MARTIAN DUST STORMS: A REVIEW ∗

WALTER FERNÁNDEZ
Laboratory for Atmospheric and Planetary Research, DFAOP/School of Physics and Centre for

Geophysical Research, University of Costa Rica, San José, Costa Rica

(Received 11 February 1998; Accepted 20 October 1998)

Abstract. A review of the dust storms observed on Mars is made. This includes the seasonal and
interannual variability of planet-encircling and regional dust storms. Although there is a significant
interannual variability, planet-encircling dust storms have been observed to form during the southern
spring and summer seasons, while regional dust storms tend to occur more frequently. Some aspects
of possible mechanisms associated with the origin, maintenance and decay of the dust storms are
also discussed.

1. Introduction

Although Mars and the Earth share some similarities (for example, axial tilt and
rotation rate), they also show marked differences between them. On Mars, the
dust raised from the surface into the atmosphere by strong winds absorbs solar
radiation and constitutes an important internal heat source. In addition, on Mars
the sublimation and condensation of carbon dioxide produces a mass flow from the
subliming polar cap to the condensing polar cap. A comprehensive understanding
of the Martian general atmospheric circulation requires knowledge of the processes
which control the seasonal and interannual variations of the injection, transport and
removal of dust, water vapor, and carbon dioxide (e.g., Zurek et al., 1992).

Planetary atmospheres offer many fascinating research topics. One of them is
the Martian dust storms, also called Martian yellow storms or Martian yellow
clouds. They have been observed for a long time (e.g., Antoniadi, 1930; Martin
and Zurek, 1993). The planet-encircling dust storms appear to form when one or
several local dust storm expand or when more than one dust storms expand and
merge (e.g., Ebisawa and Dollfus, 1986). They obscure large areas on Mars from
periods from days to months. The thermal structure of the atmosphere is modified
by the dust and, consequently, the atmospheric circulation is also perturbed. Planet-
encircling storms transport dust over long distances. In addition, the deposition and
zonal redistribution of dust may produce changes in the surface albedo and in the
surface composition and thermal characteristics (Chrristensen, 1986).

Several authors have described or reviewed the topic of Martian dust storms
(e.g., Gierasch, 1974; Zurek, 1982; Martin, 1984; James, 1985; Kahn et al., 1992;
∗ Paper presented at the Seventh UN/ESA Workshop on Basic Space Science, Tegucigalpa,

Honduras, 16–20 June 1997.

Earth, Moon and Planets 77: 19–46, 1997–1998.
© 1998 Kluwer Academic Publishers. Printed in the Netherlands.



20 WALTER FERNÁNDEZ

Martin and Zurek, 1993; Zurek and Martin, 1993). Particular case studies have also
been analysed and reported in the literature (e.g., Capen, 1974; Martin, 1974a,b,
1976; Briggs et al., 1979; Dollfus et al., 1984a,b).

The problem of Martian dust storms is posed by the question of what are the
mechanisms controlling the origin, maintainence, and decay of the dust storms.
This problem requires knowledge of the thermal structure and dynamical properties
of the Martian lower atmosphere (e.g., Zurek et al., 1992).

In this article, different aspects of Martian dust storms will be described. These
aspects include: classification of dust storms, properties of the airborne dust, sea-
sonal and interannual variability of dust storms, their evolution, their causes, sur-
face changes in atmospheric variables associated to Martian dust storms, models
for these storms, and seasonal and long-term aspects of the dust cycle.

For the following discussion it is convenient to remember that the longitude of
the Sun in Mars-centered (aerocentric) coordinates, Ls , is used as an angular mea-
sure of the Martian year with Ls 0◦, 90◦, 180◦, and 270◦ marking the beginning of
northern hemisphere spring, summer, fall, and winter (southern fall, winter, spring,
and summer), respectively. A Mars solar day or sol is equal to 24.66 hours and a
one Mars year is equal to 669.6 sols (or to 1.88 Earth years).

2. Classification of Dust Storms

Martin and Zurek (1993) compiled a comprehensive list of dust activity on Mars,
from 1873 to 1990. They developed a classification scheme which takes into ac-
count the following aspects: (1) the size of the dust “event” (planet-encircling,
regional, or local); (2) the nature of the identifying observations (e.g., photography,
polarimetry, or spacecraft data); and (3) the type of activity (obscurations, station-
ary cloud, or storm). They did not attempted to include local dust storms observed
by spacecraft, which were reviewed by Peterfreund (1981) and James (1985). Dust
storms are dust clouds that were observed to move or expand (Zurek and Martin,
1993). The classification by size include: (a) the planet-encircling storm (a runaway
storm that engulfs all longitudes, East-West, of Mars with dense clouds, although
not necessarily simultaneously); (b) regional storm (long axis over 2000 km, but
do not encircle the planet); and (c) local storm (long axis of affected area less
than 2000 km). The planet-encircling storms have also been referred as “global”
or “planetwide” storms. “Nonlocal” dust storms refers to both planet-encircling
and regional dust storms (long axis over 2000 km), the obscuration of areas being
greater than a few million square kilometers (Martin and Zurek, 1993; Zurek and
Martin, 1993). Martin and Zurek (1993) feel that the terms “great” or “major”
should be restricted to those events in either the encircling or the regional classes,
thus excluding the “locals”. Some images of dust storms are shown in Figures 1, 2,
and 3.



MARTIAN DUST STORMS: A REVIEW 21

Figure 1. The end of a Mars global dust storm on the day after perihelion, at 3 pm local time (Viking
Orbiter Image 248B57). The storm is near the edge of the shrinking south polar cap. This picture
was taken by Viking Orbiter 2: the storm is about 200 km across, the frame width is 710 km, and the
frame center is at 64 deg S. [Reproduced from “Weather, Climate and Life on Mars: Frequently asked
questions answered-what causes dust storms?” (http://humbabe.arc.nasa.gov/mgcm/faq/dust.html),
written and maintained by David Catling, Space Science Division, NASA Ames Research Center,
with inputs from other scientists who study Mars including Bob Haberle, Jeff Hollingsworth, Mike
Mellon, Jim Murphy, . . . so far.]

3. Properties of Airborne Dust

The intensity, scale and evolution of a dust storm are characterized mainly by the
opacity (τ ) of the dust. Zurek (1982) has described several attempts to estimate
the opacity of the atmosphere under dust storm conditions. As pointed out by him,
these attempts include studies which use: (a) Viking Lander images of the Sun and
Phobos, (b) Orbiter TV images of the surface or the limb of the planet, (c) Orbiter
infrared observations, and (d) less direct estimates using atmospheric temperature
or surface pressure. Based on the works of Pollack et al. (1979) and Thorpe (1981),
Zurek (1982) showed the visible (effective wavelength of 0.67 µm) optical depths,
τ , of the dust haze above Viking Lander 1 (VL1) and Viking Lander 2 (VL2). He
indicates the following: (1) the opacity increases sharply to large values (τ > 2) at
Lander 1 near Ls 205◦ and 279◦, (2) the opacities at the two Lander sites (VL1 at
22.5◦N, 48◦W; VL2 at 48◦N, 226◦W) have similar trends but it is generally smaller
at the more northern site, and (3) the opacity never seems to fall below a value of a
few tenths at any time during the year. In addition, infrared optical depths estimates
from the Viking Orbiter Infrared Thermal Mapper (Martin et al., 1979) and from



22 WALTER FERNÁNDEZ

Figure 2. Top: These Hubble Space Telescope-WFPC2 images of Mars, taken on May 17 (left)
and June 27 (right), 1997, reveal a significant dust storm which fills much of the Valles Marineris
canyon system and extends into Xanthe Terra, about 1000 km south of the Pathfinder landing site.
Visual comparison of these two images clearly shows the dust storm between 5 and 7 o’clock and
about 2/3 of the way from the center to the southern edge of the June image. Bottom: The digital
data were projected to form this map of the equatorial portion of the planet. The cross marks the
approximate location of the pathfinder landing site, and the ribbon of dust which runs horizontally
across the bottom of the map traces the location of Valles Marinaris. Most of the dust is confined
within the canyons, which are up 5–8 km deep. The thickness of the dust near the eastern end
of the storm is similar to that observed by Viking Lander 1 during the first two 1977 global dust
storms which it studied. [Reproduced from “Canyon Dust Storm and Pathfinder Landing Site”
(http://www.seds.org/hst/97-23.html), where other interesting features of these images are discussed.
Credit: Steve Lee (University of Colorado), Mike Wolff and Phil James (University of Toledo) and
NASA.]

the Lander surface pressure data (Zurek, 1981, 1982), indicate that the opacity of
the dust haze above VL1 is fairly representative of the dust haze over low latitudes
during the occurrence of great dust storms (large opacity). In the two major dust
storms of 1977, the opacities at VL1 site remained elevated above the values prior
to the storm for approximately 60 and 80 sols, respectively (Zurek and Martin,
1993).



MARTIAN DUST STORMS: A REVIEW 23

Figure 3. Two Hubble Space Telescope-WFPC2 images of Mars, taken about a month apart on
September 18 and October 15, 1996, reveal a dust storm churning the edge of the Martian receding
north polar cap. The images were assembled from separate exposures taken with the Wide Field
Camera 2 (WFC2). Top (September 18, 1996): The notch in the white north polar cap is a 1000 km
long storm. The bright dust can also be seen over the dark surface surrounding the cap, where it
is caught up in the Martian jet stream and blown easterly. The clouds at lower latitudes are mostly
associated with Martian volcanos such as Olympus Mons. Bottom (October 15, 1996): Though the
storm has dissipated by October, a distinctive dust comma-shaped feature can be seen curving across
the ice-cap. The shape is similar to cold fronts on Earth, which are associated with low pressure
systems. To help compare locations and sizes of features map projections (right of each disk) are
centered on the geographic north pole. Maps are oriented with 0 degrees longitude at the top and
show meridians every 45 degrees of longitude (longitude increases clockwise); latitude circles are
also shown for 40, 60, and 80 degrees north latitude. [Reproduced from “Springtime Dust Storm
Swirls at Martian North Pole” (http://www.seds.org/hst/96-34.html). Credit: Phil James (University
of Toledo), Steve Lee (University of Colorado) and NASA.]



