WINTER POLAR WARMING IN THE MARS THERMOSPHERE.

S. W. Bougher, J. M. Bell, B. Steers, AOSS Department, University of Michigan, USA, (bougher@umich.edu),J. R. Murphy, Department of Astronomy, New Mexico State University, USA (murphy@nmsu.edu), G. M. Keating,George Washington University at NASA Langley Research Center, USA.

Introduction:

Mars winter polar warming is a phenomenon of thelower thermosphere temperatures and densities that iswell documented by in-situ accelerometer data takenduring three previous spacecraft aerobraking campaigns[1, 2, 3, 4.] The latitudinal variation in thermosphericdensities and derived temperatures (at a constant alti-tude) is measured as the various spacecraft pass overthe local winter pole. Temperatures are expected to becolder in the high latitude winter polar regions, as ra-diative equilibrium is maintained. However, significantdepartures from radiative equilibrium temperatures wereobserved during the perihelion season.

The Mars 2001 Odyssey (MO) accelerometer sam-pled high northern latitudes during aerobraking (Ls =265 - 310) as periapsis passed over the northern winterpole. The MO aerobraking data in the northern winternight revealed a strong polar warming over 100-130 km[1]. Temperatures near periapsis (100-110 km) were dis-covered to increase with latitude (60-90N) from about100 to 170-200 K, maximizing near the North pole onthe nightside. Corresponding temperatures near 120-130 km were observed to increase from 110 K to 160-170 K as the North pole was approached. These MO ac-celerometer observations provide the first clear evidenceof strong winter polar warming at Mars thermosphericheights during the perihelion season [1]. However, therewas no concurrent lower atmosphere polar warming ob-served during this MO aerobraking period [5].

Conversely, Mars Global Surveyor (MGS) and MarsReconnaissance Orbiter (MRO) accelerometer data weretaken as the southern polar region was sampled duringthe aphelion season [1, 3, 4]. MGS Phase 2 aerobrak-ing (Ls = 30-95) and MRO aerobraking (Ls = 35-109)periods span similar latitudes and seasons as periapsispassed over the South winter pole. Densities were re-trieved and temperatures derived for two different Mar-tian years, revealing no analogous warming of tempera-tures within the south polar thermosphere during theseaphelion/winter seasons.

This dramatic seasonal variation in winter polar tem-peratures and the repeatable lack of winter polar warm-ing during the aphelion season (over two Martian years),suggests a dynamical explanation that is independent ofradiative equilibrium considerations. The winter po-lar warming in the Mars thermosphere may occur as aresult of adiabatic heating from the subsiding branchof a strong inter-hemispheric circulation cell that op-erates during the solstices. This circulation pattern

should be enhanced during the perihelion season dueto stronger solar insolation (heating the entire atmo-sphere) and the remote effects of significant (and highlyvariable) aerosol heating in the lower atmosphere [1,2]. Conversely, the aphelion season inter-hemisphericcirculation should be weakened and possibly quite con-sistent (repeatable dust distributions) from one Martianyear to the next, resulting in a weak (or non-existent)warming in the southern winter polar region.

Such a dynamical explanation for Mars is not with-out precedent. Lower atmosphere (25-80 km) north po-lar winter warming has previously been observed duringseveral events [6, 7, 8, 9]. Winter polar warming in thelower atmosphere for specific events was realisticallysimulated using revised Mars lower atmosphere modelsthat were extended to ∼90 km or more [10, 11]. Suchan extension enabled these models to capture the diabat-ically driven inter-hemispheric (Hadley) circulation thatis strong and deep on Mars, spanning the region fromthe ground to nearly 100 km. Strong meridional windsabove 50 km lead to a horizontal convergence of massand downward motion above the polar region, resultingin adiabatic warming down to ∼25 km. This warmingis shown to be most pronounced during perihelion con-ditions (Ls = 270), when maximum solar insolation andstrong dust heating prevail.

