The Mars Climate Database E. Millour, F. Forget, A. Spiga, M. Vals, V. Zakharov, Laboratoire de Météorologie Dynamique, IPSL, France, L. Montabone, Space Science Institute, Boulder, USA,

F. Lefèvre, F. Montmessin, J-Y Chaufray, Laboratoire Atmosphères, Milieux, Observations Spatiales, IPSL, France, M.A. López-Valverde, F. González-Galindo,

Instituto de Astrofísica de Andalucía, Spain, S.R. Lewis, Department of Physics and Astronomy, The Open University, UK, P.L. Read, Atmospheric, Oceanic &

Planetary Physics, University of Oxford, UK, M.-C. Desjean, CNES, Toulouse, France, F. Cipriani, ESA/ESTEC, Noordwijk, The Netherlands,

and the MCD/GCM development team

What is the Mars Climate Database? • The Mars Climate Database (MCD) is a database of meteorological fields derived from General Circulation

Model (GCM) numerical simulations of the Martian atmosphere and validated using available observational

data. The database does not only include the outputs of the GCM, but also provides complementary post-

processing schemes such as high spatial resolution interpolation of environmental data and means of

reconstructing the variability thereof. Our GCM is developed at LMD (Paris, France) in collaboration with

LATMOS (Paris, France), the Open University (UK), the Oxford University (UK) and the Instituto de Astrofisica

de Andalucia (Granada, Spain) with support from the European Space Agency (ESA) and the Centre National

d’Etudes Spatiales (CNES, France).

• The MCD is freely distributed and intended to be useful for engineering and scientific applications

which require accurate knowledge of the state of the Martian atmosphere. For moderate needs, the MCD may

be accessed online via an interactive server (head for http://www-mars.lmd.jussieu.fr). To obtain a copy of the

full version which includes all the datafiles and the advanced access software, contact us:

ehouarn.millour@lmd.jussieu.fr and/or francois.forget@lmd.jussieu.fr.

• MCD version 5.3 was released in July 2017. We are implementing MCD version 6.0 which should be

available in autumn 2019.

Atmospheric variability in the MCD To represent the wide ranges of scales over which meteorological variables vary, the MCD includes:

Year to year variability and dust content variations : Simulation of years with three different solar Extreme UltraViolet (EUV) inputs on top of the different dust scenarios

Seasonal cycle : In the MCD are stored 12 “typical” days (average over 30° of Ls) around the year. Values at a given date are obtained by interpolation.

Diurnal cycle : Environmental data are stored 12 times per day; interpolation is used to evaluate values of variables at a given time of day.

Day to day variability (i.e. representation of transient waves): Various tools are provided to reconstruct the variability of meteorological variables:

Perturbations may be added as:

• Large scale perturbations, using Empirical Orthogonal Functions (EOFs) derived from the GCM runs.

• Small scale perturbations, by adding a gravity wave of user-defined wavelength.

Standard deviations of main meteorological variables are given for:

• Surface temperature, surface pressure, dust opacity.

• Atmospheric density, pressure, temperature and winds.

These RMS day to day variabilities are given both pressure-wise and altitude-wise.

The high resolution mode • The GCM horizontal longitude×latitude computational grid is 5.675°×3.75°.

• By combining high resolution (32 pixels/degree) MOLA topography and Viking Lander 1 pressure records (used as a reference to correct the atmospheric mass) with GCM surface pressure, a corrected high resolution surface pressure may be derived (as is done in our standalone “pres0” utility).

• The high resolution surface pressure can be used to reconstruct the vertical atmospheric pressure distribution and, within the restriction of the procedure, yield high resolution values of atmospheric variables.

Validation of the Mars Climate Database climatology The MCD has been validated using observational data from many available sources: Mars Global Surveyor (TES, Radio

Science, accelerometer), Mars Express (SPICAM, PFS, OMEGA, MaRS), Mars Reconnaissance Orbiter (MRO), Viking

Landers, Pathfinder, MERs Spirit and Opportunity, Curiosity. As new measurements are made available, these are

compared to the MCD predictions. We are working on a validation document which will report on all these comparisons.

