he radiated flux at the location of the
than today's mass (1.02-1.06 Mo). The second model is based on spot. Since the total stellar ln_minosity is determined by nuclear
a standard evolutionary solar model, but the ecliptic flux is in- reactions inthe core, thefluxnau.stincreaseinregionswithoutspots.
creased due to focusing by an (expected) heavily spotted early Sun._ -Such variations influx are observed on short (days) and long (years)
The relation between_stellar mass M and luminosity L for stars timescales [13]. These observatir0s measure the anisotropy in the
near 1Mo is L- M4.7s12].Ifthe Sun's original mass were larger than distribution of spots. Amore significant effect would be the average
at present, the early planetaL'y flux would be further increase,:.] due increase in the equatorial flux if the?time-averaged location of the
to migration of orbits. Isotropic mass loss does not produce a_torque spotted regions was nearer to the stellar poles than to the equator.
on aplanet and so angular momentum is conserved. Conse_luentiy, This is not the case in today's Sun, but is observed tooccur in young
semimajor axes increase inversely with mass loss and the flux is stars such as the G2V star SV Camelopardalis, in which there is a
proportional to M6.75.To increase the flux atMars by I5% requires -10% coverage, localized in latitude andlongitude, Ioward one of
that the Sun's mass be _>!.02 Mo. On the otherhand, the flu xcannot the poles.
be so large (1.1× that of the fluxat IAU today) that Earth would have We have investigated asimple model in.which polar cap block-
lost its water [3]. This imposes an upper mas_ limit of i.06 Mo. ing focuses the stellar flux in the equatorial plane. The equatorial
Nonclimatic evidence for mass loss of this magnitude might be fluxcan be enhanced amaximum of afactor of 2over the uncapped
found in the ion implantation record of meteorites and Moon rocks. case. For atime-averaged polar coverage of 10% the equatorial flux
Such evidence does exist, but is inconclusive due to uncertainties in enhancement factor is 1.17. Refinements in this model and areview
exposure times and dating [4.51. The dynamical record of adiabatic of the relevant observational data will be presented.
mass loss is alsoinconclusive.The a_liabatic invariance of the action Acknowledgments: D.P.W. and P.G.W. thank the Louisiana
variables implies that the eccentricities and inclinations of plan- Educational Quality Support Fund for partial support of this work.
etary orbits remain constant asJhe semimajor axes increase. The D.P.W. also acknowledges aNASA Ames/Stanford ASEE sunanaer
dynamical drag of the wind would have no effect on planets, but research fellowship.
would cau sea net inward migration ofbodies of sizes less than about References: [l]KastingJ.(1991)Icarus, 94.1-13 [2]IbenI.
1 km [6]. Whether the cralering record is consistent with this (1967) Annu. Rev. Astron. Astrophys. 5. 571-626. [3] Kasling J.
dynamical consequence is unclear. Mass loss could also be an (1988) Icarus, 74,472-494. [4] Caffee M. et al. (1987)Astrophys.
additional process contributing to bringing organics into the inner J., 313, L31-L35. [5] Geiss J. and Bochsler P. (199 !) The Sun in
solar system. Time (C. Sonett etal., eds.), 99-117, Univ. of Arizona. [6] Whilmire
Amass loss of 0_.1Mo has been suggested as an explanation for D. et al. (1991) Iml. Co_ on Asteroids, Comets. Meteors. 238,
the depletion of Li in the Sun by 2 orders of magnitude over Flagstaff, Arizona. [7]Graedel T.etal. (1991) GRL, 18. 1881-1884.
primordial values [7]. However, this explanation has been reconsid- [8] Swenson F. and Faulkner J. (1992) Astrphys. J., 395,654-674.
ered by [8]. who find that mass loss cannot explain the depletion of [9] Pinsonneault M. et al. (1989) Astrophys. J.. 338. 424-452.
LiinHyadesGdwarfs. Although it is generally believed that young [i0] Bohigas J. et al. (1986) AAS. 157, 278-296. [11] Mullan D.
G stars are_spun down by mass loss, most models are insensitive to et al. (1989) Astrophys. J., 339, L33-L36. [12] Mullan D. et al.
the total mass loss required [e.g., 9]: An exception is the model by (1992) Astrophys. J. [13] Radick R. (1991) The Sun in Time (C.
[10] which predicts amass loss comparable to our lower limit. S0nett et al., eds.), 787-808, Univ. of Arizona.