Research Article

Hydrogen escape from Mars is driven by seasonal and dust storm transport of water

See allHide authors and affiliations

Science  13 Nov 2020:
Vol. 370, Issue 6518, pp. 824-831
DOI: 10.1126/science.aba5229

Dust storms cause Mars to lose water

Mars was once a wet planet, but it has lost most of its water through reactions that produce hydrogen, which escapes from the upper atmosphere into space. Stone et al. used data from the Mars Atmosphere and Volatile Evolution spacecraft to study how water is transported to the upper atmosphere and converted to hydrogen. They found that water can reach higher altitudes than previously thought, especially during global or regional dust storms. Photochemical modeling shows that this process dominates the current loss of water from Mars and influenced the evolution of its climate.

Science, this issue p. 824


Mars has lost most of its once-abundant water to space, leaving the planet cold and dry. In standard models, molecular hydrogen produced from water in the lower atmosphere diffuses into the upper atmosphere where it is dissociated, producing atomic hydrogen, which is lost. Using observations from the Neutral Gas and Ion Mass Spectrometer on the Mars Atmosphere and Volatile Evolution spacecraft, we demonstrate that water is instead transported directly to the upper atmosphere, then dissociated by ions to produce atomic hydrogen. The water abundance in the upper atmosphere varied seasonally, peaking in southern summer, and surged during dust storms, including the 2018 global dust storm. We calculate that this transport of water dominates the present-day loss of atomic hydrogen to space and influenced the evolution of Mars’ climate.

The enrichment of D/H in the martian atmosphere and surface, relative to Earth, indicates that most of Mars’ initial reservoir of H2O has been lost to space (14). The standard scenario to explain this escape is that H2O is confined to low altitudes below a hygropause (a cold layer at 40- to 50-km altitude at which H2O condenses), where photodissociation and HOx chemistry produce H2. The H2 then diffuses through the hygropause into the upper atmosphere where it is destroyed, producing H, which escapes to space (2, 57). This model would produce steady H2 diffusion and H escape. Observations have shown rapid order-of-magnitude variations in the abundance of H in the exosphere, the outermost region of the atmosphere, indicating that H escape varies seasonally and during dust storms (810). A possible cause for rapid variation in the exospheric H abundance and escape rate is direct transport of H2O—the most abundant hydrogen-bearing species in the martian atmosphere (2)—into the middle and upper atmosphere, where it could be destroyed by reactions with ions to produce H (11). This transport would require a weakening of the hygropause, which could be warmed and raised in altitude as a result of increased solar irradiation around perihelion and/or heating caused by dust in the atmosphere. The middle-atmospheric (~15- to 90-km altitude) H2O abundance is often higher than that implied by the equilibrium vapor pressure at the hygropause temperature (12, 13) and varies seasonally (14, 15), and transient events substantially alter the vertical distribution of H2O (14, 16, 17).

We test this hypothesis using data from the Neutral Gas and Ion Mass Spectrometer (NGIMS) (18) on the Mars Atmosphere and Volatile EvolutioN (MAVEN) spacecraft (19). NGIMS samples neutral and ionic species in situ as MAVEN flies through the martian upper atmosphere. We search for the chemical intermediates that lie between H2O delivered from the lower atmosphere and H escape from the top of the atmosphere; for example, H2O+ and H3O+.

In situ MAVEN measurements

The position of MAVEN’s periapsis (i.e., closest approach to Mars) varied over the course of the mission, as shown in Fig. 1. Regional dust storms occurred during Mars years (MYs) 32 and 33, and a global dust storm occurred in MY 34, also shown in Fig. 1. MAVEN instruments directly sample the atmosphere during the periapse segment of its 4.5-hour orbit, in which periapsis occurs at roughly 150-km altitude (19, 20). During the observations of the last regional dust storm of MY 32, MAVEN’s periapsis precessed from 18°N to 8°N latitude, with measurements at the onset of the dust storm obtained near 13°N. This regional storm arose in the southern hemisphere near martian solar longitude [Mars-Sun angle (LS)] 313°, well after perihelion (LS = 251°) and southern summer solstice (LS = 270°). Amid the global dust storm of MY 34, MAVEN’s periapsis precessed from 21°S to 17°S latitude, with measurements at the onset of the dust storm obtained at ~18°S. This global dust storm occurred around LS = 189°, before perihelion and just after southern spring equinox (LS = 180°). Figure 1A shows the total column dust optical depth from a combination of instruments: the Thermal Emission Spectrometer on Mars Global Surveyor, the Thermal Emission Imaging System on Mars Odyssey, and the Mars Climate Sounder (MCS) on Mars Reconnaissance Orbiter (21, 22). Figure 1B shows the dayside temperature in the middle atmosphere at the 50-Pa pressure level measured by MCS (23, 24). These quantities indicate the onset, evolution, and distribution in latitude of the dust events and the atmospheric response.

