Data Set Information
DATA_SET_NAME CLEMENTINE LWIR BRIGHTNESS TEMPERATURE V1.0
DATA_SET_ID CLEM1-L-LWIR-3-RDR-V1.0
NSSDC_DATA_SET_ID
DATA_SET_TERSE_DESCRIPTION
DATA_SET_DESCRIPTION Data Set Overview:The scientific payload on the Clementine spacecraft included aLong-Wave Infrared (LWIR) camera with a single passband of width1.5 microns centered at a wavelength of 8.75 microns. The LWIRspatial resolution ranged from 200 m per pixel near the poles to55 m per pixel at the equator. Contiguous pole to pole imagingstrips were obtained with approximately 10% overlap betweenadjacent frames. However, significant longitude gaps existbetween successive orbital passes. During the systematic mappingphase of the Clementine mission, approximately 220,000 thermal-infrared images of the lunar surface were obtained. This dataset contains LWIR images where the observed radiance has beenconverted to brightness temperature, which provides informationon various physical properties of the lunar surface.Processing:Preflight calibration for the LWIR instrument was performed atLawrence Livermore National Laboratory in an effort to measurecamera characteristics such as radiometric sensitivity, gain andoffset scale factors, temporal/spatial noise, and the dependenceof dark-noise on focal plane array (FPA) temperatures. Severalsteps were involved in the generation of brightness temperaturevalues which included converting measured data number (DN) valuesto radiance values; identifying and eliminating bad pixels;correcting for pixel response variation across the detectorarray; determining the zero-flux background of the instrument;comparing LWIR measured radiance of the Apollo 17 landing site toin situ temperature measurements in order to derive absolutecalibration adjustments; and finally, converting measuredradiance values to brightness temperatures via the Planckfunction [LAWSONETAL2000].The first step in the routine to determine brightness temperaturefrom LWIR images involved converting measured DN values (rangingfrom 0 to 255) to equivalent radiance values through a preflightcalibration equation. This equation corrected for the changinggain and offset states used throughout an imaging orbit in orderto account for the increase in surface thermal emission near theequator.The primary uncertainty in the calibration was the subtraction ofa zero-flux radiance. The dark-frame signal was extremely largeand varied with time in orbit and probably with lens and FPAtemperatures. Through the 1.5 hours of a lunar imaging orbit,the heat input to the spacecraft by the cryocooler increased thesurrounding temperature and decreased the cooling effectiveness,resulting in rising FPA temperatures with time. This zero-levelbackground increase was much less pronounced in the first monthof systematic mapping than in the second. At the mid-lunar-mapping-mission orbital-correction burn, fuel was expended andthe overall thermal mass of the spacecraft was reduced. Thisresulted in an increased operational temperature and caused theLWIR background counts generated from the pixel dark current andthe thermal emission from within the optical path to increase.Thus, after the mid-lunar-mapping orbital burn, the number ofLWIR images taken during an orbital pass was reduced to avoidsaturation of the detector. In order to subtract a zero-fluxradiance from the LWIR images, a series of space-looking frames(away from the Sun, Moon, and Earth) was used that were acquiredat the beginning and end of many lunar mapping orbits. Theincreasing background level through an orbit was accounted for byfitting a line, on a pixel-by-pixel basis, to the pre-mappingspace and post-mapping space radiance values as a function oftime.The next step in the process was to create bad-pixel maps andflat fields. After subtracting a zero-level image from eachlunar image, hundreds of lunar images from a single orbit wereaveraged together on a pixel-by-pixel basis. Thus a lunar meanimage and an associated lunar standard deviation image weregenerated. Lunar images at latitudes higher than 70 deg wereneglected in the calculation of the mean image because they wereoften either saturated (due to excessively high gain states) orthey had values less than zero after the zero-level imagesubtraction. Bad pixels were identified on the mean and standarddeviation images as pixels that did not vary (low or zerostandard deviation), varied randomly (high standard deviation),or were pegged at high values such that their dynamic range waslimited (high mean). Pixels with low means also appeared bad onlunar images. An average of approximately 11% of the 16,384detector array pixels was characterized as consistently bad.There were approximately 10 orbits where at least one map hadmore than 20% bad pixels.In order to correct for pixel response variation across thedetector array, a flat field frame was created by multiplying thelunar mean image and the bad-pixel map, smoothing over the badpixels, and normalizing the resultant image to unity. For an IRcamera, the FPA signal comes not only from the signal generatedfrom the lunar surface, but also from dark current generatedinternally within each pixel detector and from thermal emissiongenerated