24 WALTER FERNÁNDEZ

Pollack et al. (1979) found from the Viking Lander data that visible optical
depth had a value of about 1 just before the onset of each of the two major dust
storms which occurred during the period of observations; it increased very rapidly
on a time scale of a few days, to peak values of about 3 and 6 with the arrival of
the first and second storm, respectively.

Clancy and Lee (1991), from an analysis of emission-phase-function (EPF) ob-
servations from the Viking Orbiter IRTM solar channel (passband = 0.3–3.0 µm,
λeff = 0.67 µm), found that: (1) opacity ranges of 0.2–3.0 with a Ls dependence
similar to the Viking Lander results (Pollack et al., 1979); (2) dust opacities as high
as 0.4 at the summit of Olympus Mons, and as high as 0.8 near the south pole during
the global dust storm of 1977; (3) a dust scattering phase function identical to that
derived from the Viking Lander observations, but the single scattering asymmetry
parameter in both cases is 0.55–0.56 rather than 0.79; (4) a dust single scattering
albedo of 0.92, as compared to the Viking Lander result of 0.86; (5) dust particle
sizes may be 5–10 times smaller than suggested by Toon et al. (1977), based on the
Viking Lander visible scattering phase functions and the observed ratio of visible
and infrared dust opacities.

Recently, Ockert-Bell et al. (1997) developed a wavelength model of the single
scattering properties of the Martian dust using the particle size distribution, shape,
and single scattering properties at Viking Lander wavelengths given by Pollack
et al. (1995). They have listed the values (as function of wavelength) of the dust
single scattering albedo (ω0), single scattering asymmetry parameter (g), extinction
factor (Qext), real index of refraction (nr ) and complex index of refraction (ni) in
the 0.21–4.15 µm wavelength range. For the wavelength of 0.67 µm, the values
of ω0,g, Qext, nr , and ni are 0.93, 0.65, 3.04, 1.51, and 0.003, respectively. Their
values of ω0 in the 0.21-1.015 µm wavelength range, which vary from 0.72 to 0.95,
were modified by Wolf et al. (1997) in the blue/ultraviolet and at 1.015 µm to
provide consistent fits of the Hubble Space Telescope data to derive cloud opacity.

The thermal infrared (9 µm) opacity of dust has been derived from brightness
temperature measurements made by the Viking Orbiter Infrared Thermal Mapper
(IRTM) data (e.g., Martin et al., 1979; Hunt et al., 1980; Martin, 1986; Martin
and Richardson, 1993). The infrared spectral measurements are very sensitive to
the presence of dust, which has a strong absorption band at 9 µm (e.g., Hunt
et al., 1980). Since the extinction factors at the visible and infrared wavelengths
are affected by both the composition and the particle size distribution, the ratio
between the Lander (visible) and infrared (9 µm) opacities gives information about
the nature of the dust (Martin, 1986). For comparison with the Lander opacities,
the infrared opacities are scaled by some factor. A value of 2.5 for the visible-to-
infrared ratio of opacity fits the results obtained by Martin (1986) and Martin and
Richardson (1993). Zurek (1982) found values for such a ratio of 1.85 using the
derived infrared opacities of Martin et al. (1979), and of 1.2 for montmorillonite
and the size distribution of Pollack et al. (1979). A value of 1.69 for the visible-
to-infrared ratio of opacity was used by Hunt et al. (1980) to derive the visible



MARTIAN DUST STORMS: A REVIEW 25

opacity from the infrared opacity. Hunt et al. (1980) found values of 9 µm opacity
of about 2.5 in the most opaque regions of a local dust storm. Several kilometers
away from the center, the atmosphere was still very dusty, with infrared opacity up
to approximately 1.4. Correspondingly, the visible opacity is in the 2–5 range.

The use of Viking IRTM data for opacity derivation is important for mapping
purposes (Hunt et al., 1980; Martin, 1986; Martin and Richardson, 1993). The
coverage of the IRTM data is greater than that of the imaging experiment. A global
dust opacity mapping for Mars was performed by Martin and Richardson (1993)
from Viking IRTM data. The period considered is from the beginning of Viking
observations until Ls 210◦ in 1979 (1.36 Mars years). This allowed the authors to
study the evolution of the two major dust storms of 1977. In addition, their 9 µm
opacity maps have been used by Martin (1995) to make estimates of the mass
of dust suspended in the Martian atmosphere. He found that the maximum dust
content during the 1977b storm amounted to about 4.3× 1014 g, which corresponds
to about 4.3 × 10−4 g/cm2, or a layer 1.4 µm thick, if densely compacted. He also
found that during a local dust storm near Solis Planum at Ls 227◦, about 1.3 ×
1013 g of dust were lofted.

Some evidence has been found that the dust particles are nonspherical, which
has important optical effects at visible and shorter wavelengths (Thorpe, 1978;
Pollack et al., 1977, 1979). As pointed out by Clancy et al. (1991), the generally
accepted model of the dust particle size distribution is based on the analyses of
Mariner 9 Infrared Interferometer Spectrometer (IRIS) spectra of dust and on the
Viking Lander observations of the visible scattering phase function of atmospheric
dust (Toon et al., 1977). The distribution is a modified gamma distribution with a
mode radius of 0.4 µm and an average cross-section weighted particle radius (reff)
of 2.5 µm. Other particle size distributions, in which reff is smaller than 2.5 µm,
have been deduced or used by other researchers (e.g., Drossart et al., 1991; Clancy
et al., 1995; Pollack et al., 1995); see also Table I. For example, Pollack et al.
(1995) used a lognormal particle size distribution with reff = 1.85 µm, whereas
Smith et al. (1997) using Pathfinder’s data estimated reff to be 1.0 (+0.3,−0.2) µm.
Results by Clancy and Lee (1991) indicate that the Viking Lander observations are
at least consistent with reff = 0.4 µm. This is in agreement with the results of
Pollack et al. (1977). Several arguments given by Clancy and Lee (1991) provide
evidence that dust in the Martian atmosphere is composed of very fine particle sizes
(reff ≤ 0.4 µm).

With regard to composition of dust, magnetite has been identified as a particle
mineral (Pollack et al., 1979), as well as mixtures of montmorillonite and basalt
(Toon et al., 1977). Hanel et al. (1972) found that dust particles have a composi-
tion that corresponds approximately to that of rocks of intermediate SiO2 content.
Clancy et al. (1995) found that a much broader, smaller particle size distribution
coupled with a “palagonite-like” dust composition fits the complete ultraviolet-to-
30 µm absorption properties of the dust better than the distribution and mixture of
montmorillonite and basalt dust composition given by Toon et al. (1977).



26
W

A
LT

E
R

F
E

R
N

Á
N

D
E

Z

TABLE I

Compilation of martian dust particle properties inferred from spacecraft observations (from Murphy et al., 1993)

λ, Mean Other parameters Comments Source

µm radius,

µm

Ultraviolet

0.263, 2 r decrease with time; opacity South pole; the number of larger Mariner 9; Pang and Hord (1973)

0.300 decreases from 1 to 0.05 particles decreases with time

0.268, 1 Variance ∼0.4 Composition is not dominated by Mariner 9; Pang et al. (1976)

0.305 montmorillonite

0.268, 0.2 Nonspherical particle shape Nonspherical shape effects allow for Mariner 9; Chylek and Grams (1978)

0.305 montmorillonite

Visible

0.585 10 ω0 ∼ 0.7 −−0.85 Surface particles are of similar size Mariner 9; Leovy et al. (1972)

0.565, 2.5, Smaller particles at higher altitudes From clarity of surface features at Mariner 9; Hartmann and Price (1974)

0.610 5.0 various surface elevations

0.67 0.4 Nonspherical 2 northern hemisphere sites only Viking Lander; Pollack et al. (1977)

0.67 2.5 rm = 0.4 µm, ω0 = 0.86, γ=0.79; 2 northern hemisphere sites only Viking Lander; Pollack et al. (1979)

nonspherical shape
0.493, – ω0 ∼ 0.5; λ = 0.592 µm Global perspective Viking Orbiter; Thorpe (1978, 1981)

0.592 ω0 ∼ 0.8; λ = 0.443 µm, g < 0.6

0.592 – g ∼ 0.55–0.66, ω0 ∼ 0.9–0.98 Planet limb viewing Viking Orbiter; Jaquin et al. (1986)

0.3–3.0 0.4 ω0 ∼ 0.92, g ∼ 0.55 Little variation with space or time Viking Orbiter EPF; Clancy and

Lee (1991)



M
A

R
T

IA
N

D
U

ST
ST

O
R

M
S:A

R
E

V
IE

W
27

TABLE I

Continued

λ, Mean Other parameters Comments Source

µm radius,

µm

Infrared

5–50 1–10 – Based upon comparisons with Mariner 9 IRIS; Conrath et al. (1973)

synthetic spectra

0.7, 0.5–1 – Dust-induced “anti-greenhouse” Mars 2 and 3; Moroz and

1.38 effect Ksanfomaliti (1972)

15 1 – Based on particle sedimentation Mariner 9; Conrath (1975)

lifetimes

5–50 2.5 rm = 0.4µm, modified gamma Size distribution is constant at subsolar Mariner 9 IRIS; Toon et al. (1977)

distribution latitudes

7–45 – – Fewer large particles at south pole than Mariner 9 IRIS; Paige et al. (1987)

at lower latitudes

7, 9, - – Derived 9-µm opacities are a factor of Viking IRTM; Martin (1986)

15, 20 2.5 less than observed values

– – – Apparent “neutral” greenhouse effect Viking Lander; Haberle and

due to dust Jakosky (1991)

5–50 – rm ∼ 0.5 µm, variance ∼0.15 Post-1971–1972 dust storm; palagonite Mariner 9 IRIS, Santee and

is a representative mineral Crisp (1991)

λ represents the wavelength(s) employed in the various studies listed.



28 WALTER FERNÁNDEZ

Murphy et al. (1993) made a compilation of Martian dust particle properties
inferred from spacecraft observations (see Table I). Other useful information about
properties of airborne dust on Mars has been provided by Kahn et al. (1992).

