*************************************************** Caveats for End Users of IUVS Data Products *************************************************** Last updated 2017-11-14 Appropriate for v07 data release *********** Cosmic Ray Detector Events ********** The image-intensified 2-D active pixel sensors used in IUVS are sensitive to cosmic radiation. Such events result in 'splashes' and 'streaks' with random orientations on the detector images. The impact is often negligible when observing bright targets (e.g. the Martian disk) but can become a major source of noise for sensitive observations (e.g. coronal scans). A unified method for cleaning the data of these events is still under development. The user should be aware that most transient unexpected features they find in the data are attributable to these cosmic ray events. ************ Calibration Uncertainty ************ The IUVS instrument currently maintains systematic absolute calibration uncertainties of 25% and 30% for the FUV and MUV channels respectively. This makes retrievals using many emission features unreliable, with optically thick emissions being particularly impacted due to non-linear responses. Only retrievals which have been found robust to the current calibration uncertainty are included in the current data release. A calibration correction factor has been applied for disk observations that use the electronic shutter. The correction factor has inaccuracies in the 315-325 nm region on the order of 17%. ************ Slit Tilt and Wavelength Shift in L1C MLR *********** The image of the slit on the IUVS detectors has a significant tilt relative to the detector pixel columns, especially in the FUV channel. This produces artifacts during multiple linear regression (MLR) fitting of emission brightnesses in L1C apoapse products, causing significant artificial structure in FUV the disk maps. The variations in brightness associated with these artifacts are currently captured by adding a fixed factor of 10% to the reported random uncertainty. The position of the slit image on the detector can also shift due to thermal gradients across the instrument case, with movements of ~ 1 detector pixel over the course of an orbit being common. This results in small but occasionally significant shifting of the wavelength scale in both channels. This is currently accounted for only in the L1C periapse data products with a per-orbit wavelength shift correction. ****** MUV Contamination in FUV Data ******* When IUVS makes observations of the Martian surface or lower atmosphere, the FUV channel experiences significant contamination by 1st order MUV sunlight reflected from the planet. This most significantly impacts stellar occultations on the dayside and disk mapping observations. For disk observations, the multiple-linear-regression (MLR) fitting of FUV features in apoapse L1C products incorporates a term to remove this artifact, with generally good results. However, variations in disk MUV spectrum due to aerosol loading and ozone absorption from observation to observation are currently not accounted for. Data products at L1A and L1B are uncorrected, and the user is urged to exercise appropriate caution when working with them. ************ Side Segment Timing Accuracy *********** During observations of limb airglow on the MAVEN orbit side segments, reconstructed tangent altitudes are highly sensitive to the reconstructed timing relative to periapsis. There is evidence that the current reconstructions in some cases contain inaccuracies that translate to altitude translations of up to 1 atmospheric scale height. The absolute altitude of the airglow features should therefore be treated with caution, though relative changes (e.g. scale heights) are robust. The limb scans obtained during the periapse segment of the orbit are much less sensitive to the timing reconstruction and the reported altitudes are expected to be accurate. *********** Observational Constraints *********** The IUVS instrument is occasionally constrained from running certain commands due to observational constraints, e.g. the Sun being too close to its field of view. Subtle timing discrepancies in predicting when these