Detailed global model simulations are likewise re-quired to capture the energetics and dynamics of thecombined lower and upper atmospheres in order to ad-dress this issue of Mars thermospheric winter polarwarming. The processes driving these seasonal and in-terannual variations in the Martian thermosphere can beaddressed through careful interpretation of available ac-celerometer datasets using several global thermosphericmodel simulations. Coupled yet separate Mars Gen-eral Circulation Model (MGCM)-Mars ThermosphericGeneral Circulation Model (MTGCM) simulations ofthe lower and upper atmospheres of Mars are now usedto address these features and the underlying dynamicalprocesses linking these Mars atmospheric regions.

Our strategy is to focus on changing lower atmo-sphere conditions from one Martian year to the next,while holding topside forcing (e.g. solar EUV fluxes)constant. For this presentation, we have selected Mar-tian solstice simulations, which are driven by solar andaerosol heating corresponding to aphelion (Ls = 90) andperihelion (Ls = 270) conditions appropriate to existingMGS, MO and MRO observing periods (above). NewMGCM-MTGCM simulations are conducted using threeavailable mapping years of infrared (IR) optical depth

Seventh International Conference on Mars 3027.pdf

HEIGHT (KM)

LATITUDE (DEG)

100

110

120

130

140

150

160

170

180

190

200

210

220

0-20-40-60-80 20 40 60 80

HEIGHT (KM)

LATITUDE (DEG)

100

110

120

130

140

150

160

170

180

190

200

210

220

0-20-40-60-80 20 40 60 80

HEIGHT (KM)

LATITUDE (DEG)

100

110

120

130

140

150

160

170

180

190

200

210

220

0-20-40-60-80 20 40 60 80

HEIGHT (KM)

LATITUDE (DEG)

100

110

120

130

140

150

160

170

180

190

200

210

220

0-20-40-60-80 20 40 60 80

HEIGHT (KM)

LATITUDE (DEG)

100

110

120

130

140

150

160

170

180

190

200

210

220

0-20-40-60-80 20 40 60 80

HEIGHT (KM)

LATITUDE (DEG)

100

110

120

130

140

150

160

170

180

190

200

210

220

0-20-40-60-80 20 40 60 80

(a)

(d)

(b) (c)

(e) (f)

Figure 1: Zonal average neutral temperatures, for both solstice seasons at each of the three TES Mapping Years as a function ofboth altitude (vertical axis) and latitude (horizontal axis). The solar flux is held constant at F10.7-cm = 130 for all simulations. Theblack horizontal line in all plots represents 120 km altitude level. Neutral temperatures (in K) for the northern summer solstice,LS = 090 are depicted for the various lower atmospheric dust conditions : (a) TES Year 1, (b) TES Year 2, and (c) TES Year 3.Similarly, neutral temperatures (in K) for the opposite season of southern summer solstice, LS = 270, are depicted for the variouslower atmospheric dust TES Maps: (d) TES Year 1, (e) TES Year 2, and (f) TES Year 3.

data acquired by the Thermal Emission Spectrometer(TES) instrument on board MGS [5]. Interannual varia-tions in these dust opacity distributions of the Mars loweratmosphere are utilized to investigate the dust’s subse-quent impacts (both dynamical and thermal) upon theMars upper atmosphere structure and dynamics. Ourgoal is to isolate the impact of changing lower atmo-sphere conditions (e.g. dust distributions) upon the sea-sonal/interannual variations of key features in the Mar-tian thermosphere.

Subsequently, specific calculations appropriate tothe exact seasonal and solar cycle conditions of MGS,MO, and MRO aerobraking observing periods (near thesolstices) are presented. Simulated polar warming fea-tures are compared with the observed temperature anddensity trends. Implications for the seasonal and inter-annual variability of the Mars winter polar warmingphenomenon will be described; consequences for aer-obraking planning exercises will be discussed.

