Data Set Information
|
| DATA_SET_NAME |
APOLLO LUNAR SAMPLES BUG OBSERVATIONS V1.0
|
| DATA_SET_ID |
BUGLAB-L-BUG-4-APOLLO-SAMPLES-V1.0
|
| NSSDC_DATA_SET_ID |
|
| DATA_SET_TERSE_DESCRIPTION |
|
| DATA_SET_DESCRIPTION |
Data Set Information:OVERVIEW:--------The Bloomsburg University Goniometer (BUG) was used to make bidirectionalreflectance distribution function (BRDF) measurements of 6 Apollo lunar soilsamples: Apollo 11 (10084), Apollo 12 (12001), Apollo 15 (15071), Apollo 16(61141 and 68810), and Apollo 17 (70181). Each sample was measured at fourwavelengths: 450nm, 550nm, 750nm, and 950nm. Center wavelengths and passbandfor each filter can be found in the table below. Center wavelength (nm) FWHM(nm) 453 50 550 50 751 10 952 9The archive consists of 6 tabular data files, each with an accompanying detachedPDS label. Columns are incidence angle, emission angle, azimuth angle, phaseangle (all in degrees, i and e are measured from nadir), and radiance factor(the reflectance of the sample at some geometry to that of a perfect Lambertiansurface illuminated and viewed normally) at a center wavelength of 453, 550,751, and 952 nm.The data from the Apollo 11 sample and one of the Apollo 16 samples (68810)were analyzed by JOHNSONETAL2013.SAMPLE INFORMATION:------------------Dr. David Paige (UCLA) acquired the lunar samples.DATA SET ACQUISITION:--------------------A typical powdered sample is poured into a circular sample dish, 5.6 cm indiameter and 1.0 cm deep. Each sample was scraped flat with a spatula alongthe container surface. They were then gently tapped to allow the contents tosettle slightly and smooth any obvious surface structure. The sample isilluminated by a collimated spot of light approximately 2.5 cm in diameter atnormal incidence. To make the measurements, the detector reports a voltagebased on the intensity of the light detected in a given wavelength. Forreference, we first measure the voltage from a sample of Spectralon(R) viewedat incidence angle, i : 0, emission angle, e : 5 degrees. All sample voltagesare divided by this reference to give a raw measure of reflectance.Subsequently, this reflectance is multiplied by cos(i)/cos(e) for eachmeasurement to give radiance factor. We multiply by cos(i) because, as theincidence angle increases, the sample is illuminated by an increasingellipsoid spot of light. The detector 'sees' the entire spot, so there is nofall-off in incident intensity as incidence angle increases. Multiplying bycos(i) accounts for this. Dividing by cos(e) accounts for the decreasingapparent area of the sample as seen by the detector. For the standardmeasurement suite, the maximum incidence angle is 60 degrees and the spot iswholly contained within the sample dish. For higher incidence angles, we useda different sample holder (discussed below).The stepper motors that move the incidence and emission arms and rotate thesample stage in azimuth have a nameplate precision of 0.001 degrees. This isfar better than our ability to level a powdered surface in a small samplecontainer and, as a result, we only claim angular accuracies of 0.25 degrees.Subsequent tests with a spectralon standard showed that the original sampleradiance factors were 7.6% too dark (the brightness of spectralon at i:0, e:5is 7.6% too bright when compared with a perfect Lambertian surface at thatgeometry). This is nearly identical to the estimates of non-Lambertianbehavior found in JACKSONETAL1992. All measurements in this archive havetaken that factor into consideration.Because of the architecture of the goniometer, the detector sees the entiresample holder plus some additional area outside the holder. Although allsurfaces have been flocked with low reflectance material, there are residuallow-level secondary reflections from the sample off the mechanism that areobserved by the detector. These reflections have an azimuthal dependencebecause various parts of the sample platform or arms are within the detectorfield-of-view as the stage rotates, so we apply a correction factor beforereporting the radiance factor. The correction factor is determined by holdingemission angle at e : 0 deg, fixing i at 15, 30, 45, and 60 deg in steps, androtating in azimuth for each incidence angle. Because e : 0, there should beno change in the reflected light as the stage rotates in azimuth. Deviationsfrom the minimum reflection observed represent additions due to scatteredlight. A low order polynomial is fit to the change in reflectance as azimuthvaries, and this polynomial is used to correct all sample measurements at agiven incidence angle. The corrections typically increase as incidence angleincreases but are usually on order of 5% or less. Samples measured at 950 nmhave the highest amount of scattered light (black flocking was not aseffective in the IR) and have the most correction.A full set of measurements (cycle) at a given wavelength takes approximatelyone hour. We use a stabilized power supply with active optical feedback tomaintain the intensity within 1% during the measurement cycle. We check thefidelity of our data by duplicating 36 measurements at the same or mirrorpositions throughout each cycle and comparing results. We find our root-mean-square repeatability to typically be within 2% during a sample run.For these samples, we desired high incidence angle observations andconstructed a different sample holder (11.9cm x 2.4cm x 0.5cm). Samples inthis container were measured in the principal scattering plane at i:60, 70,and 75 deg, and perpendicular to (az:90) the scattering plane at i:50, 60,70, and 75 deg. The measurements at i: 50 and 60 overlapped those of thestandard measurements and allow for a consistency check. In general, the datasets were consistent. There are occasional differences, especially at thehighest emssion angles (e : 70, 80 deg) that we attribute to imperfectscattered light removal, differences in sample preparation, and the generallylower SNR at these geometries.For sample 68810, we used a different (older) movement file for these lattermeasurements that differed slightly from the movement file used for the othersamples; although it covers much of the same geometry, it is not sampled asdensely as the latter movement files. Rather than delete those non-sampledgeometries from the 68810 database, it was decided to keep the samplegeometry file identical to that of the other samples and insert 0.000 wherethere is no data.Please see the GEOMETRY directory for information regarding the observationgeometry for the data in the archive.
