The first 5 pages, shown as text so search engines and readers can find this document.
Page 1
The spatial distribution of planetary ion fluxes
near Mars observed by MAVEN
D. A. Brain1, J. P. McFadden2, J. S. Halekas3, J. E. P. Connerney4, S. W. Bougher5, S. Curry2, C. F. Dong5,
Y. Dong1, F. Eparvier1, X. Fang1, K. Fortier1, T. Hara2, Y. Harada2, B. M. Jakosky1, R. J. Lillis2, R. Livi2,
J. G. Luhmann2, Y. Ma6, R. Modolo7, and K. Seki8
1Laboratory for Atmospheric and Space Physics, University of Colorado Boulder, Boulder, Colorado, USA, 2Space Sciences
Laboratory, University of California, Berkeley, California, USA, 3Department of Physics and Astronomy, University of Iowa,
Iowa City, Iowa, USA, 4NASA Goddard Space Flight Center, Greenbelt, Maryland, USA, 5Department of Atmospheric,
Oceanic, and Space Sciences, University of Michigan, Ann Arbor, Michigan, USA, 6Institute of Geophysics and Planetary
Physics, University of California, Los Angeles, California, USA, 7UVSQ/LATMOS-IPSL/CNRS-INSU, Guyancourt, France, 8Solar-
Terrestrial Environment Laboratory, Nagoya University, Nagoya, Japan
Abstract We present the results of an initial effort to statistically map the fluxes of planetary ions on a
closed surface around Mars. Choosing a spherical shell ~1000 km above the planet, we map both outgoing
and incoming ion fluxes (with energies >25 eV) over a 4 month period. The results show net escape of
planetary ions behind Mars and strong fluxes of escaping ions from the northern hemisphere with respect to
the solar wind convection electric field. Planetary ions also travel toward the planet, and return fluxes are
particularly strong in the southern electric field hemisphere. We obtain a lower bound estimate for planetary
ion escape of ~3 × 1024 s1, accounting for the ~10% of ions that return toward the planet and assuming that
the ~70% of the surface covered so far is representative of the regions not yet visited by Mars Atmosphere
and Volatile EvolutioN (MAVEN).
1. Introduction
The escape of atmospheric particles to space may have been especially effective at Mars, relative to Earth or
Venus, due to a combination of its small size and lack of global magnetic field. The small size leads to a low
escape energy for atmospheric particles, and the lack of global magnetic field allows the incident solar wind
to approach much closer to the planet than at Earth, so that the solar wind interacts directly with the Martian
upper atmosphere. Considering that planetary ions may be more effectively removed from the atmosphere
in this interaction, a major goal over the past few decades has been to accurately evaluate the loss of ions
from Mars under a variety of solar and solar wind conditions. One reason for doing so is to determine whether
an early greenhouse atmosphere could have been removed over billions of years, causing the Martian
climate to transition from a state that allowed liquid water to play a prominent role in shaping the surface
to today’s inhospitable surface environment [Pollack et al., 1987; Jakosky and Phillips, 2001].
Previous spacecraft measurements have established that atmospheric ions escape the planet and have been
used to map the spatial distribution of ions around the planet. The Phobos 2 mission provided the first direct
measurements of escaping O+ ions [Lundin et al., 1990] and presented the idea that there is spatial structure
to the escaping ion energies and fluxes [Kallio et al., 1995]. Mars Express has measured planetary ions for an entire
solar cycle and continues to provide measurements of escape rates and the spatial distribution of ions down-
stream from and near the planet [e.g., Barabash et al., 2007; Fränz et al., 2007; Lundin et al., 2008; Fränz et al.,
2010; Nilsson et al., 2010, 2011; Ramstad et al., 2015]. These measurements lead to the present understanding
of the existence of a variety of acceleration processes for ions and different channels for escape [e.g., Dubinin
et al., 2011]. Estimates of the total present-day ion escape rate range from ~1023 s1 to 3 × 1025 s1 [Dubinin
et al., 2011, and references therein]. Discrepancies in the determination of escape rate depend partly upon the
energy range of particles considered. Both escape rates and the spatial distribution of ions near Mars provide
important constraints for models of ion loss from Mars. These models, validated by observations, can be used
to estimate atmospheric loss rates at earlier epochs in Martian history [Lillis et al., 2015].