Profiling of Mars atmospheric temperature has been obtained from Earth-based
microwave measurements of 1.3 mm and 2.6 mm rotational spectra of Mars at-
mospheric CO (Clancy et al., 1990, 1994, 1996). The disk-average (i.e., low- to
mid-latitude) atmospheric temperatures obtained with this technique as early as
1975 allow determination of dust storms, because if the atmosphere has a larger
amount of dust it is generally warmer and more isothermal than a relatively clear
atmosphere. Abrupt increases of Mars atmospheric temperature at the 0.5 hPa level
(greater than 20 K), obtained from the microwave CO spectra in 1992 (Ls 228◦)
and 1994 (Ls 254–283◦), compare favorably with the temperature increases at the
0.5 hPa level obtained by Viking IRTM at the beginning of the two 1977 global
dust storms (Clancy et al., 1996). The occurrence of two global dust storms, one
in 1992 (Clancy et al., 1996) and another in 1994 (Clancy et al., 1994, 1996), was
inferred from these observations. As pointed out by Clancy et al. (1996) neither of
the 1992 and 1994 microwave identifications of Martian global dust storms have
been verified by Earth-based imaging. Clancy et al. (1996) have indicated that this
lack of supporting Earth-based measurements is due to the difficulty of observing
Mars around conjunction, which coincides with the so called dust storm season
during the last two cycles. For this reason, the 1977 Mars global dust storms were
not observed from Earth-based imaging. Microwave observations of CO emission
lines from the Mars atmosphere will greatly augment our knowledge of interannual
variability of dust storms.

4. Seasonal Variability of Dust Storms

Martin and Zurek (1993) found that dust storms have originated in many different
regions of Mars. In general, for all size classifications, there is more activity in the
southern hemisphere and in specific regions like Hellas (light albedo).

All planet-encircling storms have been observed during the southern spring and
summer seasons, although there is significant interannual variability. These storms
began between Ls 204◦ and 310◦, during the south polar cap recession (Martin and
Zurek, 1993). The initial clouds for planet-encircling storms have occurred over
dark areas. In most cases this initial area has been Hellespontus (dark albedo).

Regional dust storms occur in nearly all seasons, with the two principal ex-
ceptions being Ls 130◦–160◦ and 330◦–20◦, and they occur more frequently than
planet-encircling storms (Martin and Zurek, 1993). They do tend to occur most
frequently (Ls 160◦–330◦) during the southern spring and summer seasons, in the
so called “dust storm season”.

The period Ls 161◦–326◦ has been referred to as “dust storm season” (Martin
and Zurek, 1993). This period is nearly centered on perihelion (Ls 250◦), which



MARTIAN DUST STORMS: A REVIEW 29

is the time of maximum insolation on Mars. This enhance atmospheric convection
during daytime, as well as thermal circulations induced by topography.

So, both seasonal ranges (for planet-encircling and regional dust storms) are
somewhat centered on perihelion.

5. Interannual Variability of Dust Storms

Martin and Zurek (1993) found that there is significant interannual variability in the
occurrence of planet-encircling dust storms. In addition, Zurek and Martin (1993)
found that “the chance of a planet-encircling dust storm occurring in any arbitrary
Mars year is estimated to be approximately one in three (18–55% at the 95% level
of confidence), if such occurrence is random from year-to-year and yet restricted
seasonally to southern spring and summer”. So, as they do not occur every Martian
year, it appears that seasonal conditions in themselves do not provide the require-
ments for smaller storms to evolve into planet-encircling storms. Thus, as pointed
out by Martin and Zurek (1993), “we must go beyond the question of how to raise
dust, to how do a few dust storms perpetuate themselves but other do not?”.

The probability for nonlocal dust storms to occur during a given storm season is
approximately 80%, reflecting the large number of detections of regional, but not
planet-encircling dust storms (Zurek and Martin, 1993).

Ingersoll and Lyons (1993) proposed the hypothesis that the global Martian
climatic system, consisting of atmospheric dust interacting with the atmospheric
circulation, produces its own interannual variability when forced at the annual
frequency. This is supported by their findings in which the interannual variability
manifests itself either as a periodic solution in which the period is a multiple of the
Martian year, or as an aperiodic (chaotic) solution that never repeats.

6. Evolution of the Dust Storms

There are only five well-documented cases of planet-encircling dust storms, which
occurred in 1956, 1971, 1973, and 1977 (two). However, as pointed out before,
Earth-based microwave measurements of Mars atmosphere have provided evidence
of the occurrence of global dust storms in 1992 and 1994 (Clancy et al., 1994,
1996). Ebisawa and Dollfus (1986) found that in 1956, 1971, and 1982, the planet-
encircling dust storms manifested sharp-cut boundaries (did not display irregular
or diffuse outlines) at their outbreak in the early stage of their evolution. They
have also pointed out that the dust is first raised in specific and precisely localized
regions and at this stage it does not spread out quickly. Actually, these local dust
clouds (bright spots in Earth-based photography) expand slowly during this initial
phase, which last usually less than a week. In the next several days, new dust
clouds develop and old ones merge, the dust cloud expands and moves toward other



30 WALTER FERNÁNDEZ

regions; its outlines are irregular and diffuse. Secondary local dust clouds may
develop at new locations. The planet-encircling dust storms expand first zonally
(east-west direction) until a latitudinal band is obscured, which is followed by a
expansion in the north-south direction (Martin, 1974a,b, 1976). Their minimum
lifetime is 3 or 4 weeks, while regional storms probably do not last more than 2
weeks (Martin, 1984). It was found that the planet-encircling dust storm of 1971
went through a daily cycle of regeneration, although each day it advanced farther
than it did the day before (Martin, 1974a).

There are only two known regions in which planet-encircling storms originate:
Helles-Hellespontis and Solis Planum-Claritas Fossae (Martin, 1984). The areas
subject to local and regional dust activity are much more numerous and have been
mapped by Martin and Zurek (1993).

7. Causes for Dust Storms

7.1. THE SALTATION MECHANISM

In order for a dust storm to be formed, dust particles must be rised into the at-
mosphere, possibly by the saltation mechanism (Bagnold, 1941). The surface char-
acteristics and the boundary layer flow are also involved in the process of rising
dust into the atmosphere. The extension of Bagnold’s work to Martian conditions
has been carried out by several authors (e.g., Gifford, 1964; Ryan, 1964; Sagan
and Pollack, 1967; Hess, 1973) and has been summarized by Gierasch and Goody
(1973). Assuming that the drag coefficient (Cd ) of a sand grain is a function only of
the particle Reynolds number, the minimum friction velocity to raise sand (U ∗min)
is given by:

(U ∗min)
3 = constant(σgν/ρ), (1)

where σ is the grain density,g is gravity acceleration, ν kinetic viscosity, ρ air
density, and the constant is 1.2 × 102 from Bagnold’s data. For CO2 at 200 K and
6 hPa, Gierasch and Goody (1973) found that U ∗min = 180 cm s−1, as compared
to 15 cm s−1 for terrestrial conditions. According to Csanady (1967), the friction
velocity (U ∗) relates to the approximately geostrophic free-steam velocity (Ug) as
follows:

(U ∗/Ug)2 = Cd, (2)

where Cd is the drag coefficient. Leovy and Mintz (1969) used the following
relationships:

U ∗ = 3× 10−2Ug, for stable conditions (3a)

U ∗ = 6× 10−2Ug, for unstable conditions. (3b)



MARTIAN DUST STORMS: A REVIEW 31

For unstable conditions on Mars, with U ∗ = 180 cm s−1, Ug = 30 m s−1. Iversen
et al. (1976a,b) obtained a threshold friction velocity for Mars of about 2 m s−1.

The particles which are most likely to be moved at the ground have an approx-
imated radius of 100 µm (Greeley et al., 1980). Saltation of particles of this size
will cause the raise of smaller particles from the ground, some of which will return
to the surface because of setting. Wind speeds larger than 25 m s−1 at 1.6 m above
the ground were recorded by VL1 during a local dust storm (Ryan et al., 1981;
James and Evans, 1981). Moore (1985) pointed out that, during the Martian dust
storm of sol 1742 at VL1, winds probably attained speeds of about 40–50 m s−1 at
a height of 1.6 m above the surface and friction speeds were probably about 2.2–
4.0 m s−1; a spectrum of particle sizes from ten or so micrometers and clods, up
to 4–5 mm, were salted by and entrained in the wind. A comprehensive discussion
of the mechanisms by which dust may be transported from the surface into the
atmosphere (which is beyond the purpose of this article) is given by Greeley et al.
(1992).

Properties of surface particulate materials have been determined from thermal
inertia data obtained from the Mars Gobal Surveyor thermal emission spectrometer,
the Viking IRTM, and the Mariner 9 infrared radiometer (e.g., Kieffer et al., 1973;
Edgett and Christensen, 1991; Presley and Christensen, 1997b). For example, Ed-
gett and Christensen (1991) found that thermal inertia data indicate that Martian
aeolian dunes have an average particle size of about 500 ± 100 µm. They also
found that the particle size transition from suspension to saltation is about 210 µm.
These results are consistent with particle sizes of 400–460 µm for 270 IU as de-
rived by Presley and Christensen (1997b). A review of the methodology available
for the study of thermal conductivity of particulate materials – with emphasis on
low atmospheric pressures – is given by Presley and Christensen (1997a). Exper-
iments carried out by Presley and Christensen (1997b) have shown that thermal
conductivity is approximately related to particle diameter and atmospheric pressure
(in the 1–100 torr range) by the relation:

κ = (CP 0.6)d−0.11 log(P/K), (4)

where κ is the thermal conductivity in W/(m K), P is atmospheric pressure in torr,
d is the particle diameter in µm, C ≈ 0.0015, and K ≈ 8.1 × 104 torr (when
these units are used). This relation is valid for particulate materials that have a
medium well-packed bulk density (Presley and Christensen, 1997b). The particle
size dependence of the thermal conductivity derived by Presley and Christensen
(1997b) is significantly different from the estimate by Kieffer et al. (1973), except
for 200 µm particles. Presley and Christensen (1997c) found that particle shape
influences the thermal conductivity primarily by affecting the bulk density and that
significant differences in bulk density influence the thermal conductivity.