Fig. 1 MAVEN location and atmospheric conditions during martian dust storms.

(A) Gridded total column dust optical depth (red to yellow) (21, 22) and (B) mean dayside temperature (red to yellow with black contour at 200 K) at the 50-Pa pressure level from MCS measurements (23, 24) between October 2014 and November 2018. MAVEN’s periapsis positions are indicated with blue crosses. Gaps in MAVEN’s periapsis location correspond to gaps in the NGIMS dataset (32). Ticks in the MY axis occur at the beginning of the year (LS = 0°). Dust storms correspond with heating in the middle atmosphere.

Figure 2 shows the variation of the abundance in the ionosphere (the charged portion of the atmosphere) of H2O+ and H3O+, normalized by the abundance of electrons, near MAVEN’s periapsis over the course of the mission. These two species are produced mainly by charge transfer between CO2+ and H2O, producing H2O+, and proton transfer from HCO+ to H2O, producing H3O+ (see below). Three forms of variations are visible in Fig. 2A. First, there is diurnal variation across two orders of magnitude. The solar zenith angle (the angle between the vertical and the Sun) in Fig. 2, A to C, corresponds to the time of day at which the measurement was made (0° is noon at the equator, 90° dawn or dusk, and 180° midnight). Second, H2O+ varies with season (LS), with repeatable maxima near LS = 270°, the peak of southern summer, and minima near LS = 40°. The relative abundance of H2O+ in the dayside ionosphere varies by more than an order of magnitude over a martian year, from tens of parts per million (ppm) to hundreds of ppm, peaking in southern summer near perihelion and declining during northern summer. The peak in ratio of H2O+ abundance to electron abundance ([H2O+]/[e]) in southern summer coincides with the peak in observed seasonal trends in middle-atmospheric H2O abundance (14, 15), the exospheric H and D abundances (9, 10, 25), and H escape rate (2628). Third, Fig. 2A shows large and rapid increases in the H2O+ abundance coincident with dust storms near LS = 315° in MY 32 and LS = 190° in MY 34. More detailed views of these events are shown in Fig. 2, B and C.

Fig. 2 Variation of H2O ions measured using NGIMS.

Mean [H2O+]/[e] for each MAVEN orbit (A) between October 2014 and November 2018, (B) during a regional dust storm in MY 32, and (C) during the MY 34 global dust storm. Dotted boxes in (A) indicate the regions shown in (B) and (C). Each data point is the mean [H2O+]/[e] (circles), [H3O+]/[e] [blue triangles in (B) and (C)], or [O2+]/[e] [green squares in (B) and (C)] between CO2 densities of 5 × 108 to 109 cm−3 (MAVEN’s nominal periapsis) for a single orbit. The value of the solar zenith angle is indicated by the color of the circle, as indicated in the colorbar. The transition from red to blue for [H2O+]/[e] corresponds to the shift from day to night.