The Mars Atmosphere and Volatile EvolutioN (MAVEN) mission arrived at Mars in September 2014 and began
making measurements of the Martian upper atmosphere and escaping particles, with the objective of
BRAIN ET AL. MARS ION ESCAPE MAPS 1
PUBLICATIONS
Geophysical Research Letters
RESEARCH LETTER
10.1002/2015GL065293
Special Section:
First Results from the MAVEN
Mission to Mars
Key Points:
• MAVEN ion measurements are
mapped to a spherical surface
around Mars
• Planetary ion fluxes are organized in
four spatial regions on the shell
• Heavy ion escape rates exceed
2 × 1024 s1 for energies >25 eV
Correspondence to:
D. A. Brain,
david.brain@lasp.colorado.edu
Citation:
Brain, D. A., et al. (2015), The spatial
distribution of planetary ion fluxes near
Mars observed by MAVEN, Geophys. Res.
Lett., 42, doi:10.1002/2015GL065293.
Received 8 JUL 2015
Accepted 18 SEP 2015
©2015. American Geophysical Union. All
Rights Reserved.
Page 2
evaluating the importance of atmospheric escape in shaping Martian climate [Jakosky et al., 2015]. MAVEN
should build upon the successes of previous missions in three ways. First, MAVEN carries instruments that
directly measure the solar and solar wind drivers of escape, including magnetic field measurements that
can be used to organize the spatial distribution of escaping ions. Second, MAVEN’s planetary ion measure-
ments typically occur at higher cadence (16 s or better) than recent missions, allowing better determination
of the 3-D distribution of ions on smaller spatial and temporal scales. Third, as a mission entirely dedicated to
the upper atmosphere and escape, MAVEN carries a full complement of instruments capable of probing the
physics of ion acceleration. One goal of MAVEN is to use the measured variability in atmospheric escape rates
via each loss process to extrapolate back in time [Jakosky et al., 2015; Lillis et al., 2015].
Here we present the first results of an effort to spatially map the measured fluxes of planetary ions observed
by MAVEN near Mars in a way that allows an estimation of the ion escape rate. We separately map ions
moving toward and away from the planet and construct maps of the net flux of ions. We use these maps to place
initial constraints on present-day global escape rates and place them in context with previous measurements.
2. Methodology
Our analysis includes measurements from the start of MAVEN science operations on 16 November 2014,
through 1 April 2015. The orbital coverage of MAVEN for this time period is shown in Figure 1.
We map MAVEN planetary ion measurements to a closed surface around the planet, in contrast to most pre-
viously published studies. This approach is taken for two reasons: (1) we seek to evaluate the flux of ions both
leaving and approaching the planet, suggesting the use of a closed surface, and (2) a “plume” of escaping
ions detected by MAVEN apparently contains significant ion fluxes on trajectories that would not intersect
the more traditional planar surface behind Mars [Dong et al., 2015]. In the coming studies we plan to compare
a variety of surfaces for mapping ion escape from Mars. Here we choose a spherical surface that ranges from
1.25 to 1.45 Mars radii (RM) or ~850–1530 km altitudes. This radial distance is chosen so that all nightside solar
zenith angles are sampled on the surface, as shown in Figure 1.
We select times when MAVEN was located on the spherical shell (i.e., between 1.25 and 1.45 RM). These per-
iods are mapped in the often used Mars-Solar-Orbital (MSO) coordinate system in Figure 2 (left), corrected for
the aberration of the solar wind at Mars by using the upstream solar wind velocity measured by the Solar
Wind Ion Analyzer (SWIA) [Halekas et al., 2013]. The coverage spans a relatively limited range of local times
Figure 1. MAVEN’s orbit coverage from 16 November 2014 to 1 April 2015. Coverage is shown in cylindrical coordinates
relative to the Mars-Sun line, with each orbit chronologically overplotted. Dashed lines indicate the inner and outer
boundaries of a spherical shell around Mars used to analyze ion fluxes for this study.
Geophysical Research Letters 10.1002/2015GL065293
BRAIN ET AL. MARS ION ESCAPE MAPS 2
Page 3
and solar latitudes. Ion motion near Mars, however, is believed to be largely structured by the directions of
the solar wind flow and interplanetary magnetic field (IMF). Therefore, we rotate MAVEN’s location at each
time into a Mars-Solar-Electric field (MSE) coordinate system, defined on an orbit-by-orbit basis using
MAVEN upstream measurements of the solar wind velocity and IMF direction recorded by SWIA and the
Magnetometer (MAG) [Connerney et al., 2014]. MAVEN’s coverage of the shell around Mars is much more
complete in this coordinate system. Of necessity, we exclude time periods for which SWIA and MAG estimates
of the solar wind and IMF are unavailable. This is the primary reason that we only include observations
through March 2015; MAVEN’s orbit precession did not allow it to observe the solar wind after this time for
a period of a few months.