32 WALTER FERNÁNDEZ

7.2. POLAR CAP RECESSIONS AND MAJOR DUST STORMS

Most of the dust events (78%) studied by Martin and Zurek (1993) occurred be-
tween Ls 161◦ and 326◦, during the recession of the south polar cap. However,
no events were observed during the two periods when the contracting caps are
close to their minimum size, and sublimation has become minimal. Therefore, it
appears that there is some relationship between polar cap recessions and major
dust events and that possibility might be strongest for the south polar cap. This
is supported by the Viking images, which show local dust near the edges of the
receeding caps (Briggs et al., 1979). The relationship between polar cap recessions
and dust storms is the result of strong local and large-scale winds due to large
temperature gradients at the polar cap edge and mass outflow of the subliming
cap (Leovy et al., 1973; Burk, 1976; French and Gierasch, 1979; Haberle, 1979;
Haberle et al., 1979; Fernández, 1995a,b). Besides this situation, the regions where
major dust storm develop show favorable topographic characteristics: slopes and
uplands which can enhance atmospheric convection and flow lifting by orography,
as well as particular surficial properties (e.g., albedo and thermal inertia). A study
of the Martian boundary layer, taking into account topography, was made by Blum-
sack et al. (1973). They found, 1 km above a Martian slope of 0.005, variations in
the temperature of ± 15 K and in the upslope wind of ± 25 m s−1. Magalhaes and
Gierasch (1982) have also studied wind induced by slopes on Mars. They found
that for a surface with low thermal inertia and large roughness length wind speeds
in the early morning hours can produce saltation. Some preliminary studies about
the effects of topography on baroclinic instability were made by Blumsack and
Gierasch (1972).

7.3. THE POLAR CAP BREEZE

The line of separation between the Martian polar caps and the “desert” (frost-free)
adjacent areas constitutes a discontinuity in meteorological boundary conditions.
The polar caps are composed of CO2 while the “desert” areas are formed mainly by
dust particles with a composition that corresponds approximately to that of rocks
of intermediate SiO2 content (Hanel et al., 1972). The difference in the response
capabilities of the polar cap and the “desert” areas to the radiation field originates a
meridional temperature gradient, which produces in the atmosphere a baroclinic in-
stability. This generates an atmospheric circulation system similar in some aspects
to the terrestrial sea breeze. Fernández (1995a,b) called this circulation system the
Martian polar cap breeze.

From calculations made by Pallmann and Frisella (1972, Fig. 5) and by Pall-
mann et al. (1973, Fig. 13), it can be seen that in summer time the smallest tem-
perature difference between the polar cap and the “desert” adjacent areas occurs at
about 04:00 local time and has a value of about 74 K. The greatest temperature
difference occurs near 15:00 local time and is then about 102 K. The surface
temperature of the polar cap is the frost point temperature of CO2 (148 K) and



MARTIAN DUST STORMS: A REVIEW 33

it is an equilibrium temperature of the solid and gaseous phases. Since the “desert”
areas are much warmer than the polar cap, and being that density decreases upward,
the surfaces of constant density slope upward from the “desert” areas to the polar
cap. Because the isobaric surfaces are usually more nearly horizontal, the two sets
of lines intersect to form a solenoid field in the vertical plane (baroclinic effect). If
we assume for a moment that Mars is a longitudinally uniform nonrotating planet
and if we ignore friction, then the solenoid field originates a circulation system
consisting of a desertward current adjacent to the surface, and a much weaker return
above. The baroclinic effect would cause progressively increasing wind speeds. In
addition to the Coriolis and friction effects, other factors influence the polar cap
breeze (Fernández, 1995a). They are: the prevailing wind, topography, irregularity
of the polar cap-edge, stability of the atmosphere, and seasonal variations.

Burk (1976) and Fernández (1995b) have made numerical simulations of the po-
lar cap breeze, utilizing models constructed in part from work previously done by
others on the terrestrial sea breeze. For a situation in which the cap edge was taken
to be at 65◦S, Burk (1976) found maximum horizontal winds of approximately
20 m s−1 when the early morning temperature profile was isothermal. Fernández
(1995b) found maximum wind speeds of about 15 m s−1 when the polar cap is
near 50◦S and based on the data given by the Mariner 6 exit, which presented an
inversion of γ = − 3.3 K/km near the ground. Although these winds are not strong
enough to rise dust, they may act together with other wind regimes.

It should be pointed out that these simulations did not considered several effects
that influence the polar cap breeze and which may produce stronger winds, such as
thermal and mechanical circulations induced by topography.

7.4. THE LARGE-SCALE POLAR CAP WIND

The polar cap breeze is a local wind phenomenon affecting only the immediate
vicinity of the polar cap-edge. There is also a large-scale polar cap-and-“desert”
wind thermally induced by the general temperature differences between the two ar-
eas. Leovy and Mintz (1969) made calculations of meridional and zonal winds for
two experiments; one simulates orbital conditions at the southern summer (northern
winter) solstice of Mars, and the other, orbital conditions at the southern autumnal
equinox. At the solstice, the mean meridional circulation shows intense thermally
directed circulation: the air rising in the subtropics of the summer hemisphere,
moving northward at the upper level, and descending in the winter subtropics.
There is some indication of a weaker reverse cell at high latitudes (approximately
30◦). However, the winds are extremely light. In the southern autumnal equinox
experiment, the meridional winds are much weaker.

Haberle et al. (1993) have made numerical simulations of Mars atmospheric
general circulation with the NASA Ames General Circulation Model. At the south-
ern summer (northern winter) solstice and when the atmosphere is relatively clear
they found that the mean meridional circulation consists of three cells: a large and



34 WALTER FERNÁNDEZ

intense cross-equatorial Hadley cell with rising and descending branches centered
at 30◦S and 30◦N, respectively; a thermally indirect Ferrel cell at high northern
latitudes, and a weak Hadley cell – with lower level horizontal winds from po-
lar cap to “desert” areas – at high southern latitudes. They also found that over
75% of the mass supplied to the rising branch of the cross-equatorial Hadley cell
flows through the lowest several kilometers of the atmosphere and that the mean
meridional motions are about a factor of 6 stronger on Mars than on Earth. At the
southern winter solstice, the cross-equatorial Hadley circulation (with its rising
branch at northern mid-latitudes) is considerably weaker compared to northern
winter. A Hadley cell at high southern latitudes is also present at the southern
winter solstice, but the horizontal flow at lower levels is toward the pole. In the
equinox experiments, the cross-equatorial Hadley cell almost disappears and there
are two Hadley cells of comparable strength, one over the Southern Hemisphere
and the other over the Northern Hemisphere. In the Southern Hemisphere, large-
scale wind at lower levels from polar cap to “desert” areas is observed in the spring
and fall equinox experiments (Haberle et al., 1993).

A numerical study of the large-scale polar cap winds on Mars was carried out by
Haberle et al. (1979). Their approach consisted of utilizing the complete nonlinear
equations in sigma-coordinates to determine the seasonal variation of the middle
and high latitude circulation system. The spacing between grid points was 2.5
(148 km) and lateral boundaries were set at 27.5◦N and 90◦N. They considered
the following factors: seasonal thermal forcing, mass exchange between polar caps
and atmosphere, large-scale topography, and polar cap size. During spring, the
thermal forcing sets up a circulation with easterlies equatorward of the cap-edge
and easterlies poleward of it; the maximum westerlies occur where the horizontal
temperature gradients are largest. Vertical motions are strongest in the vicinity of
the polar cap-edge with rising motion over the “desert” areas and sinking motion
over the polar cap. The intensity of the circulation is proportional to the cooling
power or size of the polar cap. This pattern changes for other times of the year.
In winter, surface winds are westerly over the entire cap and part of the adjacent
“desert” areas. In late winter and early spring, a two-cell circulation is observed
with an inflow at the higher latitudes of the polar cap and an outflow at the lower
latitudes. Their results also show that surface winds near the edge of a retreating
polar cap are significantly enhanced. The maximum surface winds are 20 m s−1

during spring and 30 m s−1 during winter for the Northern Hemisphere.

7.5. ATMOSPHERIC TIDES AND DUST STORMS

On the planetary scale, as pointed out by Zurek (1982), diurnal variation of surface
heating during relatively clear periods and of solar heating of the airborne dust
during dusty periods drives large-amplitude atmospheric tides, with the tidal winds
having their near surface maximum speeds 20◦ of latitude away from equator.
This may be the explanation for the tendency of local dust storms to form in the



MARTIAN DUST STORMS: A REVIEW 35

subtropics of both hemispheres. Leovy and Zurek (1979) found, from observations
of surface pressure oscillations at VL1 and VL2 sites that the thermally driven
global atmospheric tides were closely coupled to the dust content of the Martian
atmosphere.

7.6. MID-LATITUDE SYNOPTIC WEATHER SYSTEMS

Cold fronts associated with mid-latitude baroclinic waves have been proposed as a
mechanism for the generation of local dust storms, observed particularly in north-
ern hemisphere mid-latitudes during fall and winter (Leovy et al., 1972; Tillman
et al., 1979; Ryan et al., 1981; James, 1985). The wind speeds may be increased
by other factors (James and Evans, 1981). This mechanism is responsible for the
generation of a large number of terrestrial dust storms in mid-latitudes in winter
(James, 1985). Baroclinic waves could be an important factor not only in the gen-
eration of local dust storms but also in the growth and development of larger dust
storms (Ryan and Sharman, 1981). Fronts were frequently observed at the VL2 site
during fall and winter (Tillman et al., 1979). The passage of eastward travelling
baroclinic waves with centers north of the VL2 site has been analysed from surface
observations by Ryan et al. (1978) and Ryan and Sharman (1981). These surface
observations are described in the next section. The joint action of a baroclinic wave
and the diurnal thermal tide (when combined in phase) may produce strong winds
(Ryan and Sharman, 1981; James, 1985). Transient baroclinic eddy activity has
been studied from General Circulation Model (GCM) simulations by Barnes et al.
(1993). They found a strong transient baroclinic eddy activity in the extratropics of
the northern hemisphere during the northern autumn, winter, and spring seasons.
The simulated eastward propagating eddies are characterized by zonal wavenum-
bers of 1–4 and periods of 2–10 days. In addition, Barnes et al. found a very weak
transient baroclinic eddy activity in their southern winter simulations. According
to them, it appears that this is partly a consequence of the stabilizing effects of
the zonally symmetric topography in the GCM, but it also must be associated with
certain aspects of the zonal-mean circulation in the southern winter.