The dust storms shown in Fig. 2, B and C, are substantial perturbations to the observed seasonal trend. At the onset in the upper atmosphere of the global dust storm of June 2018 (MY 34, Fig. 2C), the relative abundance of H2O+ in the ionosphere rises sharply in a short period of time: [H2O+]/[e] increases from 106 to 327 ppm, a factor of 3.1, over ~2 days, from 8 June 2018 to 10 June 2018. Simultaneously, [H3O+]/[e] increases from 11 to 28 ppm, a factor of 2.5. MAVEN’s periapsis precessed onto the night side of Mars shortly after the onset of the global dust storm, and measurements in this period are affected by the diurnal variation of the ionosphere (see below). Nevertheless, there is an abrupt increase in the relative abundance of these ions, which coincides with the first effects of the dust storm on the upper atmosphere on 8 June 2018, as measured by the Imaging Ultraviolet Spectrograph on MAVEN (29) and neutral atmosphere density measurements from NGIMS (30). The regional dust storm observed in March 2015 (MY 32, Fig. 2B) occurs during the declining seasonal trend, with an upward perturbation caused by the storm: Over 2 days, [H2O+]/[e] increases from 32 to 73 ppm, a factor of 2.3, and [H3O+]/[e] increases from 2.4 to 3.9 ppm, a factor of 1.6.

Large diurnal variations are also observed in the relative abundance of H2O+ as MAVEN’s periapsis repeatedly precessed from the day side to the night side, a process which occurs over several months, as seen in Fig. 2A. The relative abundance of H2O+ increases on the night side because, relative to other species in the martian atmosphere, H2O has a low ionization potential (IP), 12.6 eV, and OH has a high proton affinity (PA), 6.14 eV. In general, atmospheric ionization flows by electron transfer from neutral species with lower IPs to positive ions whose parent neutrals have higher IPs and, by proton transfer, from species whose nonprotonated forms have lower PAs to species that have higher PAs (31). On the night side, away from a constant source of photoionization driving the production of short-lived ions (such as O2+), H2O+ becomes more abundant. Thus, we can only observe the effects of dust storms on the day side; otherwise, the diurnal variation overwhelms the signal because of the storm.

Water in the upper atmosphere

Assuming that the increases in H2O+ and H3O+ are due to an influx of H2O into the upper atmosphere, we calculate H2O abundances from NGIMS density measurements of H2O+, H3O+, HCO+, and e obtained over 4254 MAVEN orbits. We assume photochemical equilibrium for H3O+; i.e., its production rate is equal to its loss rate. This assumption is known to be valid in the thermosphere on the day side but has not been substantiated on the night side (32). H3O+ is mainly produced by proton transfer to H2O and is only destroyed by dissociative recombination with e (3134). The results of these calculations are shown in Fig. 3. The diurnal variation of H2O is much smaller than that of H2O+, as expected in the neutral atmosphere, which is more chemically stable than the ionosphere.

Fig. 3 Variation of H2O and H2.

(A to C) Same as in Fig. 2 but for the H2O mixing ratio (circles) and H2 mixing ratio [blue triangles in (B) and (C)].

A repeatable seasonal trend occurs in the H2O abundance in the upper atmosphere, which peaks in southern summer between LS = 250° and 270°, shown in Figs. 3A and 4. This trend in upper-atmospheric H2O coincides with the seasonal trends seen in previous observations of the H2O abundance in the middle atmosphere (14, 15), the exospheric H and D abundances (9, 10, 25), and H escape flux (2628). The mean upper-atmospheric H2O mixing ratio is >1 ppm at all times and can exceed 10 ppm in southern summer. A sinusoidal model fitted to the dayside H2O abundances from all MYs, but excluding regional and global dust events, is shown in Fig. 4. This model has a maximum H2O mixing ratio of 4.3 ppm at LS = 259° and a minimum of 2.0 ppm at LS = 86°, for a variation of more than a factor of two. The H2O mixing ratio is never negligible, contrary to the assumption of 0 to 4 parts per billion made in numerous modeling studies (5, 33, 35).

Fig. 4 Seasonal variation of upper-atmospheric H2O.

Each data point is the mean dayside (solar zenith angle <90°) H2O mixing ratio between CO2 densities of 5 × 108 and 109 cm−3 for a single MAVEN orbit over MYs 32 (red circles), 33 (blue triangles), and 34 (green squares). The solid line is a sinusoidal model fitted to the data from all three MYs to illustrate the seasonal variation. Dust storm data (dotted boxes) are included in the plot but excluded from the model fitting.