Measured ion fluxes are taken from the Suprathermal and Thermal Ion Composition (STATIC) instrument on
MAVEN. STATIC is an electrostatic analyzer with time-of-flight measurement capability (J. McFadden et al.,
The MAVEN Suprathermal and thermal Ion Composition (STATIC) Instrument, submitted to Space Science
Reviews, 2015) that can record ion fluxes as a function of energy (0.1eV–20 keV), mass (1–70 amu), azimuth
direction (0–360°), and deflection angle (~±45°). We use measurements recorded when the instrument was
in “pickup ion mode,” which provides a good combination of energy (32 channels), mass (8 mass bins),
and angular (4 defections and 16 azimuth directions) resolution. Wherever possible, archive data with 16 s
cadence are used. When archive data are not available (~3.5% of the time), we use survey data with cadence
of 512 s. We omit observations for which the estimated background was abnormally high or the attenuator chan-
ged state during the observation (leading to an invalid geometric factor), leaving a total of 20,836 observations
over an ~4 month period.
We treat each STATIC observation in our study individually before including it in maps of ion fluxes. Our ana-
lysis treats only heavy ions, with mass/charge of 9 amu or greater. First, we apply a rough procedure to
remove instrument counts from “straggling protons.” Straggling, described in J. McFadden et al. (submitted
manuscript, 2015), results in protons being recorded at higher masses due to the energy loss of particles in
the foils used in STATIC’s time-of-flight system. Sophisticated background subtraction techniques are
presently being developed for STATIC data; here we take a conservative approach of subtracting 20% (a value
arrived at after manual inspection of many distributions) of the counts recorded at masses less than 2.5 amu
from all higher mass bins at each energy and visually inspecting the results. This likely results in a slight
oversubtraction of heavy ion counts at energies where proton fluxes are greatest but is preferable to the
alternative of analyzing flux in heavy ion mass bins that come from protons. Overall uncertainty in the
determination of flux is estimated to be 50% or less.
We do not correct the measurements for spacecraft potential or spacecraft velocity effects. While MAVEN can
measure spacecraft potential via several methods, a mature determination of spacecraft potential has not yet
been incorporated into our analysis. Charging and velocity effects most strongly influence low-energy ion
Figure 2. MAVEN orbital coverage on a shell extending from 1.25 to 1.45 RM, for time periods from 16 November 2014 to 1
April 2015. (left) Coverage in MSO coordinates, corrected for the solar wind aberration at Mars. (right) Coverage in MSE
coordinates with the upstream electric field direction parallel with the y axis of the plot. Dayside regions are on the left half
of both panels with the subsolar point at (0,0), and nightside regions are on the right half of both panels with the antisolar
point at (180,0).
Geophysical Research Letters 10.1002/2015GL065293
BRAIN ET AL. MARS ION ESCAPE MAPS 3
Page 4
observations. We therefore do not include measurements below 25 eV in our study. Further discussion is
provided in Harada et al. [2015].
Finally, we separate each measured distribution into the portions moving toward or away from the planet
with respect to our surface, respectively. The vector ion flux for each portion of the distribution is computed
(via direct integration) and the radial component of the flux mapped into 5° × 5° spatial bins on our surface.
Separate maps for incident and escaping ions are constructed.
3. Results
Figure 3 shows the results of mapping the planetary heavy (m/q > ~9) ion fluxes measured by STATIC onto
the spherical shell around Mars. Overall, ~68% of the surface area of the shell is covered by the maps, and
a median of seven observations occupies each populated bin, with a maximum of 376 observations in a sin-
gle bin. As the MAVEN mission progresses, the surface area coverage of the maps should increase, as well as
the number of observations in each bin.
A map of the median outward ion flux in each bin is shown in Figure 3a. We select median values in this study
as proxies for the “typical” flux in each bin and as values less likely to be influenced by extreme values than a
mean flux. If we instead choose mean flux, we obtain higher values, but the relative variations in the spatial
distribution of planetary ions are not significantly changed. The flux of outward ions from the central portion
of the dayside of the planet is low compared to the flux on the nightside and at high northern (in MSE coor-
dinates) dayside latitudes.
Figure 3b shows a map of the median inward flux. The inward fluxes appear to be more uniformly distributed
than the outward fluxes, with the exception of the southern electric field hemisphere on the dayside where
fluxes are higher than in other regions.
The net flux in each bin is obtained by taking the difference between the median outward and inward fluxes.