Comparison between Opportunity entry (Mars Year

26, solar longitude Ls=339) profile, retrieved by

Paul Withers and mean MCD profiles obtained for

various dust scenarios.

Illustrative examples:

clim scenario (for Opportunity entry) temperature

profiles obtained with the three possible Solar EUV

inputs (minimum, average, maximum, thick lines),

alongside with profiles for MY24 to MY33.

Overview of MCD contents • The MCD includes 4 synthetic dust scenarios in order to better represent the range of variability of the

Martian atmosphere due to the amount and distribution of suspended dust: a climatology case (derived

from 8 Mars years of observations, Montabone et al. 2015), a cold (clear sky, 50% less dust than

observed), a warm (dusty conditions, 50% more than observed) and a global dust storm case (high

tau=5 opacity everywhere and using “darker” dust optical properties from Ockert-Bell et al. 1997, rather

than the more recent Wolff et al. 2009 ones).

• The MCD extends into the thermosphere, up to ~350 km (and more), since the GCM it is derived from

includes a thermosphere model above ~100 km. Three Extreme Ultra Violet (EUV) scenarios, which

account for various states of the solar cycle (minimum, average and maximum), are thus explored.

• Far more than the main meteorological variables (density, temperature and winds) are provided, as the

GCM includes water cycle (Madeleine et al. 2012, Navarro et al. 2014), chemistry (Lefèvre et al. 2011)

and ionosphere (Gonzalez-Galindo et al. 2013, Chaufray et al. 2014) models.

Mean climatological variables provided in the MCD:

(stored at 12 local times of a typical day for each of 12 months and reconstructed by interpolation)

• Atmospheric density, pressure, temperature and winds (horizontal and vertical)

• Surface pressure and temperature, CO2 ice cover, Seasonal H2O frost cover,

• Thermal and solar radiative fluxes,

• Dust column opacity and dust mass mixing ratio, estimation of dust effective radius and deposition rate,

• [H2O] vapor and [H2O] ice columns and mixing ratios, ice particle effective radius,

• [CO2], [CO], [O], [O2], [O3], [N2], [Ar], [H], [H2], [He] and [e-] mixing ratios and columns,

• Air specific heat capacity, viscosity and reduced molecular gas constant R,

• Planetary Boundary Layer (PBL) height, vertical convective winds in PBL,

• Surface wind stress and sensible heat flux.

• Additional scenarios, representative of

the last 10 Mars Years (MY24 to

MY33) both in terms of observed

atmospheric dust content (Montabone

et al. 2015; daily maps downloadable

from the MCD website) and EUV input

(daily varying) are also included in

order to provide our best estimate of

these years. The EUV forcing for these

scenarios is that of the observed solar

cycle.

• For intensive and precise work: You will need the full MCD, which contains the data files (in NetCDF format)

and access software (which does all the post-processing to include and account for sub-grid scales, day-to-day

variations of the Martian atmosphere, etc…) as well as the lighter standalone high resolution surface pressure

predictor “pres0”.

The software is written in Fortran; works on Linux and can be ported to Windows. Examples of Python, IDL,

Matlab, Scilab, C and C++ interfaces to the MCD are also provided.

Using the Mars Climate Database

• For moderate needs: head for the online version of

the MCD: http://www-mars.lmd.jussieu.fr which gives

access to :

• All scenarios and variables, a choice between 3

different vertical coordinates (pressure levels, altitude

above areoid or above surface), wind vector overlay,

zonal averages, various map projections, …

• An Earth date to Mars date (value of solar

longitude Ls) converter, as well as a reverse Mars

date to Earth date converter.