|
| DATA_SET_RELEASE_DATE |
2017-01-31T00:00:00.000Z
|
| START_TIME |
1965-01-01T12:00:00.000Z
|
| STOP_TIME |
N/A (ongoing)
|
| MISSION_NAME |
|
| MISSION_START_DATE |
|
| MISSION_STOP_DATE |
|
| TARGET_NAME |
MOON
|
| TARGET_TYPE |
SATELLITE
|
| INSTRUMENT_HOST_ID |
BUGLAB
|
| INSTRUMENT_NAME |
BLOOMSBURG UNIVERSITY GONIOMETER
|
| INSTRUMENT_ID |
BUG
|
| INSTRUMENT_TYPE |
RELFECTANCE SPECTROMETER
|
| NODE_NAME |
Geosciences
|
| ARCHIVE_STATUS |
|
| CONFIDENCE_LEVEL_NOTE |
The uncertainty in the data is due to many factors and cannot be easilyquantified for individual points as it might be for some instruments. Thevast majority of the data points are the result of a single measurement.Signal-to-noise varies dramatically even within a given data run. At normalincidence and emission angles with brighter samples, the noise is lower whileat high incidence and emission angles with darker samples, the noiseis higher.Evidence of the overall uncertainty comes primarily from the repeatability ofmeasurements made at the same or mirror geometries during the course of anhour-long data run. For these, we kept track of the worst case difference,both relative and absolute, between any set of repeat measurements. We alsocomputed the root-mean-square difference for the 36 repeats within ourstandard movement file. Below is a table of the relative differences computedfor each file (all in percent). 450nm 550nm 750nm 950nm max rms max rms max rms max rms10084 5.1 1.5 4.8 1.1 1.8 0.6 3.9 0.912001 6.4 1.5 6.2 1.5 5.6 1.2 7.1 2.015071 8.0 2.0 7.0 1.8 4.4 1.2 4.2 1.061141 2.6 0.6 2.5 0.7 2.5 0.5 3.5 0.868810 1.9 0.5 0.8 0.2 1.0 0.3 2.3 0.670181 3.6 0.7 1.7 0.4 2.3 0.5 3.5 0.9For each file, the maximum relative difference between repeated measurementsrange from ~1% to 8%. In every case, the maximum difference occurs whenrepeating measurements at the mirror geometry of the highest phase angle (i :60, e : 80, phase ~140 in or near the scattering plane). For example, here isthe report for sample 12001 at 750nm for the worst case. Inc emis azim phase raw radfac corrected % diff 60.00 80.00 -10.00 138.86 0.02711 0.07806 -60.00 80.00 170.00 138.86 0.02862 0.08241 5.57576Both measurements were made at the same incidence, emission, and phase angle,but the goniometer was oriented in a mirror geometry. Here we note two thingsthat may be responsible for these larger-than-typical differences: low lightlevel (the absolute brightness of samples tends to be lowest at the highestphase angles) and the likelihood of more scattered light in one orientationthan the other. Based on these observations, we assign relativeuncertainties of +/- 10% to all data points. However at most geometries, therelative uncertainties are considerably better (within 1% or 2%) as indicatedby the relatively low RMS values.The absolute uncertainty in measurements with BUG is more difficult toascertain. We have compared the BUG measured reflectance of pressed PTFE(e.g. Spectralon) white and grey scale targets with published values, and theBUG measured reflectance of a soda-lime powder with that of the JPL short armgoniometer [HAPKEETAL2009]. In all cases, the measurements agree quite welland are often indistinguishable. Other than these informal cases, however, wecannot provide a rigorous estimate of absolute uncertainty and conservativelyestimate that the absolute uncertainties in all measurements are within +/-15%.This data set underwent external peer review.
|
| CITATION_DESCRIPTION |
Shepard, M.K. D. Paige, and E. Foote, Apollo lunar sample BUG Observations, BUGLAB-L-BUG-4-APOLLO-SAMPLES-V1.0, NASA Planetary Data System, 2016.
|
| ABSTRACT_TEXT |
The Bloomsburg University Goniometer (BUG) was used to make bidirectional reflectance distribution function (BRDF) measurements of Apollo lunar samples.
|
| PRODUCER_FULL_NAME |
E. FOOTE
MICHAEL K. SHEPARD
D. PAIGE
|
| SEARCH/ACCESS DATA |
Geosciences Web Service
|
|
.png)
.png)