The results are shown in Figure 3c. A few different spatial regions are evident from the figure. On the night-
side of the planet, outward flux generally dominates inward flux and the net flow of particles is away from the
Figure 3. Measured ion fluxes mapped to a spherical shell around Mars in 5° × 5° bins, in the same MSE coordinate system used in Figure 2. The subsolar point is at
(0,0) and the antisolar point at (180,0). (a) Median outward ion fluxes. (b) Median inward ion fluxes. (c) Net flux in each bin, obtained by subtracting the inward flux
map from the outward flux map. (d) The number of observations in each bin.
Geophysical Research Letters 10.1002/2015GL065293
BRAIN ET AL. MARS ION ESCAPE MAPS 4
Page 5
planet. On the dayside at low (MSE) latitudes, inward flux weakly dominates escape flux, so that the net flow
of particles is toward the planet. At higher latitudes, especially in the northern (MSE) hemisphere, outward
flux strongly dominates return flux.
In addition to mapping the measurements spatially, we can sum the total ion flux in various regions and use
them to estimate an escape rate. Guided by Figure 3 and expectations from simulations [e.g., Fang et al.,
2008], we choose four regions: the north and south electric field polar regions (MSE latitude >±45°), with
the remaining low latitudes divided into dayside and nightside regions. Table 1 reports the total outward,
inward, and net ion rates in each region, as well as the global rates. The totals are obtained by summing
the median flux in each bin, multiplied by the area of the bin. Table 1 also gives a range in rates for each
region, provided as a rough estimate of the variability in each region. The range is obtained by computing
the lower and upper quartile fluxes in each 5° × 5° spatial bin and then summing over all lower quartile values
and upper quartile values.
Globally, the outward ion rate exceeds the inward rate by a factor of several, so that the net flow of planetary
ions through the spherical shell is away from Mars. Despite some individual spatial bins being dominated by
returning ion flux, the net flow is outward in all four of the large regions we have defined, though the rate
from both the dayside and the southern MSE region is low and, considering the range of measured values,
consistent with no net outward flow of particles. The inward rate is comparable on the dayside and nightside
but is about 3 times greater in the southern MSE region than in the north. The outward rate is 2.5–3 times
larger on the nightside than the dayside and ~2 times larger in the northern MSE region than the south.
Approximately 70% of the surface area of the shell has been sampled overall, but the coverage is nonuniform
in each region. The results of Table 1 do not incorporate any estimate of the fluxes or escape rates in regions
not yet visited by the MAVEN spacecraft. Thus, rates from the dayside and nightside are directly comparable
given the similar coverage in those regions, as are rates from the north and south electric field regions. When
we correct for the sampling bias by assuming that the median rate in each partially sampled region is repre-
sentative of the rate in the entire region, we see that the majority of the net outward flow of ions is through
the nightside (47%) and the northern electric field (31%) regions.
4. Discussion
We have been cautious to this point to refer to the mapped ion fluxes as outward and inward, but our goal is
to evaluate the escape rate of planetary ions. Different spatial regions are evident in the maps in Figure 3.
First, on the nightside of Mars, heavy ions escape from the planet, with a small fraction (~10–15%) returning
toward the planet. Second, in the north electric field region, newly born ions escape the planet in the pickup
ion plume [Dong et al., 2015] and are carried northward in MSE coordinates by the v × B electric field in the
plasma. This results in significant escaping fluxes seen in the dayside northern hemisphere. Third, a signifi-
cant fraction of particles in the southern MSE hemisphere have trajectories that carry them inside our sphe-
rical surface. Some of these particles impact the planet in the exobase region, and others escape. We
conducted test particle calculations similar to those presented in Fang et al. [2008], which suggested that
Table 1. Heavy Ion Rates Measured by MAVENa
Dayside Nightside +E Pole E Pole Global
Coverage 74% 71% 57% 54% 68%
Outward rate 4.5 9.7 4.6 2.0 21
(2.4–9.3) (5.8–15) (2.5–8.3) (1.4–3.1) (12–36)
Inward rate 1.7 1.6 0.42 1.1 4.8
(.87–3.1) (1.1–2.5) (.29–.68) (.84–1.6) (3.1–7.9)
Net 2.8 8.1 4.2 0.90 16
(–.76–8.4) (3.3–14) (1.8–8.0) (.20–2.2) (4.2–33)
aTotal heavy ion rates through the spherical shell obtained from the maps in Figure 3, in units of 1023 s1.The sphe-
rical shell is divided into four regions defined in the text, and median outward, median inward, and net outward rates are
computed for each, as well as global values for the entire shell. Ranges for each value are also provided, computed by
summing the lower quartile and the upper quartile fluxes in each 5° × 5° spatial bin.
Geophysical Research Letters 10.1002/2015GL065293
BRAIN ET AL. MARS ION ESCAPE MAPS 5