7.7. DUST LOAD PRIOR TO THE ONSET OF DUST STORMS

Observations suggest that the Martian atmosphere had a great amount of dust load
just prior to at least some of the planet-encircling dust storms (Pollack et al., 1979;
Martin et al., 1991a). This indicates that a dusty atmosphere may provide favorable
conditions for the development of these storms (Leovy et al., 1973; Leovy and
Zurek, 1979; Harberle et al., 1982; Martin et al., 1991b).



36 WALTER FERNÁNDEZ

8. Surfaces Changes in Atmospheric Variables

Ryan and Henry (1979) and Ryan and Sharman (1981) studied surface changes
in meteorological variables at VL1 (22.5◦N, 48◦W) and VL2 (48◦N, 230◦W) as-
sociated with the arrival of major dust storms. The arrivals of the storms were
accompanied by significant increase in wind speed and pressure at VL1; no such
changes were observed at VL2. A decrease in temperature range (due to a decrease
in maximum temperature and an increase in minimum temperature) was observed
at both sites. According to Ryan and Henry (1979) the wind directions during the
period of increase on sol 209 followed the same diurnal pattern as on previous
sols, switching from westerlies to easterlies in early afternoon, implying an inten-
sification of the tidal pattern rather than a response to a synoptic pressure system.
Ryan and Sharman (1981) have pointed out that the increase in diurnal pressure
amplitude is caused by increases in the tides as well as baroclinic waves. The
semidiurnal tide in particular is the basis indicator of “global” dust content. Ryan
and Sharman (1981) have also pointed out that a pressure periodicity of 2–3 sols
is evident and that the wind vector exhibits clockwise rotation in synchronization
with the pressure waves. This is due apparently to the passage of eastward traveling
baroclinic waves with centers north of the VL2 site (Ryan et al., 1978). From gen-
eral circulation model (GCM) runs, Pollack et al. (1993) found that atmospheric
dynamics have a significant impact on the seasonal pressure variations measured
at VL1 and VL2 sites.

9. Models for Dust Storms

As pointed out by Zurek (1982), any model attempting to explain the origin of the
great dust storms must address at least three critical questions: (1) why do these
precursor storms originate only in certain areas during certain seasons, (2) by what
mechanism do some, but certainly not all, local storms evolve to planetary scale,
and (3) how widespread is the source of the dust that eventually winds up in the
planetary-scale dust haze? A fourth question (also pointed out by Zurek, 1982) is:
where is the dust raised by a great dust storm deposited?

Several mechanisms have been proposed for the generation of major dust storms.
They may be divided in two classes (Gierasch, 1974; Zurek, 1982) based on the
enhancement of: (1) the planetary-scale components of the atmospheric circulation
(e.g., Leovy et al., 1973; Houben, 1982), and (2) small but expanding vortical
motion through a strong dust feedback (e.g., Gierasch and Goody, 1973; Golitsyn,
1973; Hess, 1973).



MARTIAN DUST STORMS: A REVIEW 37

9.1. MECHANISMS RELATED WITH LARGE-SCALE ATMOSPHERIC

CIRCULATION

These mechanisms may be summarized as follows (Leovy et al., 1973). During
southern spring local dust storms develop in the baroclinic zone adjacent to the
contracting south polar cap. This zone is characterized by strong local and large-
scale winds produced by large temperature gradients and sublimation of the polar
cap. These winds, enhanced by topographic characteristics (sloping terrain and up-
lands), rise the dust forming local dust storms in preferred areas such as Hellas and
Argyre. These local storms generate and dissipate during a period of several days.
The diabatic heating due to absorption of solar radiation by the dust intensifies
the general circulation, mainly over sloping terrain in the 20–40◦ latitude band,
which in turn maintain the formation of local dust storms that eventually merge and
develop into a major dust storm. The upper-level northward flow is also intensified
and the dust moves toward equator (to lower latitudes). As the dust rises in the
ascending branch of the cross-equatorial solsticial circulation it spread faster due
to vertical wind shear (wind increasing with height).

The northward expansion of the storms depends of how high the dust is raised.
Harbele et al. (1982) found, using a numerical model, that dust must raise up to
20 km in order to significant northward transport may occur. Other studies (Leovy
et al., 1972; Conrath et al., 1973) show that during planet-encircling dust storms
the dust was present in the 40–60 km altitude range.

The decay phase of the dust storms is the result of increasing static stability over
the regions where dust is rised (Leovy and Zurek, 1979; Pollack et al., 1979) and
the suppression of surface heating when opacities are greater than one (Pollack et
al., 1979). The decay phase may last approximately between 50 and 75 sols (e.g.,
Zurek, 1982).

Leovy et al. (1973) used a numerical model to simulate the characteristics of
three components of the global wind system during the planet-encircling dust storm
of 1971. These three components of the wind system are the steady axially sym-
metric component, and the diurnal and semi-diurnal tides. They found that zonally
averaged surface wind speeds in the equatorial zone were near 30 m s−1, which are
strong enough to rise dust by saltation (Golitsyn, 1973; Gierasch and Goody, 1973).
The pattern of these winds corresponds well with the pattern of observed light
surface streaks (light albedo markings trailing from craters and other topographic
features). These streaks extend from northeast to southwest in the northern tropics
and from northwest to southeast in the southern tropics (Sagan et al., 1973). Leovy
et al. (1973) also found that the surface winds due to the diurnal and semidiurnal
tides were not strong enough to rise dust, although they contribute significantly to
the total wind. The factors responsible for the intensification of meridional circu-
lation are strong insolation, low static stability, and high atmospheric absorptivity
and emissivity (Leovy et al., 1973).



38 WALTER FERNÁNDEZ

The effects of dust on the thermal structure and the dynamics of the Martian
atmosphere have been studied by several investigators (e.g., Gierasch and Goody,
1972; Haberle et al., 1982; Schneider, 1983; Pollack et al., 1979, 1990; Murphy et
al., 1990, 1993; Santee and Crisp, 1993). Some results are summarized below.

Haberle et al. (1982) used a primitive equation model to study the transport of
dust away from a southern surface source at Ls 270◦, as well as the changes due to
the absorption of solar and infrared radiation by dust and CO2. Their results show
that dust is very effectively transported by the zonal mean circulation which rapidly
increase in intensity and spatial extent as the dust spreads.

Schneider (1983) used a steady, zonally symmetric, nearly inviscid model to
study the development of planet-encircling dust storms. His results indicate that
as the heating is increased from zero, some critical value is reached at which
the response jumps from local to global, which may be related to the observed
“explosive” growth of planet-encircling dust storms.

Murphy et al. (1990) made numerical simulations of the decay of Martian planet-
encircling storms, using an aerosol transport/microphysical model. They found that
temporal and spatial variations of the size distributions of the particles suspended
in the atmosphere are important factors of the decay of dust storms.

Murphy et al. (1993) studied the size-dependent transport of dust particles in the
Martian atmosphere away from a specific surface source, using a coupled system of
a zonally symmetric primitive equation grid point model of the Martian atmosphere
and an aerosol transport/microphysical model. Their findings include: (1) When
the dust is distributed among a range of particle sizes from submicrons to tens
of microns, there is a more extensive latitudinal transport away from a specified
southern subtropical source than in the case when a single spherical particle size
(2.5 µm) is used; (2) the input spherical particle size distribution inferred by Toon
et al. (1977) from Mariner 9 IRIS spectra, is not maintained in suspension for
an extended period of time at subsolar latitudes for radii between 1 and 10 µm;
(3) when slower falling nonspherical (disk-shaped) particles are considered, an
improved maintenance in suspension of the input particle size distribution is ob-
tained, in better agreement with inferred values; and (4) the maximum calculate
visible dust opacities at the latitudes of the Viking landers generally fall short of
the values inferred during the 1977 global dust storms.

Planet-encircling dust storms do not occur during northern spring and summer
seasons, even when there is an intense baroclinic zone adjacent to the north polar
cap which produces local dust storms (Wolf et al., 1996). Figure 3 shows a local
dust storm churning the edge of the Martian receding north polar cap. As pointed
out by Zurek (1982), because of the ellipticity of the orbit of Mars and the present
occurrence of aphelion in late northern spring, the insolation available at Ls near
20◦ is only 74% of that available at Ls near 200◦ during the present epoch. This
reduces the intensity of the planetary-scale circulation and local convective activity.



MARTIAN DUST STORMS: A REVIEW 39

9.2. MECHANISMS RELATED WITH EXPANSION OF VORTICAL MOTION

9.2.1. Large Vortices
Gierasch and Goody (1973) modified a terrestrial hurricane model, developed by
Carrier (1971), to predict the evolution of a planet-encircling dust storm on Mars.
They did so, considering that the planet-encircling storm of 1971 suggest a self-
sustaining system converting solar to kinetic energy on a large organized scale. In
the dust storm case the increase in enthalpy comes from absorption of solar radia-
tion by the dust core, instead of by latent heat as in the case of terrestrial hurricanes.
In this model, as adapted to Martian conditions, an axisymmetric swirling flow field
is assumed to exist outside the central core and above the surface boundary layer
(this might occur above an elevated plateau). The radial pressure gradient drives a
flow radially inward toward the core where it rises. Then heating of the dusty core
by absorption of solar radiation intensify the horizontal gradients of temperature at
the core’s periphery, accelerating the cyclonic swirling flow inward into the core.
Above the core, there is a region where the flow is forced upward and outward
from the core. Gierasch and Goody (1973) found that: (1) In the growth phase the
area of the compact core of large dust particles grows by a factor of 10 in about
3 days, starting somewhere between 105 and 107 km2. (2) In the mature phase the
compact core exist for at least 10 more days, until day 13. At the same time a cloud
of comparatively small dust particles spreads sideways covering the whole planet
sometime after day 23. Particles 1µm in diameter are convected up to 40 km. (3) In
the decay phase gradual clearing takes place after days 13–23 with 1 µm particles
finally clearing about day 140. They pointed out that these findings agree quite well
with the observed behavior of the planet-encircling dust storm of 1971.

