MRO ACC Level 0 review Overall impression The structure is good. The promised data products are good. The documentation is good, but I hope it will be supplemented by a detailed technical paper in the peer-reviwed literature. There are plenty of minor inconsistencies and missing files. I have plotted most of the data products in DATA/ and no entries appeared invalid or crazy. ROOT DIRECTORY and general comment No AAREADME file No ERRATA file No VOLDESC file Some XXXINFO.TXT files list all files except XXXINFO.TXT, but some include the XXXINFO.TXT file. e.g. CATINFO.TXT does not list CATINFO.TXT, but DOCINFO.TXT does list DOCINFO.TXT. CALIB/ CALINFO.TXT lists: AERODYN.FMT AERODYN.TAB SC_DATA.TAB But actual contents are: ACCEL.TAB ORBELEM.TAB RATEQUAT.TAB SC_DATA.TAB THRUSTER.TAB ACCEL.TAB, ORBELEM.TAB, RATEQUAT.TAB, and THRUSTER.TAB appear to be the Level 0 DATA files for orbit 100. AERODYN.TAB will be valuable for data users. I presume it will contain Cx, Cy, Cz and the corresponding moment coefficients as functions of two angles describing MRO attitude and density. The conventions used to define the angles should be clearly stated. SC_DATA.TAB Contents are OK. "Phase 1 of aerobraking" is referred to. Did MRO have Phases? The uncertainties in A, L, and m should be stated. CATALOG/ CATINFO.TXT lists: MISSION.CAT INSTHOST.CAT INST.CAT REF.CAT PERSON.CAT DATASET.CAT What controls the order of this list? Files are not listed alphabetically. But actual contents are: INST.CAT PERSON.CAT REF.CAT INST.CAT The spacecraft y-direction is discussed. Is the spacecraft reference frame used here the MRO_SPACECRAFT frame (NAIF ID -74000, ftp://naif.jpl.nasa.gov/pub/naif/MRO/kernels/fk/mro_v01.tf)? If so, this should be stated. If not, the frame should be defined clearly. References to published papers, or an ASCII diagram like those in ftp://naif.jpl.nasa.gov/pub/naif/MRO/kernels/fk/mro_v01.tf, can be helpful here. "The accelerometer measures the change in velocity of the spacecraft in the spacecraft y-direction over 1 second intervals." No, it doesn't. Changes in velocity due to gravitational accelerations are not measured by an IMU. INST.CAT is a good place to restate the IMU position (given in SIS_ACC), plus the uncertainties in IMU position. What is the uncertainty in the alignment of the y-axis accelerometer with the MRO_SPACECRAFT frame y-axis? Does any of the documentation say (A) that the velocity of the spacecraft relative to the atmosphere was approximately aligned with the -y axis, causing the acceleration experienced by the spacecraft to be approximately aligned with the +y axis and (B) that the velocity/acceleration vectors were typically aligned with the +/- y axis to within XXX degrees (eg 5 degrees)? This information should be stated somewhere. PERSON.CAT is fine REF.CAT has two publications, a small number. Are there other engineering publications or documents that can also be referenced? If the MRO special issue of Science appears before this archive is released, then those papers can be added. Tolson paper uses "et al." whereas Bougher paper lists all authors. This is inconsistent. Bougher paper includes co-author Zurek, not Zerek. DOCUMENT/ DOCINFO.TXT lists: SIS_ACC.TXT But actual contents are: SIS_ACC.RTF SIS_ACC.TAB SIS_ACC.TAB looks like bad LaTeX with many \par tags SIS_ACC.RTF Is Shaun Brown at U Maryland (SIS_ACC) or NCSU (PERSON.CAT)? On my computer, lines in Section 1 are formatted like this, with one long line followed by one very short line. Does this happen to other users? If so, the linebreaks should be fixed. The applicable documents are not defined clearly enough for a user to find them. The 2007 AIAA/AAS paper needs to list the authors. The paper number (AAS 07-183) could also be stated. The title of the AIAA/AAS paper does not agree with the title in the conference program (http://www.space-flight.org/AAS_meetings/2007_winter/conference program 1-9-07.pdf, page 39 of 60). The EPSC paper also needs authors and the paper number. The EPSC paper should be included in REF.CAT The location of the CM is stated, which is good. What frame is used to describe the CM location? MRO_SPACECRAFT? Something else? The text says "located at approximately (-0.85, -0.55, -0.57) m" How approximate is approximate? +/- 10 cm? Is the CM location constant in time and the approximate location is approximate because of the quality of the CM determination pre-flight? Or does the CM location change with time due to fuel consumption (monotonic effect) or due to fuel slosh (variable effect)? Perhaps the SIS_ACC is not the best place for such information, but it should be discussed somewhere in the archive. Temperature sensitivity is less than 30 micro-g per degree C. How big is one g? 10 ms-2? 9.8 ms-2? Some other value? Planned data products (section 4.2) are good. The 1.26 nbar level will be archived. No previous ACC archive or publication has explained how this level is calculated. Presumably dp/dz = density x g is used, but this is never stated. What is the upper boundary condition? This comment applies primarily to the higher level deliveries. Altitude will be archived for Level 2 profiles and data products presented at constant altitude levels in Level 3. Altitude is defined relative to an ellipsoid areoid with stated a and f. This information should be repeated as often as possible in connection with the Level 2 and 3 data products, because many users might expect the altitude to be relative to something like the 1 deg x 1 deg MOLA areoid. This comment applies primarily to the higher level deliveries. 