2 0 0 9 P a r i s / M e u d o n

IWCMO Conference


Talk at the IWCMO meeting, 19th September 2009



he purpose of this talk is to present a recently identified kind of dust storm on Mars, the 把ross-equatorial dust storms, and to compare them with what will be called here the 把lassical dust storms. During the last ten years, our knowledge about planet Mars has been greatly improved by the launch of several orbiters (Mars Global Surveyor, Mars Odyssey, Mars Express and so on). During the meantime, the technical tools available to amateur astronomers have known a big increase in quality, thanks to the introduction of CCD cameras, then webcams, and finally fast uncompressed video cameras. We propose to review the last 10 years dust activity of Mars thanks to both professional and amateur data.


1 Two models of dust storms

The model for the 把lassical dust storms has been historically drawn to explain the global or encircling dust events that are sometimes observed on Mars (like in 1909, 1956, 1971…). It痴 declined with the following criterions :

   They start in the southern hemisphere of Mars, from favourable sites mostly located around the Hellas basin ;

   They begin preferentially during late southern spring or southern summer (Ls 250-270ー), as we have to wait for the atmospheric pressure痴 increase to generate winds strong enough to lift dust clouds. This season corresponds to the passage of the planet at its perihelion ;

   The main mechanism responsible for the global diffusion of dust is the ォ positive feed-back サ. The dust lifted in the atmosphere absorbs the infrared emission from the Sun, heating the upper atmosphere of Mars; this increase in temperatures enhances the dominant winds, and they lift finally more dust in the atmosphere, and so on.

   The unique spring / summer Hadley cell is the second mechanism that allows global spreading of dust. A Hadley cell is a vertical, latitudinal, atmospheric circulation in a planet痴 atmosphere. On Mars, this unique cell has its ascending branch near 60ーS, and its descending branch near 60ーN.

Over the last ten years, the global dust storm of 2007 fits perfectly in this model. The global 2001 storm also fits nicely, with the notable exception of the seasonal criterion, the storm being raised one season earlier after the southern spring equinox.


The model of the #cross-equatorial dust storms has been inferred by scientists working on the Mars Global Surveyor (MGS) data. They observed in 1999 the southward descent of a dust activity raised on the northern polar region over Mare Acidalium, that crossed the equator in a few days and triggered more important dust activity in the southern hemisphere. This model responds to the following criterions:

   The atmospheric circulation in the northern polar region (NPR) during fall and winter is governed by three stationary #waves created by the topography (see below), over Mare Acidalium, Utopia, and Arcadia. These are 敗torm zones where cloud fronts will be generated, carrying both dust and water vapour ;

   They develop during two #seasonal windows in the northern fall (Ls 210-240ー) or northern winter (Ls 310-350ー). Around the winter solstice (Ls 270ー), there is only one wave circulating and the winds are not strong enough.

   The cloud fronts are pushed southward if a certain number of local conditions are fulfilled. If so, they will travel to the equator, and eventually raise more dust clouds in the south hemisphere of Mars.

Figure 1 Left : the cross-equatorial model : in the lowland of Acidalium during fall or winter, a storm zone exists, organized with one high pressure system over Tharsis and one low pressure system over Acidalium. Between the two, southward winds push a dust fronts toward the equator. Right : image of such a southward front by MGS. Copyright NASA/JPL/MSSS.


2Focus on the storm zones

The three storm zones are created by the eastward polar jetstream over the lowlands of the northern hemisphere, located just after higher terrains. As the jetstream arrived over the lowland, it痴 deflected toward the south, and this generates anticyclonic circulation at the south-west of the scene, and cyclonic circulation at the north-east. Between the two systems, dominant winds are southward; they correspond also to strong returns of descending branches from the Hadley cell. Local winds are then deflected slightly to the west, because of trade winds.

Because of topographic differences, the storm zone over Mare Acidalium is the stronger of the three, and the one responsible for the vast majority of cross-equatorial storms. The Arcadia storm zone is the weakest.