During the global dust storm of June 2018, the mean H2O mixing ratio increased by a factor of 2.4, from a mean value of 3.0 to 7.1 ppm over 2 days. The water abundance then continued to increase following the seasonal and dust storm trends, reaching values >60 ppm at LS = 204°. This event included the highest persistent H2O abundances we observed, holding at tens of ppm for more than 5 months at the end of 2018. During the regional storm observed in MY 32, the H2O mixing ratio increases by 7%, from 4.6 to 4.9 ppm in 2 days. Thus, dust storms drive perturbations in upper-atmospheric H2O that are far more rapid than the diurnal variations, and even small dust storms have a measurable impact. The MY 32 dust storm occurred when MAVEN was near the subsolar point and is superimposed on a declining seasonal trend. We interpret these variations as being due to enhanced transport of H2O into the upper atmosphere during the dust storms.

Previous and coincident measurements of water

The effects of both regional and global dust storms on the H2O in the lower and middle atmosphere have been observed before. In MY 32, Mars Express observed large seasonal increases in the H2O mixing ratio, including an increase around LS = 315° (14), coinciding with the increase we observe in Figs. 2 and 3. These measurements indicate an increase in the hygropause altitude and an injection of water into the middle atmosphere, coinciding with our observed increases in the abundances of H2O, H2O+, and H3O+ in the upper atmosphere. Similarly, MCS detected an increase in the hygropause altitude, rising to 65 to 70 km, in this period (15). During the global dust storm of MY 34, Trace Gas Orbiter (TGO) measured a peak H2O abundance of ~100 ppm at 50-km altitude in the southern hemisphere (80°S to 83°S) and a peak H2O abundance of ~250 ppm just below 40-km altitude in the northern hemisphere (51°N to 59°N), indicating an injection of water into the middle atmosphere coincident with our observations (16). Both of these measurements were obtained at LS = 196.64°. Further TGO measurements showed that H2O abundances of ~5 to 10 ppm extended to altitudes up to 90 to 100 km (13, 17). The time scales of dust storm–induced variations in the upper-atmospheric H2O mixing ratio shown in Fig. 3 are consistent with these simultaneous measurements of the middle-atmospheric H2O mixing ratio during the 2018 global storm. MCS observed an H2O mixing ratio close to 300 ppm near 50-km altitude in the middle atmosphere (36). At the same time, we measure a mean H2O mixing ratio of 7.5 ppm near MAVEN’s periapsis (~150-km altitude) over an LS range of 191.05° to 193.21° and a latitude range of 13°S to 18°S. These measurements in the middle atmosphere, below the NGIMS sampling range, provide context for the NGIMS measurements. Seasonally and during dust storms, the H2O abundance in the middle atmosphere increased substantially in the span of a few martian days, coincident with NGIMS observations in the upper atmosphere.

Molecular hydrogen in the upper atmosphere

Molecular hydrogen, H2, may contribute to variations in protonated ions and is a source of H in the upper atmosphere. Measured H2 variations obtained by NGIMS are shown in Fig. 3, B and C. H2 reacts with numerous ions, mainly CO2+, to produce H atoms that may escape the planet’s atmosphere. H2 itself can escape, although much more slowly than H because H2 is twice the mass of H and lighter species have lower escape energies. The absolute sensitivity of NGIMS to H2 is unknown, so we cannot derive an absolute H2 abundance. However, relative variations of H2 are measurable. No seasonal trend or impact of dust storms are apparent in the upper-atmospheric H2 abundance (Fig. 3, B and C). H2 is therefore not the source of H responsible for seasonal and dust storm–induced variations in the upper atmosphere.

Water as a source of escaping hydrogen

We constructed a photochemical model to investigate the effects of the H2O abundances in the Mars ionosphere on the H escape rate. We constrain the model with data from a MAVEN Deep Dip (DD) campaign. During weeklong DD campaigns, the MAVEN spacecraft penetrates to lower altitudes, ~125 km. The measured altitude profiles provide constraints on ionospheric composition. We use the second DD campaign (DD2) executed around LS = 329° in MY 32, because MAVEN’s periapsis was then close to the subsolar point. We average the measurements over the DD using a previously described technique to remove wave structure in the altitude profiles (20). The solar zenith angles and latitudes do not vary appreciably during a DD, so the resulting profiles are effectively averages over longitude for a latitude of 4°S and local time of 12 p.m.