The local storms observed by Viking (e.g., Briggs et al., 1979) do not show
vortical motions. The possible explanation is that the winds are relatively strong in
the regions where the dust storms originate to allow the organization of the initial
cyclonic swirling flow (Zurek, 1982).

9.2.2. Dust Devils
Dust devils have been proposed as a mechanism to rise dust into the Martian
atmosphere (e.g., Ryan, 1964; Neubauer, 1966; Sagan and Pollack, 1969; Sagan
et al., 1971). Ryan and Lucich (1983) made a study of local convective vortices
(which includes dust devils) as observed by VL1 and VL2. By convective vortices
the authors meant those that transport heat upward, being formed in the presence of
superadiabatic temperature lapse rate and atmospheric vorticity. Ryan and Lucich
found that: (1) convective vortices occurred more frequently during Martian spring
and summer, (2) they occurred during the period from several hours after sunrise
to few hours before sunset, (3) they were more common at the more southerly VL1
site where convection is expected to be more active, (4) in the majority of cases
(about 70%) an increase in temperature was clearly associated with the vortex, (5)
no preference in the formation of cyclone or anticyclone vortices was detected, and



40 WALTER FERNÁNDEZ

(6) seven of the 118 vortices recorded had wind speeds that may have risen dust
from the surface.

Dust devils have been detected from Viking Orbiter images by Thomas and
Gierasch (1985). The overall shapes of the observed clouds (identified by those
authors as dust devils) were columnar, cone, funnel, or irregular, and they were
generally four to ten times as high as their were wide. The heights of most of
the clouds (61 of 99 observations) were in the 1.0–2.5 km range; the maximum
height was 6.8 km. The bases of the clouds were apparently narrow, and the widest
areas near the tops of the clouds were about 1 km across. In addition, Thomas and
Gierash have pointed out that all clouds identified as dust devils were imaged at
14 to 15 hours local time in local summer. Dust devils have also been detected
analysing Pathfinder’s meteorological data (Schofield et al., 1997).

The coexistence of a number of dust devils has been postulated as a mechanism
which may produce local dust storms (Hess, 1973). Once a dust cloud at least
10 km thick and several tens of km in radius is formed by the dust devils, the
absorption of solar radiation generate temperature gradients which, in turn, produce
winds capable of rising more dust (Hess, 1973; Golitsyn, 1973).

10. Seasonal and Long-Term Aspects of the Dust Cycle

Although it is not the purpose of this article to discuss all the different aspects
of the dust cycle on Mars, some seasonal and long-term aspects of the dust cycle
should be briefly mentioned. A detailed discussion of this subject is given by Kahn
et al. (1992).

Dust can be raised over large areas of Mars and redistributed over a large part of
the planet during the occurrence of planet-encircling dust storms. In the years when
planet-encircling dust storms are formed during southern hemisphere’s summer,
the dust raised in the southern hemisphere subtropics is transported to the northern
hemisphere subtropics by large-scale cross-equatorial upper level winds associated
with a Hadley type circulation (Kahn et al., 1992). Baroclinic waves in the northern
hemisphere may transport dust toward the northern polar region (Tillman et al.,
1979; Barnes et al., 1993). The incorporation of dust into the residual north polar
cap was proposed by Pollack et al. (1979). This suggests that the northern polar
region may act as a sink of dust in the present epoch. Other possible long-term
sinks of dust during the current epoch, associated with the descending branch of the
Hadley circulation, are the bright, low-thermal-inertia regions of Arabia, Tharsis,
and Elysium in the northern hemisphere (Kahn et al., 1992). It is unlikely that
the south polar cap may be a sink of dust in the present epoch, since in southern
hemisphere’s winter condensation is taking place at the south polar cap and tran-
sient baroclinic eddy activity is much weaker than in northern hemisphere’s winter
(Barnes et al., 1993). On the other hand, the more frequent formation of dust storms
in some specific areas (southern hemisphere subtropics and near the retreating polar



MARTIAN DUST STORMS: A REVIEW 41

cap-edges) indicate that these regions may be net sources of airborne dust (Kahn et
al., 1992).

Transport of dust may also take place in periods when planet-encircling storms
are not observed, as suggested by changes in surface albedo. The dust deposited in
some regions may be transported by winds close to the surface to neighboring re-
gions, which can explain changes in surface albedo. In general, changes in surface
albedo can be explained by seasonal and interannual redistribution of very small
amounts of dust (Kahn et al., 1992).

As pointed out by Kahn et al. (1992), if some regions are long-term sources
or sinks, losing or gaining the equivalent of a few µm a year, meters of material
can be removed or accumulated on time scales of 1 Myr. The observed polar- and
low-latitude layered terrains indicate a net deposition of airborne dust. Since the
obliquity cycle of Mars is about 51,000 years, the season of perihelion (when max-
imum insolation occurs) will reverse about every 25,000 years. This implies that
in the next reverse of the obliquity cycle the present sink regions (at the northern
low-latitudes and the northern polar region) will become source regions and the
southern polar region will act as a sink for airborne dust. However, arguments
given by Kahn et al. (1992) in relation to long-term dust deposition rates suggest
that either the dust cycle has operated differently in the past or/and that the current
understanding of the dust cycle is incomplete.

11. Concluding Remarks

The above description shows that important results have been obtained concerning
the preponderant forcing mechanisms controlling the origen, maintainence, and
decay of the dust storms. Nevertheless, further research is needed to provide addi-
tional information related to these systems. New missions to Mars would provide
valuable observational data to evaluate the numerical model results.

Acknowledgments

The helpful suggestions and comments of an anonymous referee improved the
manuscript and are greatly appreciated.

References

Antoniadi, E. M.: 1930, La Planète Mars 1659–1929, Hermann et Cie, Paris, pp. 39–47.
Bagnold, R. A.: 1941, The Physics of Blown Sand and Desert Dunes, Methuen, London, 265 pp.
Barnes, J. R., Pollack, J. B., Haberle, R. M., Leovy, C. B., Zurek, R. W., Lee, H., and Schaeffer,

J.: 1993, ‘Mars Atmospheric Dynamics as Simulated by the NASA Ames General Circulation
Model, 2. Transient Baroclinic Eddies’, J. Geophys Res. 98, 3125–3148.



42 WALTER FERNÁNDEZ

Blumsack, S. L. and Gierasch, P. J.: 1972, ‘Mars: the Effects of Topography on Baroclinic
Instability’, J. Atmos. Sci. 29, 1081–1089.

Blumsack, S. L., Gierasch, P. J., and Wessel, W. R.: 1973, ‘An Analytical and Numerical Study of
the Martian Planetary Boundary Layer Over Slopes’, J. Atmos. Sci. 30, 66–82.

Briggs, G. A., Baum, W. A., and Barnes, J.: 1979, ‘Viking Orbiter Imaging Observations of Dust in
the Martian Atmosphere’, J. Geophys. Res. 84, 2795–2820.

Burk, S. D.: 1976, ‘Diurnal Winds Near the Martian Polar Caps’, J. Atmos. Sci. 33, 923–939.
Capen, C. F.: 1974, ‘A Martian Yellow Cloud, July 1971’, Icarus 22, 345–362.
Carrier, G. F.: 1971, ‘The Intensification of Hurricanes’, J. Fluid Mech. 49, 145–158.
Christensen, P. R.: 1986, ‘Regional Dust Deposits on Mars: Physical Properties, Age, and History’,

J. Geophys. Res. 91, 3533–3545.
Chylek, P. and Grams, G. W.: 1978, ‘Scattering by Nonspherical Particles and Optical Properties of

Martian Dust’, Icarus 36, 198–203.
Clancy, R. T. and Lee, S. W.: 1991, ‘A New Look at Dust and Clouds in the Mars Atmosphere: Analy-

sis of Emission-Phase-Function Sequences from Global Viking IRTM Observations’, Icarus 93,
135–158.

Clancy, R. T., Muhleman, D. O., and Berge, G. L.: 1990, ‘Global Changes in the 0–70 km Thermal
Structure of the Mars Atmosphere Derived from 1975–1989 Microwave CO spectra’, J. Geophys.
Res. 95, 14543–14554.

Clancy, R. T., Lellouch, T. E., Billawala, Y. N., Sandor, B. J., and Rudy, D. J.: 1994, ‘Microwave
Observations of a 1994 Mars Global Dust Storm’, Bull. Am. Astron. Soc. 26, 1130.

Clancy, R. T., Lee, S. W., Gladstone, G. R., McMillan, W. W., and Rousch, T.: 1995, ‘A New Model
for Mars Atmospheric Dust Based Upon Analysis of Ultraviolet Through Infrared Observations
from Mariner 9, Viking, and Phobos’, J. Geophys. Res. 100, 5251–5263.

Clancy, R. T., Grossman, A. W., Wolf, M. J., James, P. B., Rudy, D. J., Billawala, Y. N., Sandor, B.
J., Lee, S. W., and Muhleman, D. O.: 1996, ‘Water Vapor Saturation at Low Altitudes Around
Mars Aphelion: A Key to Mars Climate?’, Icarus 122, 36–62.

Conrath, B. J.: 1975, ‘Thermal Structure of the Martian Atmosphere During the Dissipation of the
Dust Storm of 1971’, Icarus 24, 36–46.

Conrath, B., Curran, R., Hanel, R., Kunde, V., Maguire, W., Pearl, J., Pirraglia, J., Welker, J.,
and Burke, T.: 1973, “Atmospheric and Surface Properties of Mars Obtained by Infrared
Spectroscopy on Mariner 9”, J. Geophys. Res. 78, 4267–4278.

Csanady, G. T.: 1967, ‘On the “Resistance Law” of a Turbulent Ekman Layer’, J. Atmos. Sci. 24,
467–471.