Figure 2: simplified wind pattern for the NPR during mid-fall and mid-winter, on a relief map with albedo features superimposed. Dashed arrows are paths for the cross-equatorial storms.



3 Focus on the Acidalium atmospheric wave

Scientists said that the waves in the NPH (north polar hood) show a #two sols periods. This means that each storm zones will generate clouds fronts every two days when the wave arrives. This explains the changes observed on the polar hood on amateur images over the apparitions of 2005 and 2007 (up to late November). The NPH over Acidalium show a 派igh phase on one day, before or after the wave, when the latitude of its southern border is near 40-50ーN. On the second day, a low phase is observed, as the NPH show a shape : this is the cloud front. The point of the V shape is located at the east of Nilokeras and can descend as low as 15-20ーN, although 30ーN is more common.

Small dust clouds will be preferentially generated only during the 罵ow phase, at the point of the V, on a very narrow region located near Nilokeras and Niliacus Lacus.

This pattern of activity is precise and repeatedly observed from year to year during the critical seasons.



Figure 3 : amateur images of the Acidalium atmospheric wave reveal the two days period of the wave, with an alternation of high an low phase of the NPH. On this sequence in 2007, small dust clouds are raised during one low phase on November 2nd. Note the V shape of the cloud front on this day and on the 31 of October.

Figure 4 : the 50-hours oscillation of the NPH during October 2005, measured at the longitude 35ー on amateur images show the wavy nature of the circulation there. Two dust events have been imaged, on the 13th and the 17th, corresponding also the arrival of the cloud front during the 斗ow phase (arrowed). The cloud on the 17th marks the beginning of the regional storm of 2005 (see below).

Figure 5 : another example, taken by the MGS orbiter on march 2000, 19th (Ls 321ー). This again shows the V shape dust front, and small dust clouds raised on Nilokeras and Niliacus Lacus. Note the striking resemblance with Dave Tyler痴 image above on 2 november 2007.


4 The cross-equatorial dust storm of July 2003 (Ls 211-220)

The alert was sent by Don Parker with images of his on the first day of July, that show dust activity over Hyapigia Viridis (northern Hellas). The subsequent evolution was a nice regional event, expanding mainly to the east, as did the 2001 storm, but with also interesting activity at the west.

Analysis of MGS data show that this episode was triggered by a small dust cloud travelling southward from the Utopia storm zone. Dust clouds of this nature are easily missed by the partial amateur coverage (this one was certainly well observed from the Atlantic Ocean!) and also because they are very small.


Figure 6 : RG 610 images (R+IR) from Don Parker on July 1, 2, 3, 4, 6, 2003, showing the core of the regional dust event. South is up.

Figure 7 : MGS images showing the descent of the precursory dust cloud, on June 28, 29, 30, and July 1st. North is up, with Syrtis Major on the left or central sides of the frames. Credit: NASA/JPL/MSSS. Processing: C. Pellier.


5 The cross-equatorial event of December 2003 (Ls 314-320)

Again, the alert is sent by Don Parker with images on December 13th, that show a very bright dust cloud over Chryse. Analysis of data from the Thermal Electro Spectrometer (TES) instrument of the MGS probe, which provides thermal infrared images of Mars, shows that this is the result of a long-lived low scale dust activity near Nilokeras during the first weeks of December.


Figure 8: frames taken from a TES movie (now unavailable). On December 12th, the activity bursts near the equator in Chryse, but the northern path of dust from Acidalium is still visible. On December 16th, again the northern path of dust is visible while bright dust clouds expand on the southern hemisphere.

Figure 9: first amateur images by Don Parker on December 13th, 2003. Dust is crossing the equator from Chryse to Mare Erythraeum.