The photochemical calculations are one dimensional (1D) and use the DD2 geometry. Neutral and ionic composition is calculated from 80 to 250 km, including photolysis, chemical reactions, and diffusion of neutrals (32). We include both molecular and eddy diffusion, adopting an eddy diffusion coefficient of 108 cm2 s−1. Ions are assumed to be in local photochemical equilibrium. Rate coefficients are discussed below.

The model is constructed to match the primary neutral constituents measured by NGIMS, as shown in Fig. 5A. We calculate H2O by assuming a mixing ratio of 2 ppm at 80 km, consistent with our measurements obtained near aphelion (LS = 71°) in the absence of dust storms (Fig. 4). Photolysis causes the H2O mixing ratio to decrease slightly with altitude until ~160 km, where molecular diffusion begins to dominate and the mixing ratio starts to increase with height. For O, we assume a mixing ratio of 2% at 80 km, chosen to produce a vertical profile consistent with NGIMS measurements at higher altitudes.

Fig. 5 Calculated 1D model abundances and reaction rates.

(A) Neutral species abundances from our low-H2O photochemical model (lines) compared with the NGIMS DD2 mean data (crosses). (B) Same as in (A), but for ionic species. Chemical reaction rates for selected reactions in the (C) low- and (D) high-H2O models. The legend in (D) also applies to (C).

Figure 5B shows that the altitude profiles of H2O+ and H3O+ in the model are consistent with the observations, validating our earlier inferences. The model also matches measurements of CO2+, O2+, and HCO+, species involved in the chemical destruction of H2O. Our calculations also match the O+ observations. Models of the Mars ionosphere have often assumed that O+ was produced by CO2+ + O → O+ + CO2 with a rate coefficient of 10−10 cm3 s−1 (37); however, laboratory work has placed an upper limit of 6 × 10−13 cm3 s−1 on this rate coefficient (38, 39). Adopting this value, O+ is produced primarily by CO2 + hν → CO + O+ (hν, photon energy) and lost by O+ + CO2 → CO2+ + O and O+ + CO2 → O2+ + CO. Previous models assumed a combined rate coefficient for these reactions of 10−9 cm3 s−1 (40). However, state-specific measurements indicate a value of 4 × 10−10 cm3 s−1 (41). Our model uses a value of 3 × 10−10 cm3 s−1, chosen to match DD2 observations. Between 140 and 150 km, the predicted and observed protonated ion densities agree to within 25% (Fig. 5B). There are larger discrepancies at higher altitudes and near MAVEN’s periapsis, but the observed densities near periapsis are uncertain because the spacecraft is moving primarily horizontally (20) and the model densities are uncertain at higher altitudes because ion diffusion has been neglected in the model.

We use the model to calculate the net destruction rate of H2O and the production rate of H in the ionosphere (Table 1). The NGIMS measurements are only available above the ionospheric peak, the location of the maximum total ion density (Fig. 5B, 135 km), so we rely on model predictions to investigate chemical processes down to 80 km. In addition to the results for the DD2 model, we present rates for a similar model calculated for the same conditions and input data, but in which we adopt an H2O mixing ratio of 430 ppm at 80 km, consistent with our measurements obtained during the global dust storm of MY 34 (Fig. 4). H production in the ionosphere for the low-H2O model is 5 × 107 cm−2 s−1, at the low end of the range of H escape rates inferred by previous studies, 107 to 5 × 109 cm−2 s−1 (2628). The column-integrated H2O destruction in the ionosphere for the high-H2O model is 1.4 × 109 cm−2 s−1, corresponding to an H production rate of 2.8 × 109 cm−2 s−1, at the high end of the range determined in previous studies (2628). H production from destruction of H2 by CO2+ is 9.6 × 107 cm−2 s−1, less than twice as large as H production from ionospheric destruction of H2O in the low-H2O models, and substantially less than that in the high-H2O models.

Table 1 H2O budget and H production.

Two models with low and high H2O are listed, representing periods between and during dust storms, respectively (32). The H2O mixing ratios at 80 km for the low- and high-H2O cases are 2 and 430 ppm. Total H production from H2O is twice the difference between all H2O destruction and all H2O production. Total H production from H2 is twice the rate of destruction of H2 by CO2+.