MGCM-MTGCM Model Formulation:

The Mars Thermospheric General Circulation Model(MTGCM) itself is a finite difference primitive equationmodel that self-consistently solves for time-dependentneutral temperatures, neutral-ion densities, and threecomponent neutral winds over the globe [2, 12]. MT-GCM prognostic and diagnostic thermospheric fields aresimulated on 33-pressure levels (at altitudes above 1.32µbar corresponding to ∼60-300 km), with a 5x5 ◦ lati-tude and longitude resolution. Recently, a fast non-LocalThermodynamic Equilibrium (NLTE) 15-micron cool-ing scheme was implemented in the MTGCM, alongwith corresponding near-IR heating rates. These im-provements are based upon recent detailed one dimen-sional (1-D) NLTE model calculations for the Mars up-per atmosphere [13].

The MTGCM is currently driven from below by theNASA Ames Mars General Circulation Model (MGCM)code [14] at the 1.32 µbar level (near 60-80 km), permit-ting a detailed coupling across this boundary. This cou-pling scheme captures both migrating and non-migrating

Seventh International Conference on Mars 3027.pdf

HEIGHT(KM)

LATITUDE (DEG)

100

110

120

130

140

150

160

170

180

190

200

210

220

0-20-40-60-80 20 40 60 80

HE

IGH

T (

KM

)

LATITUDE (DEG)

100

110

120

130

140

150

160

170

180

190

200

210

220

0-20-40-60-80 20 40 60 80

HEIGHT(KM)

LATITUDE (DEG)

100

110

120

130

140

150

160

170

180

190

200

210

220

0-20-40-60-80 20 40 60 80

HEIGHT(KM)

LATITUDE (DEG)

100

110

120

130

140

150

160

170

180

190

200

210

220

0-20-40-60-80 20 40 60 80

HEIGHT(KM)

LATITUDE (DEG)

100

110

120

130

140

150

160

170

180

190

200

210

220

0-20-40-60-80 20 40 60 80

HEIGHT(KM)

LATITUDE (DEG)

100

110

120

130

140

150

160

170

180

190

200

210

220

0-20-40-60-80 20 40 60 80

(a)

(d)

(b) (c)

(e) (f)

Figure 2: Zonal average adiabatic heating and cooling in (colored contours) and zonal mean stream functions (black lines) forboth solstice seasons at each of the three TES Mapping Years as a function of both altitude (vertical axis) and latitude (horizontalaxis). The solar flux is held constant at F10.7-cm = 130 for all simulations. The black horizontal line in all plots represents 120 kmaltitude level. Adiabatic heating and cooling rates (in K/day) and zonal stream function (black contours with directional arrows) forthe northern summer solstice, LS = 090 are depicted for the various lower atmospheric dust conditions: (a) TES Year 1 (maximummeridional wind at 120 km of -100 m/s), (b) TES Year 2 (maximum meridional wind at 120 km of -90 m/s), and (c) TES Year 3(maximum meridional wind at 120 km of -120 m/s). Similarly, adiabatic heating/cooling (in K/day) and zonal mean winds (in m/s)for the opposite season of southern summer solstice, LS = 270, are depicted for the various lower atmospheric dust TES Maps: (d)TES Year 1 (maximum meridional wind at 120 km of 130 m/s), (e) TES Year 2 (maximum meridional wind at 120 km of 150 m/s),and (f) TES Year 3 (maximum meridional wind at 120 km of 120 m/s)..

upward propagating tides plus the thermal expansion andcontraction of the Mars lower atmosphere with the pas-sage of the seasons and dust storm events [2, 12]. Keyprognostic (temperatures, zonal & meridion‘al winds)and diagnostic (geopotential height) fields are passedupward from the MGCM to the MTGCM at the 1.32-µbar pressure surface at every MTGCM gridpoint on2-minute timestep intervals. No downward couplingfrom the MTGCM to the MGCM is presently activated.The inclusion of the Ames MGCM in providing a realis-tic lower atmosphere is critical for achieving a realisticsimulation of the Mars upper atmosphere within the MT-GCM domain [2].