6 The cross-equatorial dust storm of October 2005 (Ls 308-320)

This regional event has been the first one completely imaged by amateurs, from the very first dust clouds. Alert day is on October 18th, 2005, with several images showing a triangle-shape dust cloud over Eos by many observers. But the first dust cloud is visible on the 17th, descending from Niliacus Lacus on images in Europe by Marc Stemmelin, and on the USA by Bill Flanagan, Don Parker and Ed Grafton. Activity remained during two weeks with an expansion highly comparable to the December 2003 event (the two events are in fact very similar, raised almost at the same location on the same season.

Figure 10 : sequence of the beginning of the 2005 regional storm from amateur images. Jim Phillips’image on the 18th show the dust storm when it痴 first identified. But images taken the day before reveal a very small dust triangle on northern Chryse, that is the true start of the episode (not present on the 16th, see Flanagan images). The subsequent strong southern movement of the storm is another evidence of its cross-equatorial nature.


7 The cross-equatorial storm model reviewed with observations

Observations bring two nuances to the model :

1)   Dust clouds do not travel. As shown by Masatsugu Minami, Director of the OAA Mars section, dust clouds are stationary during the day time [See Masatsugu own intervention]. Dust clouds look to be generated by daytime convection. But, there must be some lifted dust travel with the dominant winds, that will lift convective dust clouds each day farther from the original site.

2)   The main dust front created on the storm zone rarely manages to escape and travels south. Cross-equatorial dust activities are generated by small dust clouds lifted in advance of the main front.


Figure 11 (left): dust activity at the Acidalium storm zone is much more common in a form of small clouds raised in advance of the main cloud front.

Figure 12 (right): If dust clouds don稚 travel, then we must imagine another way of dust diffusion to explain why they can follow precise paths and are not always andom.


8ー Cross-equatorial storms vs. classical storms

During the last ten years, a good number of dust events have been observed, from global to local scale, storms that have lasted months or only days. The majority of the observed events belongs to the northern polar region, among them the three big regional events observed in 2003 and 2005. But, it is quite curious to see that no cross-equatorial storm manage to evolve into a global storm, when the two global events observed (2001 and 2007) belong clearly to the 把lassical storm models, having been originally raised by southern dust clouds.

One would then try to look if there is any reason that might prevent cross-equatorial storms to reach the global stage (even if we might not consider this as being completely impossible). Here are some propositions on this topic:

1)   The most favourable period for global storm occurs near perihelion and northern winter solstice (Ls 250-270ー), but this is also a period when northern dust activity is very low, with no southward path opened, as scientists showed ;

2)   Any cross-equatorial dust activity must travel with the descending branch of the Hadley cell, closer to the ground, while southern generated dust clouds have an immediate access to the ascending branch, so they are sent to the higher atmosphere.

3)   Preferential starting sites for global storms lie away from the majority of the cross-equatorial storm paths, so those one can hardly trigger clouds where they have more chance to go global. This because the Hellas basin, which as been recently the most favourable site, is not located at the same longitudes than the Mare Acidalium northern storm zone, which is by far the most active one.


Figure 13 : schema of the Hadley cell, with respective path for southern and northern dust clouds. Figure 15 : example of cross-equatorial storm trapped into Valles Marineris (MGS, 22 march 2000, Ls 322ー). Those clouds have more chance of being stopped by unfavourable topography. The 2000 activity did not survive the trap, but the 2005 activity did so, nonetheless.


Figure 17 : map of the planet showing the main paths of cross-equatorial storms as observed over the last decade, and the starting sites of the 2001 and 2007 global storms. There has been only one cross-equatorial storm near the favourable site of Hellas for global dust storm. The majority of northern dust clouds arrive on Mare Erythraeum, away from Hellas.



Further investigations might be carried out in two directions:

1)   A study of past dust storms (before the 1990) to try figure out which one might have been a cross-equatorial event, as they look to be quite frequent

2)   A comparative study of cross-equatorial storm seasons from martian year to martian year


Christophe PELLIER, SAF Mars Section Director

Back to the Index