View this table:

It is possible that ionospheric destruction of H2O is equally or more important than neutral photolysis of H2O for H escape. We cannot compare the upper- and lower-atmosphere contributions to escape in a more quantitative way because escape rates in these models are closely dependent on the boundary conditions on the H and H2 mixing ratios at 80 km, which are assumed values chosen to replicate observations at higher altitudes. Boundary conditions on mixing ratios can introduce spurious fluxes into photochemical models. Physical boundary conditions, such as deposition velocities at the surface, are required to accurately calculate fluxes. Our calculations and the high H2O densities in the thermosphere found in the MAVEN observations show that ionospheric chemistry is a source of H, which can escape from the Mars atmosphere.

Implications for climate evolution

Early studies of H production and escape neglected ionospheric destruction of H2O because H2O was assumed to be confined to low altitudes by a hygropause (5, 6, 42). Later, it was found that the hygropause varies in altitude with season, leading to speculation that H2O saturation may not occur at all during dust storms, owing to elevated temperatures (43). More recent 1D photochemical models (35, 44, 45) and 3D global circulation models (4648) have likewise not included chemical destruction of H2O in the ionosphere. Occultation and limb profile measurements in the middle atmosphere demonstrated that the hygropause does not confine H2O to low altitudes (13, 1517), and it has been postulated the elevated exospheric H densities during dust storms could be explained by these enhancements in the middle-atmospheric H2O abundance (11). However, this model neglected ionospheric chemistry, assuming the H was produced solely by neutral photolysis. We have shown that H2O is present in the ionosphere with substantial abundance throughout the martian year continuously since the arrival of MAVEN and that H is produced from this H2O by ion chemistry.

The time scales for destruction of H2O and production of H by ionospheric chemistry are short. Once in the ionosphere, H2O is destroyed primarily by reactions with CO2+ with a rate coefficient of 2.4 × 10−9 cm−3 s−1 (40). Adopting a CO2+ density of 3 × 104 cm−3 (Fig. 5B, 135 km) implies a lifetime for H2O of 4 hours. This is more than an order of magnitude shorter than the H2O photolysis time constant of 2 × 105 s (48). Assuming an electron density of 105 cm−3 (Fig. 5B, 135 km) and electron recombination rates of 4.4 × 10−7 cm−3 s−1 (49) for H2O+ and H3O+ yields a time constant of ~20 s for production of H from electron recombination. Thus, H2O transported to the ionosphere is quickly broken apart and converted to H. We expect the H produced at these high altitudes to have a high probability for escape. This implies that the limiting factor is the supply rate of H2O to the ionosphere from lower altitudes. This has implications for our view of H escape from Mars. In previous models, H escape was controlled by production of H2 in the lower atmosphere and its slow diffusion through the middle atmosphere. In our interpretation, H escape depends on how H2O moves past the hygropause and the rate at which it is transported from the middle atmosphere to the ionosphere.

The relative importance of neutral and ion chemistry in the production and escape of thermal and nonthermal H remains unclear. Because we observe that the H2 abundance is constant during the dust storms whereas the H2O abundance increases strongly, ionospheric destruction drives the elevated H escape rate during dust events. The importance of ion chemistry to H production has been discussed previously (34), although without direct information on the H2O abundance in the ionosphere. The H escape flux can be expressed asF = 1.6 × 108 + 1.4 × 1013 f cm−2 s−1(1)where f is the H2O mixing ratio at 80 km (34). The constant term of 1.6 × 108 cm−2 s−1 is the limiting flux for an H2 mixing ratio of 15 ppm, the value imposed at the lower boundary in those models (34). Adopting an H2O mixing ratio of 2 ppm in Eq. 1 produces a contribution to escape from ion chemistry of 2.8 × 107 cm−2 s−1, roughly half the column-integrated production rate in Table 1, suggesting an escape probability of 50% for the H produced in the ionosphere. This result is strongly dependent on the H mixing ratio at 80 km, which is an assumed value. Thus, in the previous model (34) and ours, the relative contributions of neutral and ion chemistry to escape depend on the boundary conditions for H and H2 at 80 km. The abundance of, exact mechanism of production for, and contribution to escape of a nonthermal component of the H corona at Mars have not been determined because of degeneracies in the models used (27, 28) and cannot be determined from our results.