from the camera optical path. For the LWIR themagnitude of these contributions was very large and varied boththrough an orbit and through the mission. Thus three separatebad-pixel maps and flat fields were required for each orbit. Thenumber created was constrained by the number of lunar images overwhich to average. The minimum number of images to average wascontrolled by the need to eliminate structure due to scenevariations that result from surface topography. The maximumnumber of images to average was limited by the varying thermalemission of the lunar surface through an orbit; the measuredradiance values were much higher near the equator than near thepoles. To create the mean and standard deviation images,latitude bins were averaged together in the following way: 70S to20S, 20S to 20N, and 20N to 70N. Lunar images from latitudeshigher than 70 degrees used flat fields and bad-pixel maps fromadjacent latitude bins.Inflight absolute calibration was accomplished by comparing theLWIR-derived temperatures at the Apollo 17 landing site totemperatures determined in situ from the heat-flow experiment.Although in situ temperatures were also measured at the Apollo 15site, LWIR frames are available only for the Apollo 17 site. Thedifference between Apollo temperatures and LWIR temperatures wasapproximately 17 K, corresponding to a 22% difference inradiance. This 22% change in radiance reflects both thelimitations of the preflight calibration and the differencesbetween prelaunch and inflight calibrations. Therefore, all ofthe LWIR-derived radiance values were multiplied by 0.82.Brightness temperatures were then calculated using the Planckfunction assuming unit emissivity.The bad pixel routine did not identify all bad pixels due totheir shifting locations from image to image. The few remainingbad pixels were eliminated via a sigma-filter routine thatreplaced pixels that were greater than 3-sigma away from the meanof a 3 x 3 pixel region around each pixel. Finally, the LWIRtemperatures were corrected for the varying Sun-Moon distancethroughout the mission.
DATA_SET_RELEASE_DATE 2002-07-01T00:00:00.000Z
START_TIME 1994-02-27T07:04:02.174Z
STOP_TIME N/A (ongoing)
MISSION_NAME DEEP SPACE PROGRAM SCIENCE EXPERIMENT
MISSION_START_DATE 1991-11-19T12:00:00.000Z
MISSION_STOP_DATE 1994-05-07T12:00:00.000Z
TARGET_NAME SKY
MOON
TARGET_TYPE CALIBRATION
SATELLITE
INSTRUMENT_HOST_ID CLEM1
INSTRUMENT_NAME LONG WAVELENGTH INFRARED CAMERA
INSTRUMENT_ID LWIR
INSTRUMENT_TYPE CAMERA
NODE_NAME Geosciences
ARCHIVE_STATUS
CONFIDENCE_LEVEL_NOTE Preflight calibration data were acquired with an automatedcalibration facility at Lawrence Livermore National Laboratory.The preflight calibration measurements included radiometricsensitivity; FPA uniformity; gain and offset scale factors;temporal/spatial noise; dark noise dependence on FPAtemperatures, integration times, input voltage levels, andspectral response of FPA; optical distortion map; point spreadfunction; electronic warm-up time; and cryocooler cool-down time.Absolute uncertainties were found from the zero-level radiancedetermination and are constant in radiance but vary in equivalenttemperature throughout an orbit, being lower near the equatorwhere the surface temperatures are higher. For the entiremission, a typical 2-sigma uncertainty is 7 K or greater.Absolute uncertainties were propagated on a pixel-by-pixel basisthroughout the calibration routine, and then image-averaged.Adjacent LWIR frames overlap and there is consistently less thana 1% difference in brightness temperature between overlappingregions. Thus, the statistical uncertainty of a singlemeasurement is much less than the absolute uncertainties throughan orbit.The LWIR camera cannot measure temperatures below approximately150 K. In the calibration routine, a zero-level image issubtracted from each lunar radiance image. Sometimes thisresults in pixel values less than zero, particularly near thepoles where the temperatures are low. All pixel values that areless than zero in radiance are set to 0.0001 and then convertedto temperature. This results in temperatures around 45 K thatare not real and have large associated uncertainties.Temperatures below 150 K are not to be trusted.Several orbits worth of data were not reduced, either becausethere were no space images taken for that orbit or because thedetector saturated. LWIR images from 95% of the orbits from thefirst month of lunar mapping and from 77% of the orbits from thesecond month of lunar mapping were reduced. Thus data from atotal of 86% of the 266 Clementine systematic mapping orbits wasreduced.For more information on the LWIR calibration routine, see[LAWSONETAL2000].
CITATION_DESCRIPTION Lawson, S. L., and B. M. Jakosky, Clementine LWIR Brightness Temperature Archive, CLEM1-L-LWIR-3-RDR-V1.0, NASA Planetary Data System, 2002.
ABSTRACT_TEXT The Clementine Long-WaveInfrared (LWIR) Brightness Temperature data set containscalibrated brightness temperature images that provide informationon physical properties of the lunar surface.
PRODUCER_FULL_NAME BRUCE JAKOSKY
STEFANIE L. LAWSON
SEARCH/ACCESS DATA
  • Geosciences Web Services
  • Lunar Orbital Data Explorer
  • Geosciences FTP Resource
    		•