Dollfus, A., Bowell, E., and Ebisawa, S.: 1984a, ‘Polarimetric Analysis of the Martian Dust Storms
and Clouds in 1971’, Astron. Astrophys. 131, 123–136.

Dollfus, A., Bowell, E., and Ebisawa, S.: 1984b, ‘The Martian Dust Storms of 1973: A Polarimetric
Analysis’, Astron. Astrophys. 134, 343–353.

Drossart, P., Rosenqvist, J., Erard, S., Langevin, Y., Bibring, J.-P., and Combes, M.: 1991, ‘Martian
Aerosol Properties from the Phobos ISM Experiment’, Ann. Geophys. 9, 754–760.

Ebisawa, S. and Dollfus, A.: 1986, ‘Martian Dust Storms at the Early Stage of Their Evolution’,
Icarus 66, 75–82.

Edgett, K. S. and Christensen, P. R.: 1991, ‘The Particle Size of Martian Aeolian Dunes’, J. Geophys.
Res. 96, 22,765–22,776.

Fernández, W.: 1995a, ‘Description of the Martian Polar Cap Breeze’, Earth, Moon, and Planets 70,
183–191.

Fernández, W.: 1995b, ‘A Numerical Simulation of the Martian Polar Cap Breeze’, Earth, Moon,
and Planets 70, 193–205.

French, R. G. and Gierasch, P. J.: 1979, The Martian Polar Vortex: Theory of Seasonal Variation and
Observations of Eolian Features’, J. Geophys. Res. 84, 4634–4642.

Gierasch, P. J.: 1974, ‘Martian Dust Storms’, Rev. Geophys. Space Phys. 12, 730–734.



MARTIAN DUST STORMS: A REVIEW 43

Gierasch, P. J. and Goody, R. M.: 1972, ‘The Effect of Dust on the Temperature of the Martian
Atmosphere’, J. Atmos. Sci. 29, 400–402.

Gierasch, P. J. and Goody, R. M.: 1973, ‘A Model of a Martian Great Dust Storm’, J. Atmos. Sci. 30,
169–179.

Gifford, F. A.: 1964, ‘A Study of Martian Yellow Clouds that Display Movement’, Mon. Wea. Rev.
92, 435–439.

Golytsyn, G. S.: 1973, ‘On the Martian Dust Storms’, Icarus 18, 113–119.
Greeley, R., Leach, R., White, B. R., Iversen, J. D., and Pollack, J. B.: 1980, ‘Threshold Wind Speeds

for Sand on Mars: Wind Tunnel Simulations’, Geophys. Res. Lett. 7, 121–124.
Greeley, R., Lancaster, N., Lee, S., and Thomas, P.: 1992, ‘Martian Aeolian Processes, Sediments

and Features’, in H. H. Kieffer, B. M. Jakosky, C. W. Snyder, and M. S. Matthews (eds.), Mars,
University of Arizona Press, Tucson, Arizona, pp. 730–766.

Haberle, R. M.: 1979, ‘The Influence of the Martian Polar Caps on the Diurnal Tide’, Icarus 39,
184–191.

Haberle, R. M. and Jakosky, B. M.: 1991, ‘Atmospheric Effects on the Remote Determination of
Thermal Inertia on Mars’, Icarus 90, 187–204.

Harberle, R. M., Leovy, C. B. and Pollack, J. B.: 1979, ‘A Numerical Model of the Martian Polar
Cap Winds’, Icarus 39, 151–183.

Haberle, R. M., Leovy, C. B., and Pollack, J. B.: 1982, ‘Some Effects of Global Dust Storms on the
Atmospheric Circulation on Mars’, Icarus 50, 322–367.

Haberle, R. M., Pollack, J. B., Barnes, J. R., Zurek, R. W., Leovy, C. B., Murphy, J. R., Lee, H.,
and Schaeffer, J.: 1993, ‘Mars Atmospheric Dynamics as Simulated by the NASA Ames General
Circulation Model, 1. The Zonal Mean Circulation’, J. Geophys. Res. 98, 3093–3123.

Hanel, R. A., Conrath, B. J., Hovis, W. A., Kunde, V. G., Lowman, P. D., Pearl, J. C., Prabhakara,
C., and Schlachman, B.: 1972, ‘Infrared Spectroscopy Experiment on the Mariner 9 Mission:
Preliminary Results’, Science 175, 305–308.

Hartmann, W. K. and Price, M. J.: 1974, ‘Mars: Clearing of the 1971 Dust Storm’, Icarus 21, 28–34.
Hess, S. L.: 1973, “Martian Winds and Dust Clouds”, Planetary Space Sci. 21, 1549–1557.
Hunt, G. E., Mitchell, E. A., and Peterfreund, A. R.: 1980, ‘The Opacity of Some Local Martian Dust

Storms Observed by the Viking IRTM’, Icarus 41, 389–399.
Ingersoll, A. P. and Lyons, J. R.: 1993, ‘Mars Dust Storms: Interannual Variability and Chaos’, J.

Geophys. Res. 98, 10,951–10,961.
Iversen, J. D., Pollack, J. B., Greeley, R., and White, B. R.: 1976a, “Saltation Threshold on Mars:

The Effect of Interparticle Force, Surface Roughness, and Low Atmospheric Density’, Icarus 29,
381–393.

Iversen, J. D., Greeley, R., and Pollack, J. B.: 1976b, ‘Windblown Dust on Earth, Mars, and Venus’,
J. Atmos. Sci. 33, 2425–2429.

James, P. B.: 1985, ‘Martian Local Dust Storms’, in G. Hunt (ed.), Recent Advances in Planetary
Meteorology, Cambridge University Press, New York, pp. 85–100.

James, P. B. and Evans, N.: 1981, ‘A Local Dust Storm in the Chryse Region of Mars: Viking Orbiter
Observations’, Geophys. Res. Lett. 8, 903–906.

Jaquin, F., Gierasch, P., and Kahn, R.: 1986, ‘The Vertical Structure of Limb Hazes in the Martian
Atmosphere’, Icarus 68, 442–461.

Kahn, R. A., Martin, T. Z., Zurek, R. W., and Lee, S. W.: 1992, ‘The Martian Dust Cycle’, in H. H.
Kieffer, B. M. Jakosky, C. W. Snyder and M. S. Matthews (eds.), Mars, University of Arizona
Press, Tucson, Arizona, pp. 1017–1053.

Kieffer, H. H., Chase Jr., S. C., Miner, E., Münch, G., and Neugebauer, G.: 1973, ‘Preliminary
Report on Infrared Radiometric Measurements from the Mariner 9 Spacecraft’, J. Geophys. Res.
78, 4291–4312.



44 WALTER FERNÁNDEZ

Leovy, C. and Mintz, Y.: 1969, ‘Numerical Simulation of the Atmospheric Circulation and Climate
of Mars’, J. Atmos. Sci. 26, 1167–1190.

Leovy, C. B. and Zurek, R. W.: 1979, ‘Thermal Tides and Martian Dust Storms: Direct Evidence for
Coupling’, J. Geophys. Res. 84, 2956–2968.

Leovy, C. B., Briggs, G., Young, A., Smith, B., Pollack, J., Shipley, E., and Wildey, R.: 1972, ‘The
Martian Atmosphere: Mariner 9 Television Experiment Progress Report’, Icarus 17, 373–393.

Leovy, C. B., Zurek, R. W., and Pollack, J. B.: 1973, ‘Mechanisms for Mars Dust Storms’, J. Atmos.
Sci. 30, 749–762.

Magalhaes, J. and Gierasch, P.: 1982, ‘A Model of Martian Slope Winds: Implications for Eolian
Transport’, J. Geophys. Res. 87, 9975–9984.

Martin, L. J.: 1974a, ‘The Major Martian Yellow Storm of 1971’, Icarus 22, 175–188.
Martin, L. J.: 1974b, ‘The Major Martian Dust Storms of 1971 and 1973’, Icarus 23, 108–115.
Martin, L. J.: 1976, ‘1973 Dust Storm on Mars: Maps from Hourly Photographs’, Icarus 29, 363–380.
Martin, L. J.: 1984, ‘Clearing the Martian Air: The Troubled History of Dust Storms’, Icarus 57,

317–321.
Martin, L. J. and Zurek, R. W.: 1993, ‘An Analysis of the History of Dust Activity on Mars’, J.

Geophys. Res. 98, 3221–3246.
Martin, L. J., Beish, J. D., and Parker, D. C.: 1991a, ‘Martian Dust Storms in 1990’, Bull. Am. Astron.

Soc. 23, 1217.
Martin, L. J., James, P. B., and Zurek, R. W.: 1991b, ‘Observed Changes in Limb Clouds Immedi-

ately Prior to the Onset of Planet-Encircling Dust Storms’, in MSATT Workshop on the Martian
Surface and Atmosphere through Time, Lunar and Planetary Institute, Houston, Texas, p. 86.

Martin, T. Z.: 1986, ‘Thermal Infrared Opacity of the Mars Atmosphere’, Icarus 66, 2–21.
Martin, T. Z.: 1995, ‘Mass of Dust in the Martian Atmosphere’, J. Geophys. Res. 100, 7509–7512.
Martin, T. Z. and Richardson, M. I.: 1993, ‘New Dust Opacity Mapping from Viking Infrared

Thermal Mapper Data’, J. Geophys. Res. 98, 10941–10949.
Martin, T., Peterfreund, A., Miner, E., Kieffer, H., and Hunt, G.: 1979, ‘Thermal Infrared Properties

of the Martian Atmosphere. 1. Global Behavior at 7, 9, 11, and 20 µm’, J. Geophys. Res. 84,
2830–2842.

Moore, H. J.: 1985, ‘The Martian Dust Storm of Sol 1742’, J. Geophys Res. 90, Supplement, D163–
D174.

Moroz, V. I. and Ksanfomaliti, L. V.: 1972, “Preliminary Results of Astrophysical Observations of
Mars from Mars 3’, Icarus 17, 408–422.

Murphy, J. R., Toon, O. B., Haberle, R. M., and Pollack, J. B.: 1990, ‘Numerical Simulations of the
Decay of Martian Global Dust Storms’, J. Geophys. Res. 95, 14,629–14,648.