4.3.3 Opeational - typo 4.3.3 Version 1 and 2 data products are discussed. This does not seem to mean Level 0, 2, 3, but rather indicates a preliminary data delivery followed later by a revised and improved delivery. Is this interpretation correct? I did not see any other discussion of Versions. 4.3.4 Is the data file *.tab or *.TAB (same for .FMT) 4.5 Data validation is TBD accel.tab, ratequat.tab, thruster.tab, orbelem.tab Should these names be upper case or lower case? 5.2 Level 0 arrays described as n-by-1 (should be n-by-2), n-by-8, k-by-9, and 1-by-6. Why change from n to k for thrusters? Column 2 corresponds to thruster 1. References to documents that describe the orientation of thruster 1 in the MRO_SPACECRAFT frame and that describe the (approximate) force of thruster 1 when it is firing would be helpful. Otherwise the thruster data is useless to users. Level 2 Pxxx.tab has 11 columns numbered 1 to 10. Typo. The text here describes 8 columns, but the data file contains 11. The last six columns are 1-sec z, 1-sec rho, 1-sec sigma, 39-sec z, 39-sec rho, 39-sec sigma I understand how 1-sec densities and sigmas are averaged to obtain 39-sec densities and sigmas, but why are the 1-sec and 39-sec altitudes different? How are the 1-sec altitudes processed to obtain a different set of altitudes? 5.1 says Level 3 data up to 170 km, 5.2 says up to 200 km. Which is it? 5.1 promises a max density line in the Level 3 data tables, 5.2 does not. The sequence of lines in the Level 3 data table is described one way in 5.1, but another way in 5.2. 5.1 has periapsis, constant altitude levels, 1.26 nbar, max density. 5.2 has periapsis, 1.26 in, 1.26 out, 100 km in, 100 km out, ..., 200 km in, 200 km out. Level 3 will include the longitude of the Sun. I can never remember the exact name of this property, Ls, but "longitude of the Sun" doesn't sound quite right. Is it "aerocentric longitude of the Sun"? Having an acronyms list here is excellent. Missing are CM, GDS INDEX/ INDXINFO.TXT promises INDEX.LBL and .TAB. .TAB is missing. LABEL/ LABINFO.TXT lists five files. The directory contains those five, plus CONSTALT.FMT and PROFILE.FMT. The format files for the level 0 data look fine, but see my comments on DATA/ DATA/ RATEQUAT.TAB Quaternions. Quaternions contain 4 elements, 3 of which define an axis and 1 of which is an angle. Or something like that. Anyway, 1 of the elements is distinct. Some conventions have this as the first element, others as the last element. Which convention is used here? Since the last element is always positive, I presume this is the distinct element. These quaternions describe the attitude of MRO with respect to some other frame. Relative to which frame? Mars-fixed? Sun-fixed? Earth-fixed? THRUSTER.TAB Thrusters Column 2 corresponds to thruster 1. References to documents that describe the orientation of thruster 1 in the MRO_SPACECRAFT frame and that describe the (approximate) force of thruster 1 when it is firing would be helpful. Otherwise the thruster data is useless to users. The time interval between thruster data entries varies greatly (1s for P016, 30s for P386) and often varies during one drag pass. The change in data collection procedure should be described somewhere in the archive's documentation. ORBELEM.TAB P016/ORBELEM.TAB 2.546819626799306E+004 8.579538575990308E-001 1.630682659854011E+000 1.052114194718479E-001 5.053107399226665E+000 1.972150908230000E+008 The last entry (1.97E8) is, according to ORBELEM.FMT, tau of periapsis in radians. 1.97E8 radians? I don't use orbit elements much, so I'm not familiar with typical representations of them. I know a, e, and i. I've seen longitude of the ascending node and argument of periapsis in textbooks, but I don't recognise tau (http://en.wikipedia.org/wiki/Orbital_elements). What is tau? ACCEL.TAB ACCEL.FMT says date/time has: START_BYTE = 1 BYTES = 21 and accel has: START_BYTE = 26 BYTES = 13 Here is the first entry in P016/ACCEL.TAB: 2006/092T01:55:48.000 -3.403540e-008 Byte 26 is in the middle of the acceleration data entry. The acceleration data entry is about 14 characters long, not 13. I haven't checked the other data products for similar errors. Why are ax and az not archived? Future users might find them useful. The first accel measurements on orbits before about P350 have a good baseline of data outside the atmosphere. The mean of data points 10-100 for orbits before P350 is about 2E-6 ms-2. The standard deviation of data points 10-100 for orbits before P350 is about 1.3E-6 ms-2. These pre-entry data points are scattered around the mean in manner suggestive of being Gaussian noise, but I haven't checked that quantitatively. Things are different for the later orbits, eg P418, P445. These have a systematic offset of 3E-5 ms-2 and some brief spikes. What's changed?