Example n°1: Surface pressure at Viking Lander 2 site

• Surface pressure cycle over a Martian year, as

predicted by the MCD climatology scenario at Viking

Lander 2 site, with an envelope of sqrt(2) times its

standard deviation, compared to the recorded values.

• Switching from the baseline climatology scenario to

the Dust Storm scenario enables to recover the change

in behavior recorded by Viking Lander 2 during the 1977

global dust storm.

The Viking landers reached Mars in Mars Year 12 and collected data for a few Mars years (3 for VL1 and 1.5 for VL2).

Example n°2: MSL Curiosity surface pressure

The REMS station onboard Curiosity (137.4°E and 4.6°S) measures surface pressure since the rover's arrival on Mars

(Mars Year 31, solar longitude Ls=151°) and is currently still collecting data.

• Measured surface pressure at MSL site compared to

the MCD predicted climatology scenario over the first

Mars year of measurements. See also Ordonez-

Etxeberria et al. 2019 for in-depth comparisons.

• Surface pressure at MSL site compared to MCD

predictions using the climatology scenario with added

large scale perturbations.

Towards MCD version 6.0 • Many developments have been made in our GCM physics package over the last years:

- The water cycle has been improved by introducing a representation of sub-grid scale clouds (when only a fraction of the mesh is covered by clouds, Pottier et al. 2017), and by increasing the vertical resolution, which better represents the intense convection in some night-time clouds (Spiga et al. 2017).

- An improved atmospheric chemistry solver (Cariolle et al. 2017) has been implemented.

- A parametrization of the impact of non-orographic gravity waves, which seem to be crucial components of the atmospheric circulation, has been developed and is found to significantly improves the simulated thermal structure (Gilli et al. submitted).

- A parameterization of « rocket dust storms » (Spiga et al. 2013) generating detached dust layers observed between 20 and 30 km altitude has been implemented (Wang et al. 2018). But these only arise during the dust storm season. To also obtain them during the clear season requires a parametrization of dust injection by winds along mountain slopes (Vals et al. this conference).

• The updated GCM will be used to generate MCD version 6.0, which we plan to released in October 2019.

Example n°3: TES atmospheric temperatures

The Thermal Emission Spectrometer (TES) onboard Mars Global Surveyor has nearly continuously monitored the

Martian Atmosphere for almost 3 Martian years (from early in Mars Year 24 to the beginning of Mars Year 27), yielding

information on the local and seasonal evolution of atmospheric conditions on Mars.

Top left & middle plots: Distributions of binned (using 1K bins) temperature

differences (at 106 Pa pressure level) between MCD climatology predictions

and TES (2pm or 2am) measurements over Mars Years 26 and 27 (up to Ls=85°,

i.e. end of TES measurements) and for latitudes ranging from 50°S to 50°N.

Displayed MEAN and RMS values are computed from the obtained histograms

and the curves correspond to normal distributions of same MEAN and RMS.

Top right plot: Same distributions evaluated using the clim, cold and warm

MCD scenarios depicting how these bracket observations.

Bottom right plot : Similar evaluation, comparing warm, MY25 and dust storm

scenarios to the Mars Year 25 global dust storm (from Ls=190° to Ls=230°)

measurements.

The Mars Climate Sounder (MCS) on-board Mars Reconnaissance Orbiter has been monitoring the Martian Atmosphere

since Mars Year 28 and is still collecting data, more than 7 Martian years later.

• Left & middle plots: Distributions of binned (using 1K bins) temperature differences (at 106 Pa pressure level) between

MCD climatology predictions and MCS (3pm or 3am) measurements over Mars Years 30 and 31 for latitudes ranging

from 50°S to 50°N. Displayed MEAN and RMS values are computed from the obtained histograms and the curves

correspond to normal distributions of same MEAN and RMS.

• Right plot: Same distributions evaluated this time using different MCD scenarios (cold, climatology and warm).

Example n°4: MCS atmospheric temperatures