Murphy, J. R., Haberle, R. M., Toon, O. B., and Pollack, J. B.: 1993, ‘Martian Global Dust Storms:
Zonally Symmetric Numerical Simulations Including Size-Dependent Particle Transport’, J.
Geophys. Res. 98, 3197–3220.

Neubauer, F. M.: 1966, ‘Thermal Convection in the Martian Atmosphere’, J. Geophys. Res. 71, 2419–
2426.

Ockert-Bell, M. E., Bell III, J. F., Pollack, J. B., McKay, C. P., and Forget, F.: 1997, ‘Absorption
and Scattering Properties of the Martian Dust in the Solar Wavelengths’, J. Geophys. Res. 102,
9039–9050.

Paige, D. A., Herkenhoff, H. E., and Murray, C. B.: 1987, ‘Martian Dust Properties from Mariner 9
IRIS Observations of the South Polar Cap’, Bull. Am. Astron. Soc. 19, 815.

Pallmann, A. J. and Frisella, S. P.: 1972, ‘Numerical Simulation of Radiative-Conductive Heat
Transfer in the Martian Atmosphere-Polar Cap Utilizing Mariner 9 IRIS Data’, Conference on
Atmospheric Radiation, August 7–9, Fort Collins, Colorado; published by AMS, Boston, Mass.

Pallmann, A. J., Dannevik, W. P., and Frisella, S. P.: 1973, ‘Martian Thermal Boundary Layers:
Subhourly Variations Induced by Radiative-Conductive Heat Transfer within the Dust-Laden
Atmosphere-Ground System’, NASA Contractor Report, Washington, D.C.



MARTIAN DUST STORMS: A REVIEW 45

Pang, K. and Hord, C. W.: 1973, ‘Mariner 9 Ultraviolet Spectrometer Experiment: 1971 Dust Storm’,
Icarus 18, 481–488.

Pank, K., Ajello, J. M., Hord, C. W., and Egan, W. G.: 1976, ‘Complex Refractive Index of Martian
Dust: Mariner 9 Ultraviolet Observations’, Icarus 27, 55–67.

Peterfreund, A. R.: 1981, ‘Visual and Infrared Observations of Wind Streaks on Mars’, Icarus 45,
447–467.

Pollack, J., Colburn, D., Kahn, R., Hunter, J., Van Camp, W., Carlston, C., and Wolfe, M.: 1977,
‘Properties of Aerosols in the Martian Atmosphere as Inferred from Viking Lander Imaging
Data’, J. Geophys. Res. 82, 4479–4496.

Pollack, J., Colburn, D., Flasar, F. M., Kahn, R., Carlston, C., and Pidek, D.: 1979, ‘Properties and
Effects of Dust Particles Suspended in the Martian Atmosphere’, J. Geophys. Res. 84, 2929–
2945.

Pollack, J. B., Haberle, R. M., Schaeffer, J., and Lee, H.: 1990, ‘Simulations of the General
Circulation of the Martian Atmosphere 1. Polar Processes’, J. Geophys. Res. 95, 1447–1474.

Pollack, J. B., Haberle, R. M., Murphy, J. R., Schaeffer, J., and Lee, H.: 1990, ‘Simulations of the
General Circulation of the Martian Atmosphere 2. Seasonal Pressure Variations’, J. Geophys.
Res. 95, 3149–3181.

Pollack, J. B., Ockert-Bell, M. E., and Shepard, M. K.: 1995, ‘Viking Lander Image Analysis of
Martian Atmospheric Dust’, J. Geophys. Res. 100, 5235–5250.

Presley, M. A. and Christensen, P. R.: 1997a, ‘Thermal Conductivity Measurements of Particulate
Materials, 1, A Review’, J. Geophys. Res. 102, 6535–6549.

Presley, M. A. and Christensen, P. R.: 1997b, ‘Thermal Conductivity Measurements of Particulate
Materials, 2, Results’, J. Geophys. Res. 102, 6551–6566.

Presley, M. A. and Christensen, P. R.: 1997c, ‘The Effect of Bulk Density and Particulate Size Sorting
on the Thermal Conductivity of Particulate Materials Under Martian Atmospheric Pressures’, J.
Geophys. Res. 102, 9221–9229.

Ryan, J. A.: 1964, ‘Notes on the Martian Yellow Clouds’, J. Geophys. Res. 69, 3759–3770.
Ryan, J. A. and Henry, R. M.: 1979, ‘Mars Atmospheric Phenomena During Major Dust Storms, as

Measured at Surface’, J. Geophys. Res. 84, 2821–2829.
Ryan, J. A. and Lucich, R. D.: 1983, ‘Possible Dust Devils, Vortices on Mars’, J. Geophys. Res. 88,

11,005–11,011.
Ryan, J. A. and Sharman, R. D.: 1981, ‘Two Major Dust Storms, One Mars Year Apart: Comparison

from Viking Data’, J. Geophys. Res. 86, 3247–3254.
Ryan, J. A., Henry, R. M., Hess, S. L., Leovy, C. B., Tilllman, J. E., and Walcek, C.: 1978, ‘Mars

Meteorology: Three Seasons at the Surface’, Geophys. Res. Lett. 5, 715–718.
Ryan, J. A., Sharman, R. D., and Lucich, R. D.: 1981, ‘Local Mars Dust Storm Generation

Mechanism’, Geophys. Res. Lett. 8, 899–901.
Sagan, C. and Pollack, J. B.: 1967, ‘A Winblown Dust Model of Martian Surface Features and

Seasonal Changes’, Smithsonian Astrophysical Observatory, Special Report 255, Cambridge,
Mass.

Sagan, C. and Pollack, J. B.: 1969, ‘Windblown Dust on Mars’, Nature 223, 791–794.
Sagan, C., Veverka, J., and Gierasch, P.: 1971, ‘Observational Consequences of Martian Wind

Regimes’, Icarus 15, 253–278.
Sagan, C., Veverka, J., Fox, P., Dubisch, R., French, R., Gierasch, P., Quam, L., Lederberg, J.,

Levinthal, E., Tucker, R., Eross, B., and Pollak, J.: 1973, ‘Variable Features on Mars. 2. Mariner
9 Global Results’, J. Geophys. Res. 78, 4163–4196.

Santee, M. and Crisp, D.: 1992, ‘Atmospheric and Surface Temperatures and Airborne Dust from
Mariner 9 IRIS Data’, in R. M. Haberle et al. (eds.), Workshop on the Martian Surface and At-
mosphere Through Time (MSATT), LPI Tech. Rep. 92-02, Lunar and Planetary Institute, Houston,
pp. 128–130.



46 WALTER FERNÁNDEZ

Santee, M. and Crisp, D.: 1993, ‘Thermal Structure of Dust Loading of the Martian Atmosphere
During Late Southern Summer: Mariner 9 Revisited’, J. Geophys. Res. 98, 3261–3279.

Schneider, E. K.: 1983, ‘Martian Great Dust Storms: Interpretive Axially Symmetric Models’, Icarus
55, 302–331.

Schofield, J. T., Barnes, J. R., Crisp, D., Haberle, R. M., Larsen, S., Magalhaes, J. A., Mur-
phy, J. R., Seiff, A., and Wilson, G.: 1997, ‘The Mars Pathfinder Atmospheric Structure
Investigation/Meteorology (ASI/MET) Experiment’, Science 278, 1752–1758.

Smith, P. H., Bell III, J. F., Bridges, N. T., Britt, D. T., Gaddis, L., Greeley, R., Keller, H. U.,
Herkenhoff, K. E., Jaumann, R., Johnson, J. R., Kirk, R. L., Lemmon, M., Maki, J. N., Malin, M.
C., Murchie, S. L., Oberst, J., Parker, T. J., Reid, R. J., Sablotny, R., Soderblom, L. A., Stoker,
C., Sullivan, R., Thomas, N., Tomasko, M. G., Ward, W., and Wegryn, E.: 1997, ‘Results from
the Mars Pathfinder Camera’, Science 278, 1758–1765.

Thomas, P. and Gierasch, P. J.: 1985, ‘Dust Devils on Mars’, Science 230, 175–177.
Thorpe, T. E.: 1978, ‘Viking Orbiter Observations of the Mars Opposition Effect’, Icarus 36, 204–

215.
Thorpe, T. E.: 1981, ‘Mars Atmospheric Opacity Effects Observed in the Northern Hemisphere by

Viking Orbiter Imaging’, J. Geophys. Res. 86, 11,419–11,429.
Tillman, J. E., Henry, R. M., and Hess, S. L.: 1979, ‘Frontal Systems During Passage of the Martian

North Polar Hood over the Viking Lander 2 Site Prior to the First 1977 Dust Storm’, J. Geophys.
Res. 84, 2947–2955.

Toon, O. B., Pollack, J. B., and Sagan, C.: 1977, ‘Physical Properties of the Particles Composing the
Martian Dust Storm of 1971–72’, Icarus 30, 663–696.

Wolf, M. J., Lee, S. W., Clancy, R. T., Martin, L. J., Bell III, J. F., and James, P. B.: 1997, ‘1995
Observations of Martian Dust Storms Using Hubble Space Telescope’, J. Geophys. Res. 102,
1679–1692.

Zurek, R. W.: 1981, ‘Inference of Dust Opacities for the 1977 Martian Great Dust Storms from
Viking Lander 1 Pressure Data’, Icarus 45, 202–215.

Zurek, R. W.: 1982, ‘Martian Great Dust Storms: An Update’, Icarus 50, 288–310.
Zurek, R. W., and Martin, L. J.: 1993, ‘Interannual Variability of Planet-Encircling Dust Storms on

Mars’, J. Geophys. Res. 98, 3247–3259.
Zurek, R. W., Barnes, J. R., Haberle, R. M., Pollack, J. B., Tillman, J. E., and Leovy, C. B.: 1992,

‘Dynamics of the Atmosphere of Mars’, in H. H. Kieffer, B. M. Jakosky, C. W. Snyder and M. S.
Matthews (eds.), Mars, University of Arizona Press, Tucson, Arizona, pp. 835–933.