NGIMS data are confined to altitudes above 125 km, whereas occultation and limb-sounding experiments have measured H2O up to ~100 km, leaving a large gap in coverage between 100 and 125 km. Figure 5A shows our predicted H2O mole fraction from the middle atmosphere through the ionosphere, but this is based on a 1D model, and the real atmosphere is likely to be more complex. The mixing ratios inferred for the middle atmosphere are often far greater than the values we calculate for the ionosphere (13, 16, 17). This contrast between the inhomogeneous H2O distribution derived from the middle-atmosphere measurements and the roughly constant altitude profile predicted by the 1D model (Fig. 5) suggests that the middle-atmosphere measurements are sensing a local enhancement of H2O that is transported horizontally around the planet as it diffuses upward. Such horizontal transport is not included in our 1D model. The largest peak in the lower-atmospheric H2O mixing ratio occurs in northern summer (2), the period in which we observe a minimum in the upper-atmospheric H2O mixing ratio (Fig. 4). The cause of this discrepancy between the seasonal variations in lower- and upper-atmosphere H2O abundances is unknown and cannot be explained by a 1D model.

The seasonal and dust storm–mediated delivery of water to the upper atmosphere could have played a substantial role in the evolution of the martian climate from its warm and wet state billions of years ago to the cold and dry planet we observe today. Mars has likely lost enough H2O to cover the planet’s surface with an ocean tens to hundreds of meters deep, and loss rates must have been higher in the past (4). In a martian year with no dust storms, assuming the H2O mixing ratio at 80 km is constant at 2 ppm, we calculate that 3.0 × 1015 cm−2 H would be produced from H2O and 1.1 × 1016 cm−2 H by destruction of H2 by CO2+, assuming all H, including that in HCO2+, is eventually liberated. During a global dust storm such as the one observed in MY 34, in which the H2O mixing ratio at 80 km reaches 430 ppm, 1.1 × 1016 cm−2 H would be produced from H2O in just 45 days. Thus, a single global dust storm can produce more than an entire martian year’s worth of H production and escape. If we extrapolate these numbers into the past, we find that over a billion years, a 44-cm-deep global layer of H2O would be lost owing to destruction of H2O in the ionosphere, and, assuming a rate of about one global dust storm every 10 years (50), an additional 17 cm of H2O would be lost owing to global dust storms. Larger loss estimates would be obtained if this extrapolation included the seasonal fluctuation of H2O in the upper atmosphere (Fig. 4) and regional dust storms, which, relative to global dust storms, likely have a smaller but non-negligible impact on H escape. These effects, though, are smaller than uncertainties related to our lack of knowledge of the martian climate billions of years ago (4). H escape driven by direct transport of H2O to the upper atmosphere would have accelerated the loss of water from Mars as the martian hygropause weakened owing to a decrease in atmospheric pressure, itself driven by atmospheric escape and the development of a dry, dusty surface.

Supplementary Materials

Materials and Methods

Fig. S1

Table S1

References (5263)

References and Notes

  1. Materials and methods are available as supplementary materials.
Acknowledgments: We acknowledge the contributions of the NASA MAVEN teams responsible for designing, constructing, and operating the spacecraft and its instruments, without which this work would be impossible. Funding: All authors were supported by funding from the MAVEN mission, part of the NASA Mars Exploration Program. Author contributions: S.W.S. and R.V.Y. conceived the study and wrote the draft manuscript. S.W.S., R.V.Y., M.B., M.K.E., and P.R.M. collected, calibrated, and analyzed the NGIMS data. R.V.Y. and D.Y.L. constructed the 1D photochemical model and analyzed the output, with the assistance of S.W.S. All authors contributed to the interpretation of results and preparation of the manuscript. Competing interests: We declare no competing interests. Data and materials availability: MAVEN data are available on the NASA Planetary Data System at and the MAVEN Science Data Center at We used NGIMS Level 2, version 8, revision 1 data (32), which can be searched at The 1D photochemical model code and output files are available at Zenodo (51).
View Abstract

Stay Connected to Science

Navigate This Article