constrains occur can result in unexpected commanding behavior for scheduled observations. For example, the echelle grating may not be commanded out of the optical path, the MCP voltage may remain off or at a low default value, or the scan mirror may remain in a stowed position. Unfortunately, this behavior is not reflected in the science telemetry stored in the current FITS files, as this is derived from commanded values, not those actually executed by the hardware or measured by sensors. An automated process is being developed to identify and correct this issue. Below is a list of orbits with observations known to be affected by at least one of the above commanding issues. The list is grouped by orbit segment, and each line is indicates a range of orbit numbers with commanding issues. For an isolated problematic orbit the start and end orbits will be the same. This list is not necessarily complete. Periapse: 346 348 356 360 368 368 449 449 458 526 566 566 624 624 690 690 808 808 828 828 945 950 955 970 1216 1232 1244 1244 1295 1295 1554 1556 1581 1582 1656 1656 1670 1670 1855 1858 1890 1926 1959 1960 1993 1994 2029 2030 2194 2194 2230 2230 2287 2287 2503 2503 2520 2520 2540 2554 2558 2574 2594 2594 2892 2892 3201 3201 3205 3205 3210 3210 3320 3323 3518 3518 3600 3600 3611 3611 3634 3666 3669 3681 3687 3687 3691 3691 3709 3709 3711 3715 3717 3717 3719 3719 3721 3725 3731 3732 3962 3962 3964 3964 3980 3980 3998 3998 4683 4683 4840 4840 Outbound: 346 346 356 358 464 464 512 516 536 572 604 620 672 672 700 700 730 748 768 780 816 886 948 968 1218 1232 1274 1274 1710 1710 1818 1818 1890 1926 2024 2028 2194 2194 2230 2230 2542 2542 2558 2558 2612 2612 2728 2736 2940 2940 3557 3557 3559 3563 3565 3569 4840 4840 Apoapse: 337 337 345 347 355 359 367 367 419 419 429 429 464 465 509 509 512 517 534 535 569 569 623 623 688 689 806 807 826 827 945 950 955 969 1214 1215 1218 1233 1244 1245 1273 1273 1295 1295 1325 1327 1346 1348 1350 1350 1362 1365 1384 1385 1401 1402 1422 1423 1581 1581 1710 1711 1854 1857 1890 1925 1959 1959 1992 1993 2028 2029 2405 2409 2692 2699 2891 2891 2893 2893 3009 3009 3043 3043 3074 3074 3458 3458 3496 3517 3521 3521 3610 3610 3632 3632 3687 3688 5445 5446 5465 5465 5481 5482 5500 5501 5519 5519 Inbound: 248 248 346 346 356 361 366 366 434 434 464 464 508 512 568 568 948 968 1214 1232 1468 1468 1580 1580 1710 1710 1752 1752 1854 1854 1890 1924 2028 2028 2764 2764 2780 2780 3458 3458 ************** Special Observations ************* During the nominal science mission, IUVS has occasionally carried out special calibration activities and opportunistic scientific observations. For the most part, these currently lack any special identification in the file names. Below are a list of these observations with segment name and orbit identified. Not all experiments were successful. Stray Light Experiments: outbound-orbit00676 apoapse-orbit00676 inbound-orbit00676 apoapse-orbit01293 apoapse-orbit01795 Stellar Occultation of Opportunity: outbound-orbit00516 outbound-orbit00700 Shutter Experiments: inbound-orbit00759 apoapse-orbit00823 inbound-orbit00898 inbound-orbit01585 Star Raster: apoapse-orbit01236 apoapse-orbit01597 IPH Echelle Experiment: apoapse-orbit01274 apoapse-orbit01992 apoapse-orbit02692 to apoapse-orbit02698 apoapse-orbit05042 to apoapse-orbit05055 Solar Occultation Experiment: periapse-orbit02503 periapse-orbit02540 to periapse-orbit02574 inbound-orbit05060 to inbound-orbit05094 inbound-orbit05114 to inbound-orbit05304 Stray Light/Detector Background Experiment: inbound-orbit04252 to inbound-orbit04286 Airglow Slit Stellar Calibration: inspace-orbit04876 inspace-orbit04886 inspace-orbit04896 inspace-orbit04906 inspace-orbit04912 to inspace-orbit04946 star-orbit05170 to star-orbit05290 star-orbit05306 to star-orbit05396 ********* Occultation Campaigns ********* Not all stellar occultation events are suitable for retrievals. This is due to various issues, such as unanticipated observational constraints or contamination from stray sunlight. The table below enumerates the useful events from each occultation campaign. Campaign (orbit) FUV & MUV only FUV only MUV #1 (orbit 009xx) 3 6 - #2 (orbit 012xx) - - 25 #3 (orbit 016xx) 