Title: Trapped Proton Models: TPM1 and Beyond
1Trapped Proton ModelsTPM-1 and Beyond
- Stuart L. Huston
- Science Applications International Corporation
- Working Group Meeting on New Standard Radiation
Belt and Space Plasma Models for Spacecraft
Engineering - College Park, MD
- 2004 October 58
Work supported by NASA Space Environments and
Effects Program (SEE) and Living With a Star
Targeted Research and Technology Program
(LWS-TRT), and by Science Applications
International Corporation, McDonnell Douglas, and
Boeing
2Outline
- NOAAPRO/LATRM
- TPM-1
- Recent Work
- Implications/Needs/etc.
3The Basic Question
- Q Why spend ? to get data and develop a
radiation environment model? - A
- Uncertainty in the environment can increase total
life cycle costs by many G - Trading system lifetime, reliability,
replenishment, performance, mass, numbers of s/c,
etc. - Risk management/mitigation
- Overdesign vs. underdesign
- Uncertainty in the environment may limit or
prohibit certain technologies and/or missions - Without good environment models, effects models
are useless of limited value
xxxxxxx
4NOAAPRO/LATRM
- Objective
- Develop solar-cycle dependent model of low
altitude (lt1000 km) protons - Use NOAA/POES MEPED data (gt16, gt 36, gt 80 MeV)
- Implementation
- Empirical curve fit for solar cycle variation
- Input is latitude/longitude/altitude/date
- Model calculates magnetic coordinates
- Output is simple integral flux
5Proton Flux is Controlled by Atmospheric
Density/Solar Activity
6Introduction of Phase Lag Gives Good Correlation
with Data
Simple exponential relationship between proton
flux and phase-lagged solar flux Note that there
is still factor of 2 uncertainty in flux
prediction Even larger errors during transient
events
7Model/Data Comparison Magnetic Equator
8TPM-1
- Problems w/NOAAPRO
- Poor energy resolution
- Limited spatial coverage
- Objective
- Combine CRRESPRO and NOAAPRO
- Model valid from low-altitude to near GEO
- Challenges
- Poor spatial resolution/biasing in CRRESPRO at
low altitudes - Intercalibration of very different detectors
- Combining integral flux data (POES) with
differential data (CRRES) - Approach
- Convert CRRESPRO from bins to grids
- Scale spectra based on NOAAPRO
- Use NOAAPRO solar cycle variation
9CRRESPRO 16.9 MeV Quiet
Flux is defined in bins of L/l assumed constant
within bin Resolution is too coarse to capture
gradients near atmospheric cutoff Fluxes are
biased towards higher values due to gradients
within bins
10TPM-0 16.9 MeV Quiet
TPM-0 was an interim model based on
CRRESPRO CRRESPRO fluxes converted to grids,
allowing interpolation within grid Fluxes
normalized so that average flux within bin equals
CRRESPRO flux
11Calibration of POES Detectors
SEM-2 (NOAA-15 later) re-designed to reduce
contamination
12Electron Contamination(calculations by T.
Cayton, LANL)
13Comparison Low Altitude Models
14TPM-1 Spectra (High-L)
15Flux vs. AltitudeTPM-1 Quiet Model
16Flux vs. AltitudeTPM-1 Quiet Model vs. AP-8
- NOTE population is variable, TPM-1 represents a
snapshot at solar maximum
17TPM-1 Summary As Delivered
- Energy Range 1 81.5 MeV
- Regional coverage
- 300 - 36,000 km
- 1.2 lt L lt 5.5
- 0 lt l lt 50
- Inputs
- Latitude, Longitude, Altitude (internal field
model calculates magnetic coordinates - Can perform orbital integration
- Date (i.e., phase of solar cycle)
- F10.7 history
- Time coverage
- Data cover 1978 1995 (with special attention to
1990-1991) - Valid for any point in solar cycle
- Time scale (vs. resolution) 1 month
18TPM-1 Summary (concluded)
- Calibration/Intercalibration
- Detector calibration issues as discussed
- Very limited validation/characterization
performed - Comparison with other models
- Absolute magnitude within about a factor of 5 of
AP-8 (depending on energy/region) - Spectra generally harder than AP-8
- Radial profiles show that flux peaks at lower
altitudes than AP-8 - Low altitude fluxes are lower than CRRESPRO,
higher than AP-8 - Additional comments
- Another solar cycle of POES data for
incorporation into model - High altitude portion (CRRESPRO) based on a
snapshot at solar max - Quiet and active models what do we do about
active periods?
