VOL. 80, NO. 17 JOURNAL OF GEOPHYSICAL RESEARCH JUNE 10, 1975 Acquisition and Description of Mariner 10 Television ScienceData at Mercury G. EDWARD DANIELSON, JR., AND KENNETH P. KLAASEN Jet PropulsionLaboratory, Pasadena,CalijCornia 91103 JAMES L. ANDERSON California Institute of Technology,Pasadena, Calij•ornia 91125 The Mariner 10 televisionscience subsystem was an improved versionof the Mariner 9 system,using 1500-ram-focal-lengthoptics. An elaboratepicture-taking sequence resulted in transmissionof over 4000 framesback to earth during two flyby encounterswith Mercury. Thesesequences utilized a real-timedata rate of 117.6 kbit/s, resultingin coverageof about 75% of the lightedportion of Mercury's surfaceat a resolutionof betterthan 2 kin. The completeset of usefulimages, which amounted to about 3000 frames, was processedwith three different types of digital image-processingenhancements. INTRODUCTION in Figure 3. The relativefilter factors(ratio of integratedsys- tem spectralresponse without any filter to that with a given The Mariner 10 spacecraftencountered the planet Mercury filter) for each filter are listed in Table 1 for each camera,for for the first time on March 29, 1974. Closest approach oc- input radianceswith both solar and mercurianspectra. curred on the dark side of the planet at an altitude of about A detailed summary of all of the functional characteristics 700 km. A secondencounter occurred on September21, 1974. of the cameras is given below. The aim point for thisflyby was chosen to be on the brightside at a range of about 50,000 km and about 45 ø south of the Characteristic Value equatorial plane. The spacecraftcarried twin, long focal length Focal length 1500mm television cameras to photograph the surface of Mercury. f/number f/8.4 About 3000 scientificallyuseful pictures were returnedfrom Field of view 0.36 ø X 0.48 ø both encounterswith surfaceresolution of up to 120 m. About Scanned area 9.6 X 12.35 mm 40% of the surface of Mercury was photographed at better Format 700 X 832 pixels than 2-km resolution. Encodinglevel 8 bits Frame time 42 s CAMERA CHARACTERISTICS Resolutionper TV line 9.5 X 10-6 rad The Mariner 10 television sciencesubsystem is similar in PICTURE NUMBERING AND DATA many respectsto its predecessoron Mariner 9, which so suc- t TRANSMISSION cessfully mapped the surface of Mars. One of the major differenceswas the optics. In order to increasethe high-resolu- Each Mariner 10 picture was assigneda unique identifica- tion coverageon the chosenflyby trajectory [Dunne, 1974] the tion number. The on-board electroniclogic in the flight data focal length of the Mariner 10 telescopewas increasedto 1500 subsystem(FDS) began numberingframes starting on the mm, 3 times that of the Mariner 9 high-resolution telescope. launch pad about 6 hours before lift-off. The numbersin- An auxiliary, wide-angle (50-mm focal length) optical system, crementedby one for eachsuccessive 42-s frame.The FDS al- accessedthrough the filter wheel, was added as well. A de- waysassigned A camera frames odd numbers and B camera taileddescription ofthe optical design, illustrated inFigure 1, frameseven numbers. Several times in flight,spacecraft is given by Larks [1974]. anomalies caused the counter in the FDS to reset itself to zero. The sensorwas an improved version of the seleniumsulfur (One resethappened just as the spacecraftwent into solar oc- photosurfaceslow-scan vidicon. The electronicdesign was im- cultation at Mercury encounter. Thus the incoming pictures proved to reduce the camera's susceptibilityto random elec- have been uniquely separatedfrom those taken on the out- tronic noise. A significantdesign change which resultedin an going leg of the trajectory at the first Mercury encounter. improvement over Mariner 9 was the incorporation of light Around closestapproach the numbersare 274XX on the in- flooding (F. Vescelus,unpublished data, 1975), which solved coming side and 40-200 on the outgoing side.) However, any the residual image problem that had plagued earlier Mariner picture can be uniquely identified (to the National Space television data reduction, especiallyon Mariner 6, 7, and 9 ScienceData Center, for example) by stating the major target [Young, 1974]. body along with the FDS number (e.g., earth FDS 14553 or The systemspectral transmission characteristics are shown Mercury FDS 48). in Figure 2. The opticsspectral transmission is plotted with the The Mariner 10 spacecraft could handle data in several spectraltransmission for eachfilter. The completesystem spec- modes. Data were recorded automatically on the on-board tral response(in amperesper unit area of vidicon surface)for tape recorderand playedback at a slowerrate (22.05 kbit/s) as each filter has been calcalatedfor an illuminating sourcewith much as possible.Since the tape recordertook 2.24 hours to