Brief Model Results Illustrating Variability:

Results from the MGCM-MTGCM simulations com-paring TES years #1, 2, and 3 dust forcing for aphelion

(Ls = 90) and perihelion (Ls = 270) conditions are pre-sented in Figures 1 and 2 [15]. Simulated temperaturesin Figure 1 reveal that the polar lower thermosphere ex-hibits a high degree of variability during perihelion, withreduced yet significant variability near aphelion. Thisbehavior is consistent with that observed in the loweratmosphere, as documented by [16]. The underlyingadiabatic cooling rates (summer hemisphere upwelling)and heating rates (winter hemisphere subsidence) thataccompany the interhemispheric circulation shown inFigure 2 indicate that variations in winter polar tem-peratures are tightly coupled to significant variations inadiabatic heating.

Finally, specific simulations for MGS, MO, MROaerobraking periods (near the solstices) are conducted.Latitudinal temperature variations near 130 km are il-lustrated in Figure 3. Notice that simulated MO north-ern winter polar temperatures (Ls = 270) approach 190-

Seventh International Conference on Mars 3027.pdf

Figure 3: MGS, MO, and MRO simulated latitudinal temperature structures at 130 km from several MTGCM calculations. MGS(blue), MO (red), and MRO (orange) curves are illustrated to show temperature variations and their seasonal and inter-annual trendsfor comparison to available aerobraking datasets.

200K, while southern winter polar temperatures drop toas low as 130K. These are not zonal average values,but rather specific to the local solar time (and averagedlongitudes) of the accelerometer measurements. Thisseasonal trend is generally consistent with observations[1, 2, 3, 4]. However, simulated MO winter polar tem-peratures are slightly warmer than observed.

Conclusions:

It is demonstrated from this current research and previ-ous studies that the entire Mars atmosphere is an inte-grated system that is highly coupled dynamically, ther-mally, and chemically. The proper characterization ofthe processes giving rise to observed winter polar warm-ing in the Mars thermosphere is possible using such anintegrated model treatment. Future aerobraking exer-cises will benefit from the ability to better predict thewinter polar densities that accompany these winter po-lar temperatures.

References

[1] Keating, G. M. et al., (2003), Proceedings - Inter-national workshop: Mars atmosphere modeling and ob-servations, Abst. 17-3, Granada, Spain. [2] Bougher, S.W., J. R. Murphy, and J. M. Bell (2006), Geophys. Res.Lett., 33, L02203. [3] Keating, G. M. and S.W. Bougher(2006a), Proceedings of 38th DPS Meeting, B.A.A.S.,3, 631. [4] Keating, G. M., et al., (2006b), Proceedingsof Fall AGU Meeting, P33A-06, 234. [5] Smith, M. D.,(2004), Icarus, 167, 148-165. [6] Deming, D., et al.,(1986), Icarus, 66, 366-379. [7] Jakosky, B. M., and T.Z. Martin, (1987), Icarus, 72, 528-534. [8] Theodore,B., et al., (1993), Icarus, 105, 512-528. [9] Smith, M. D.,J. C. Pearl, B. J. Conrath, and P. R. Christensen, (2001),JGR, 106, 23,929-23,945. [10] Wilson, R. J., (1997),Geophys. Res. Lett., 24, 123-126. [11] Forget, F. etal., (1999), JGR, 104, 24,155-24,175. [12] Bougher, S.W., S. Engel, D. P. Hinson, and J. R. Murphy (2004),JGR, 109, E03010. [13] Lopez-Valverde, M. A. et al.,(1998), JGR, 103, 16799-16811. [14] Haberle, R. M.,et al., (1999), JGR, 104, 8957-8974. [15] Bell, J. M., S.W. Bougher, and J. R. Murphy (2007), JGR, submitted.[16] Liu, J., M. I. Richardson, and R. J. Wilson (2003),JGR, 108, E8, 5089.

Seventh International Conference on Mars 3027.pdf