19New Directions
20Recent Work (Sponsored by SEE LWS)
- Statistical solar cycle variation (Xapsos, 2002)
- Slot Dynamics Study (2003)
- Investigated statistics of slot region
- Long-Term Dynamics (2004 2005)
- Extend energy range by combining w/SAMPEX data
(collaboration with BIRA) - Extend high energy/low altitude data to equator
with analytical model Salammbô (collaboration
with ONERA) - Develop statistical model for low-altitude solar
variation (collaboration with GSFC)
21Statistical Solar Cycle Variation
Standard TPM-1 requires knowledge of solar flux
and its history at the time of interest Xapsos
combined historical solar cycle data to determine
flux history as a function of confidence level
22Statistical Solar Cycle Variation
23Transient Belts Problem
TPM-1 contains quiet and active models No
guidance as to how often the belts are
active Under LWS, we looked at these belts
using POES long-term data
24NOAA-06/08 1986
(a) Survey plot showing color-coded intensity as
a function of time and L-shell.
(c) Probability of exceeding a given flux as a
function of L-shell.
(d) Time period from which data are taken in
relation to solar cycle.
(b) Flux vs. time at several discrete L-shells.
25Variability vs. L
Proton fluxes at L 2.5 are quite dynamic Flux
is enhanced 25 of the time 90 confidence
level is 10X higher than mean Enhancements of
up to 150X, lasting up to 1.5 years March 1991
event was unique in its inward extent (as well as
magnitude) Electron contamination use SEM-2
data from NOAA-15, -16, -17
26TPM-1 Summary 2
- Attempt to provide successor to AP-8, but limited
energy range limits usefulness - Can be used to assess uncertainty in AP-8
- Very limited validation performed
- No provision for maintenance/upgrades
- Work done since initial release has not been
incorporated into TPM-1
27Modeling/Software Issues
- ALL particles plasma, trapped protons/electrons,
solar (protons, heavies), GCR - protons .01(?) lt E lt 400 MeV
- electrons 1 keV(?) lt E lt 10 MeV
- Average, worst-case, of time above threshold,
maximum time continuously above threshold - Time scale of variations
- Estimate of uncertainty ? safety margins
- How much due to variability of environment
statistical model - How much due to uncertainty in measurement,
- How much due to modeling assumptions/simplificatio
ns - Verification/Validation
- What aspects of software implementation need to
be considered in determining model architecture? - Elliptical orbits different regions of space
with different time scales, phasing, etc.
28Data Issues
- Level of processing
- time averaging
- calibration, contamination, etc.
-
- Intercalibration different detectors, different
orbits, different epochs - Access to data treatment of sensitive/proprietar
y data - operational/piggyback detectors on commercial
s/c - classified s/c
- High fidelity vs. low fidelity detectors
- Is there any archival data worth using?
29Programmatic Issues
- Long-term commitment to model development and
maintenance - e.g., IRI, original TREMP
- Understanding of financial benefit
- Current approach makes the job difficult
- new proposals every few years ? small chunks of
the total based on anticipated funding levels - development of ad hoc models for specific
engineering applications - need a global roadmap and commitment to follow it
- Inter-Agency, Inter-Program collaboration
- Modeling is critical, but not a PI activity
(Vette, 1991) - Need an impartial modeling center with access
to data a la Aerospace/NSSDC activities - Can engineering modeling needs impact scientific
missions? - Will there be an engineering mission to
map/monitor the environment? - Export restrictions
- collaboration, data sharing
- model distribution
- How to implement model
- E.g., GEOSpace, SPENVIS
- Stand-alone vs. web-based
- Commercial tools (e.g., Space Radiation)