1. Future Developments of the Lunar Calibration System
Thomas C. Stone
U.S. Geological Survey, Flagstaff, AZ USA
GSICS Lunar Calibration Workshop
EUMETSAT
01–04 December 2014

2. Outline Short-term improvements
Longer-term: lunar model re-development
Refinements to the ROLO dataset
New measurements of the Moon
Propagating results of the refinements

3. Immediate Need: Addressing Residual Geometry Dependencies
If a consistent pattern shows up in measured/model lunar irradiance comparisons for many instruments, it gives an indication of a residual geometry dependency in the lunar model.
Before addressing any geometry effect, the dependency must be attributed to the model, with no influences of sensor image processing
How? Consistency across instruments is the key indicator.
particularly important for phase angle effects
But: the basis dataset for the lunar model also involves image processing, i.e. ROLO data
Geometry effects have been revealed by new, more stable instruments
phase angle dependency seen in instruments that capture a wide range of phase angles: PLEIADES, MSG
libration dependency seen in long-term datasets: MODIS, VIIRS
the relative accuracy of the lunar radiometry is an important consideration

4. Immediate Need: Addressing Residual Geometry Dependencies (2)
Presuming an effect is shown to be a dependency in the lunar model, what can be done?
Look for potential causes in the basis (ROLO) dataset
look for correlations with the geometry parameters, using the observed effect as a guide
Apply constraints in the process of fitting the ROLO dataset, and re-generate the coefficients of the lunar model
develop constraints to address the specific dependencies
The PLEIADES data have the quality and consistency to address the phase angle dependency
High density of measurements acquired over a short period of time
several months with Moon acquisitions every orbit
long-term sensor drift is not an issue

5. The PLEIADES Moon Image Dataset
Two figures S. Lachérade CNES

6. Longer-term: Redevelopment of the Lunar Model
Starts with reprocessing the ROLO dataset:
Refined Moon image processing to irradiance
revised method for handling spatial scattering around the Moon disk
Refined correction for atmospheric transmission
star image processing to irradiance
constraints in the stellar extinction processing
includes Vega images → absolute calibration
A new fit to the reprocessed ROLO dataset will produce:
New coefficients for the lunar reflectance model
A new form for the model, including:
wavelength-dependent libration terms
possibly a solar selenographic latitude term

7. New Measurements of the Lunar Irradiance
A new basis dataset for lunar modeling is needed, to address known deficiencies of ROLO:
SI traceability
Absolute scale uncertainty
Residual geometry dependencies
Multi-band spectral coverage
A new lunar measurement program must acquire many years of observations, to characterize the variations of the lunar brightness with phase and librations.
Minimum 3 years to capture the libration dependence within the bounds of its relative effect
ROLO operated for more than 8 years, which greatly contributed to the success of building the lunar irradiance model

8. New Measurements of the Lunar Irradiance (2)
Discussions with NIST about a lunar measurement program led to several experiment design considerations:
A long-term ground-based observation campaign, plus a flight component, e.g high-altitude balloon or space-based deployment
ground measurements to characterize the geometry dependences
flight measurements to set the absolute scale, free from atmospheric effects
Full spectral coverage is needed, for both the lunar irradiance and the atmospheric transmission measurements (by stellar extinction)
A dedicated irradiance (non-imaging) radiometer instrument, to avoid the pitfalls of image processing
Such a project may require the support of a National Measurement Institute (NMI), since a calibration measurement program is too costly for any one satellite operation agency to fund, and a substantial commercial value for lunar calibration has not been identified (yet).

9. New Measurements of the Lunar Irradiance (3)
Lunar Spectral Irradiance (LUSI) project
NIST initiative; inspired by the earlier discussions
Prototype instrumentation developed and tested at Mt. Hopkins, AZ
non-imaging telescope feeding an integrating sphere and spectrometer
observations of a calibrated sphere source several times each night
described by C. Cramer at 2013 CALCON (
current status: seeking funding
HyperSpectral Imager for Climate Science (HySICS)
Univ. Colorado Laboratory for Atmospheric and Space Physics (LASP)
Earth reflectance sensor for benchmark climate measurements
CLARREO prototype; simultaneous Earth and Sun observations
HySICS also views the Moon — can provide reference lunar reflectance
flight instrument; balloon-based test flights: Sept and August 2014

10. New Measurements of the Lunar Irradiance (final)
A new dataset can initially be used to constrain and/or scale the existing lunar model, until enough measurements are acquired for valid model development using the new data.
The goal: an absolute calibration reference that is the Moon.
Requires reducing absolute uncertainties to under ~1%
The lunar surface is exceptionally stable, and the Moon’s brightness variations are periodic and predictable
allows modeling, and a lunar reflectance model is valid for all times
The level of accuracy achievable for a lunar model is limited only by the accuracy of the basis irradiance measurements
i.e. the accuracy limits for field measurements of a natural target — 0.5% ?
improved accuracy ultimately should require accounting for solar variability, thus introducing a time dependence into the lunar model

11. Propagating Lunar Model Improvements
After validating refinements to the lunar model, how to get new results to the users?
For sponsored sensors (e.g. OLI, VIIRS), reference lunar irradiances for Moon observations will be generated by USGS and delivered to the instrument teams
For updating the GIRO, the new model formulation and coefficients must be obtained from published materials
U.S. restrictions on technology transfer (ITAR)
online journals have goals for fast review and publication processes

12. Timeline Addressing residual geometry dependencies:
Collaboration between USGS and CNES on using the PLEIADES dataset has already started; visit in November 2014
Goal is to publish before the end of 2015
Refinements to the ROLO dataset and lunar model redevelopment:
ongoing effort, sponsored by USGS Land Remote Sensing Program
expect to be completed within 2 years
New measurements of the Moon:
long-term effort; constrained by the physical ability to view the Moon with a sufficient sampling of phase angles and libration states
LUSI (NIST) and HySICS (LASP/CLARREO) projects have done initial data acquisitions, but neither is making continuous measurements
will require a nearly mission-level commitment
a consortium of NMIs and/or satellite operation agencies?