20 Jun 2001 Dear Colleague: The purpose of this message it to bring to your attention certain aspects of the issues raised in Randy Kirk's email, specifically related to the decision to change the Mars cartographic coordinate system to east-longitude, aerocentric coordinates. Summary of USGS Position: the plan is to junk more than three decades of cartographic products and experience using a surface-oriented planetographic coordinate system with the subsolar location longitude increasing with time (west longitude) for a body-centered planetocentric coordinate system based on a right-hand biased, spherical coordinate mathematical construct, with positive east longitude. The first paragraph under "Coordinate System" in Randy's message describes the differences. Summary of Mars Orbiter Camera position: this change is unnecessary and counterproductive. No compelling argument can be made in favor of the change, and it is counter to three decades of Mars cartography. Because of the wording of Randy's message (he will implement the change "...unless we receive an overwhelming response to this email indicating that a majority of the Mars research community favors retaining the [present] system..."), I STRONGLY encourage you to communicate your views directly to Randy Kirk at rkirk@usgs.gov with a copy to Joe Boyce at jboyce@mail.hq.nasa.gov What follows is a more detailed treatment of specific items noted in Randy's message. 1) Randy comments on the adoption of the east/planetographic system by the MGS MOLA and TES and MS'01 THEMIS (and possibly GRS) for their data products. The implication is that this somehow validates the choice. The MOLA team presented their arguments to the Mars Observer Project Science Group in the late 1980s. After discussion, the PSG overwhelmingly voted to stick to the existing planetographic, west-longitude system. The reasons, as will be noted below, were that the west/planetographic system had substantial heritage, better approximated the true location of features on the surface, and that neither system had REAL advantages in computing over the other, thus reducing the argument for change. The MOLA team stated that they would use planetocentric, east-longitude regardless of the PSG's decision, and in spite of Project requirements as stated in the Planetary Constants document with which each experiment was required to be consistent. After the loss of MO, the MOLA team approached the Mars Global Surveyor Project with the same proposal. It was again discussed and rejected. MOLA restated their plan to use planetocentric east longitude. The USGS fails to note that the MOC uses a planetographic/west longitude system in all of its processing. The MS'98 MARCI wide and medium angle cameras would have used this coordinate system, and the MRO'05 MARCI WA and Context cameras will use this coordinate system. Perhaps the difference is that MOLA and TES provide access to their data only through coordinates, and so people using them must deal with whatever system those teams chose, while images are images "despite" the coordinate system. In any case, the community should be aware that there was considerable discussion of this topic by the Mars Observer and MGS PSGs and that both rejected the coordinate change. 2) Randy states that other groups the USGS has contacted, including the NASA Geological Mapping Subcommittee (GEMS) and the ESA Mars Express High Resolution Stereo Camera (HRSC) team have expressed a willingness to follow the lead of the USGS and adopt our choice of coordinate system for their future products. This is not surprising, given that either coordinate system is relatively simple to implement, and that the USGS is considered by most to be responsible for taking the lead in establishing and maintaining cartographic standards for Mars within the US. However, it should not, I believe, be taken as indicating that these other groups have ENDORSED the change. Rather, that they'll go along with it. What else could they say? 3) The email notes that the planetary research community will always have to deal with the two competing systems. It states "In this respect the situation is very different than before the MGS mission when only the west/'ographic system was in use in the literature and opposition to the use of east longitudes was vehement (though ultimately ineffectual)." The community need not be held hostage by a small group of investigators who contend that their way is the best way and that everyone MUST follow their lead. Randy continues "If the USGS retains the west/'ographic system for future maps, these will not be compatible with products generated by the MOLA, TES, and Odyssey teams. If future USGS maps use the east/'ocentric system for compatibility with these teams, they will be incompatible with the bulk of old map products from the USGS and elsewhere." It is trivial to translate MOLA, TES, and Odyssey team products to planetographic west longitude. I do it routinely for both MOLA and TES data. I cannot speak for my colleagues on those teams, nor on the Odyssey team, but if they want to work in east longitude that's fine with me. I don't see why I should force them to use a system they don't like (although it will make communication more difficult), nor do I see why the rest of us have to change because they think they know best. One could ask why the MOLA team, if it were interested in communicating with the rest of us, insist on reporting their results in their own coordinate system rather than conforming to agreed-upon and commonly practiced standards. In any case, since the MOLA and TES data sets are relatively small, already digital, and will (eventually) be complete, transforming them to adhere to the planetographic west longitude system so that they are compatible with previous maps seems a better solution than striking off in a new direction and thus reducing the utility of exisiting products. 