Model-dependent deconvolution of HEND neutron data by MCNPX code

1
Model-dependent deconvolution of HEND neutron
data by MCNPX code

• M. Litvak, I. Mitrofanov, Charyshnikov, V.
  Grinkov, A. Kozyrev, A. Sanin
• V. Tretyakov
• Institute for Space Research
• D. Drake
• Santa Fe, USA

2
MODEL
Atmosphere
CO2 deposit
Dry regolith
Wet regolith
3
MODEL PARAMETERS

• Flux of CGR
• Atmosphere thickness at different times
• Chemical composition of atmosphere
• Thickness of CO2 frost
• Presence of water () in CO2 frost
• Thickness of dry regolith
• Chemical composition of regolith
• Presence of water () in dry regolith
• Presence of water () in wet regolith

4
MODEL PARAMETERS Martian Soil
Chemical analyses of Martian soil and rocks
obtained by the Pathfinder Alpha Proton X-ray
spectrometer Radiation Physics and Chemistry 61
(2001)191 197
5
MODEL PARAMETERS Martian Atmosphere Chemical
composition and thickness
Major gases Carbon Dioxide (CO2) - 95.32
Nitrogen (N2) - 2.7 Argon (Ar) - 1.6
Oxygen (O2) - 0.13 Carbon Monoxide (CO) -
0.08
http//nssdc.gsfc.nasa.gov/planetary/factsheet/mar
sfact.html
Thickness of atmosphere for given Mars region and
season is taken from the Ames global climate
model
6
HOW TO CALCULATE GEOMETRY The layer produced the
neutrons
7
HOW TO CALCULATE DATA SELECTION

• The Surface of Mars may be divided into the set
  of regions of interest.
• The properties of surface should be homogenous
  inside selected region
• The size of each region should be more than
  footprint.

8
HOW TO CALCULATE ALGORITHM
1. Calculation of neutron flux for model of
homogeneous sphere with some initial parameters
2. Convolution of model spectra with efficiency
functions for different detectors to get model
counts
3. Comparison of model count rate and real count
rate of HEND sensors for this Martian area of
interest and for the same seasons
4. Fitting of parameters of geometry, water
percentage in model of homogeneous sphere
hCO2
hdry
H20dry
H2Owet
9
HOW TO CALCULATE MINIMIZATION
Search of model best fit parameters based on ?2
minimization

• Where
• Ci is detector counts, i1,5 (SD,MD,LD,SC1,SC2).
  
• Mi is model counts predicted by MCNPX for
  (SD,MD,LD,SC1,SC2).
• ?i is statistical error for Ci, Mi
• SD, MD, LD are integral counts detected in
  proportional counters
• SC1 is fast neutron counts in energy range
  (0.8-2.5 MeV) registered in stilben
• SC2 is fast neutron counts in energy range gt2.5
  MeV registered in stilben

10
BEST FIT PARAMETERS CGR flux

• The selection of HEND data accumulated above
  large region near Solis Planum. It is driest
  place on Mars where variation of neutron flux
  explained only by variations of CGR flux or by
  Solar activity.
• The selection of different time intervals Ls
  330o-360o , 0o -30o ,30o -60o , 60o-90o,
  90o-120o , 120o-150o
• Applying of the homogenous model to estimate flux
  of CGR
• Atmosphere homogenous regolith layer with 2
  of water

11
BEST FIT PARAMETERS FROST FREE AREAS

• The selection of data for summer surface of Mars
  (season when CO2 frost is absent)
• First stepApplying of the homogeneous regolith
  model with one free
• parameter describing content of water in the
  regolith.
• Second step If homogeneous model was rejected
  two layers regolith model should be tested with
  2 of water in upper layer and two free
  parameters thickness of upper layer and water
  content in
• bottom layer.

12
BEST FIT PARAMETERS CO2 FROST THICKNESS

• The best fitting parameters describing summer
  surface are fixed for given region.
• The data for given region are taken for different
  winter periods of time.
• The real data are fitted by model of summer
  surface with added layer of CO2 frost with
  variable thickness.