4 14 - #4 (orbit 019xx) - - - #5 (orbit 021xx) 12 25 - #6 (orbit 025xx) - 5 - #7 (orbit 028xx) 15 - - #8 (orbit 032xx) - 10 - #9 (orbit 034xx) 5 - - #10 (orbit 038xx) - 5 - #11 (orbit 041xx) 40 - - #12 (orbit 044xx) 10 - - #13 (orbit 048xx) 30 - - #14 (orbit N/A) N/A N/A N/A #15 (orbit 052xx) 27 - - #16 (orbit 053xx) 36 - - ********* L1C and L2 Product Limitations ******** Level 2 contains derived data products with physical values retrieved from radiances in L1C. Many retrievals are limited to particular illumination conditions or altitude coverage. L1C limb scan brightnesses are currently limited to dayside observations (SZA < 80), and do not attempt to fit nitric oxide night glow. L2 limb scan retrievals utilize neutral atmospheric structure from the Mars Climate Database for initialization purposes. It has been found that the retrieved temperature of the limb scans has high sensitivity to the details of this initialization. More robust methods of retrieving the temperature are in development. L1C and L2 coronal observations are not processed during seasons with small phase angles due to solar contamination and off-nominal pointing. Orbits with “inward” looking coronal scans on both the outbound and inbound side segments are not handled by the current pipeline. L1C disk scans currently have FUV brightnesses set to NaN when the spectral binning is 36, 43, 72, or 96, as these have coarse binning which is not properly handled by the reduction pipeline. Most data uses 1024, 512, 256, 200, and 184 bins. L2 limb scan retrievals are limited to solar zenith angles < 60 deg. When the spacecraft uses a Fly-Z orientation at periapse (Science Scenarios 2b and 4b) the APP does not perform the “nod” necessary for IUVS to capture complete profiles of the airglow layer, and retrievals are not attempted. This occurred during the following periods: 2015-Aug-25 to 2015-Aug-31 2015-Sep-15 to 2015-Oct-12 L2 nadir viewing disk retrievals are limited to solar zenith angles < 80 degrees. The current look-up table used for the retrieval is also limited in the range of solar EUV fluxes it has indexed. As result, some orbits in the Nov 15 2015 - Feb 15 2016 time range when Mars was near apoapse are not current processed to L2. L2 nadir viewing disk retrievals are limited to solar zenith angles < 80 degrees. For v07 L1C disk products pixels near the limb which do not actually intersect the disk are geographically mapped using the geometry beneath the tangent point of the line-of-sight. This can introduce artifacts in the latitude/longitude maps at high emission angle. Also note that all geometry for disk observations is reported for a surface intercept, which introduces artifacts at high emission angles for airglow occurring at altitude. L1C and L2 disk scans are not produced for high spatial resolution observations (beginning in 2016). Validation of pipeline processing for these products is ongoing. **************** L2 Periapse FUV **************** The following caveats apply to L2 Periapse FUV retrievals: 2202 very slight striping in CO 4PG intensities 2206 striping in all intensities 2213 striping in all intensities 2224 very slight striping most apparent in vv=0/vv>0 intensity ratio 2240 slight striping in all intensities 2360 striping in all intensities 2399 slight striping in all intensities 2400 striping apparent in all intensities 2404 striping apparent in all intensities 2418 striping in all intensities 2487 striping most apparent in CO/CO2 intensity ratio 2551 very slight striping most apparent in error of CO/CO2 density ratio or 1356/CO4PG vv>0 intensity ratio 2552 very slight striping most apparent in error of CO/CO2 density ratio or 1356/CO4PG vv>0 intensity ratio 2553 very slight striping most apparent in error of CO/CO2 density ratio or 1356/CO4PG vv>0 intensity ratio 2559 very slight striping most apparent in error of CO/CO2 density ratio or 1356/CO4PG vv>0 intensity ratio 2560 slight striping most apparent in error of CO/CO2 density ratio or 1356/CO4PG vv>0 intensity ratio ****************** Other Issues ***************** A geometry bug causes the solar zenith angle (incidence angle) to be 0 deg for the first integration of echelle scans.