4) Technical merits - Randy states that the east longitude/planetocentric system has slight advantages over west/planetographic, and cites its use in the reduction of "space remote sensing data" (the aforementioned TES, MOLA, and THEMIS/GRS data?), its use in spacecraft navigation, and the need to account for planetary figure in the latter. As already noted, translating the TES, MOLA, and THEMIS/GRS derived products to west/planetographic coordinates is relatively effortless, and how those teams choose to "reduce" their data shouldn't be forced upon the rest of the community "just because." Spacecraft navigation is in fact NOT performed in the planetocentric east longitude system, but in a mass-centered inertial cartesian coordinate system that is NOT tied to the surface. The navigation system is well understood, and translations between it and either the east/planetocentric or west/planetographic systems are routine. MOC routinely targets off-nadir locations on Mars. This is accomplished by designating targets on the MDIM2.0 map, and software translates this to the navigation system (and in fact then generates pointing "quaternions" for the spacecraft). We have incorporated the MOLA planetary radii into this software because planetary figure IS important at the scale of high-resolution imaging (see next paragraph). Perhaps the most important technical problem is the need to address planetary figure. Latitude/longitude pairs with an implied analytic figure (spheroid, ellipsoid, sphere, etc.) are simply shorthand notations for a point in cartesian three-space which is most likely not on the planet's surface. Any precision application (e.g., targeting of a high-resolution image from an orbiter, or a lander aim point) will require a triple of coordinates (be they cartesian, spherical, spheroidal, cylindrical, etc.) that unambigously define a point on the surface. Rank the following "models" by which you think best describes where the surface of Mars actually "is": a sphere, an oblate spheroid, a tri-axial ellipsoid, a high degree and order harmonic model, or the actual planetary radii. At present, we're "debating" whether Mars is better approximated by a sphere or an oblate spheroid. If you think a sphere is a better approximation, then you might favor the planetocentric system; if you think the several km difference resulting from oblateness needs to be considered, you might favor the planetographic system as presently implemented. If you think Tharsis should also be accounted for, then you would favor the triaxial ellipsoid. The CORRECT approach is to use planetographic coordinates computed with the ACTUAL topography of the planet, which we know for nearly every square kilometer on the planet. But in any case, we need to adjust for the radial location of the surface in our maps or they will not accurately indicate the location of features ON that surface. The MOLA team has in fact used a multiplicity of coordinate systems for various products: sphere, spheroid, ellipsoid, ellipsoid with the center of figure offset from center of mass, spherical harmonic expansion, etc. Radii or altitude have been variously referenced to all of these models. They have used and, as you would expect given their past history, reserve the right to continue to use, all of these models in various contexts. Thus, MOLA data will always be subject to caveat emptor. To reiterate: processing of image data requires a model of the true figure of the body to be used, regardless of the map projection used to present the results. Even if a sphere were adopted for map PROJECTIONS, software will still have to internally use a spheroid or better model for locations to be input to the projection. This requirement to use two models will inevitably lead to more, not less, confusion. The spheroid is a demonstrated acceptable approximation for the majority of global Viking coverage; a sphere is not. Thus, in the planetocentric approach, calculations using two models as described above would have to be made, even through the final product would ultimately be compromised in accuracy RELATIVE TO REALITY. From my perspective, accurate rendition of high-resolution images will in fact require better accuracy than a spheroidal model provides. The images must be orthorectified to the actual topography, as is done routinely for large map-scale terrestrial maps. Summary of the attributes of the two systems: East Planetocentric West Planetographic Inconsistent with all previous Mars Consistent with all previous Mars mapping conventions and maps mapping conventions and maps Assumes planet is a sphere for map Acknowledges that the planet is not projections a sphere and provides a simple, docu- mented representation of that figure Requires second model of planet shape Can use same model for deriving to derive actual planetary positions positions as is used to map project to then be mapped to a sphere Requires software modules to be Requires no software to be changed rewritten I hope you agree that the proposed change is not in the best interest of planetary cartography and make your view known to Randy and Joe Boyce. Respectfully, Mike Malin