Mercury's spectrum, as published by McCord and Adams play ba• 36 pictures at this rate and surface resolution [1972]. The responsecurves have beennormalized and plotted changedat a rate of about 0.8 kin/h, full-disk, high-resolution coveragenear Mercury encountercould not be accomplished Copyright¸ 1975by the AmericanGeophysical Union. by using this mode. Thus the pictures near closestapproach 2357 2358 DANIEL•ONETAL.: MARINER 10 MISSION I I I I I I ! I I J I I I I •- --__ I ! i I I I ! Fig. 1. Opticalschematic of Mariner 10 high-resolutiontelevision subsystem. were sent back in real time at 117.6 kbit/s (i.e., one picture Far Encounter transmittedevery 42 s with no intermediatestorage) but with a The incomingfar-encounter sequence began as soonas Mer- correspondinglynoisier signal. During the missionthis trade cury could be viewed within the camera-pointingconstraints off betweendata rate and noisecould be madeindependently of the other sciencedata, since those data were telemetered in of the spacecraft.The primary objectivesof the far-encounter an independenttelemetry channel. A third modeof data return sequencewere (1) to obtain imagery of Mercury at pro- was transmissionin real time at 22.05 kbit/s. This slower, real- gressivelybetter resolutionwhile approachingthe planet and (2) to checkout and calibrate the televisionsubsystem and the time data rate allowedonly aboutone fifth of the total picture data to be returned for each frame. Two edit modes were avail- telemetry link. The intent of the initial sequence6 daysprior to encounter able to selectthe data to be returned.In one, only the center was to photographMercury through each spectralfilter at a strip (one quarter of the frame wide) was returned. In the minimum of two different exposurelevels (five levelsfor the other, calledthe skip-slidemode, only everyfourth pixel in a clearfilter) in orderto verify systemsensitivity as a functionof line was returned,with the first pixel returnedin a line being exposureover the dynamicrange of the instrument.Scan plat- shiftedover by two pixels from the previousline. In both of form pointing offsetsand motion of the spacecraftwithin its theseedit modesthe data were encodedwith only 6 bits. attitude control dead band resultedin obtainingonly about SEQUENCE one-half the desired number of exposure levels (Table 3). Therefore a completedetermination of the systemresponse A summaryof the imagingsequence for the first Mercury encounter is shown in Table 2. overthe entire dynamic range was not Possible. Enough data were obtained, however, to allow revised exposurecalcula- 1.0 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I tions. The sequenceson subsequentdays consisted of imagery CLEAR through all filters at midscaleexposure levels and at pro- MINUSULTRAVIOL.E• gressivelybetter resolution.After flying by Mercury on the 0.8 - - ULTRAVIOLET BLU ' le0 -- 0.9 0.8 0.6 II • • OPTICSAND i 0.7 • 0,,6 0.4 ,.q 0.5 • 0.4 OZ0.3 0.2 0.2 0.10 7/ 300 400 5•0 600 70• 0 300 400 500 600 700 WAVELENGTH, NM WAVELENGTH• nm Fig. 3. Integratedoptics, filters, and vidiconsystem response in- Fig. 2. Spectraltransmission of Mariner 10 televisionsubsystem op- dependentlynormalized for each spectralfilter on the basisof the ab- ticsand spectralfilters in percentplotted as a functionof wavelength. solute Mercury spectrumand plotted as a functionof wavelength. DANIELSONET AL.: MARINER I0 MISSION 2359 TABLE I. Mariner 10 Relative Filter Factors Solar Radiance Mercurian Radiance Filter A Camera B Camera A Camera B Camera Clear (CLR) 1.08 1.08 1.08 1.08 Minus ultraviolet(MUV) 1.57 1.49 1.44 1.39 Ultraviolet polarizing(UVP) 43.93 54.23 60.25 75.02 Blue (BL) 2.64 2.46 2.66 2.48 Orange (OR) 5.37 5.66 4.17 4.47 Ultraviolet (UV) 11.63 13.70 16.01 19.05 TABLE 2. First Mercury Encounter Sequence Resolution, Phase Range, km km Frames FDS N umbers Incomingfar encounter,-6 to - 1 day 5,700,000-800,000 280-20 546 14339-25728 Incomingcolor mosaicking,- 16 to 635,000-100,000 14-3 162 25927-27104 -4 hour Close encounter, -4 to -I-4 hours 100,000- 5500 3-0.12 592 27207-27477 0-392 Outgoingcolor mosaicking,-I-4 to 100,000-635,000 3-14 144 494-1354 q- 16 hours Outgoing far encounter,q- 1 to q-3days 800,000-2,800,000 20-60 108 2055-2590 5996-6049 Satellite search,+ l to q-3 days 1,000,000-3,500,000 627 3932-5418 8045-8106 Total 2179 dark side a similar type of far-encountersequence was per- The 117.6-kbit/s data rate was used,yielding full-frame, full- formed,beginning about 1 day after closestapproach. Both in- resolutionpictures at an averagebit error rate of about 1 in 40 coming and outgoing data were played back from the space- [Clarke and Evanchuk, 1974]. Figure 6 shows the planned craft tape recorder at 22.05 kbit/s and had a bit error rate of mosaic patterns and the resolution ranges for the first four lessthan I in 1000 bits.