1993 ALVIN DIVES (CoAxial event response cruise - Delaney/Embley chief sci.) Updated notes 10/14/94: The 1993 Alvin nav is getting quite convoluted (...what's new, right?). Originally we got all our nav from Russ McDuff in the form of AlvinX/Y files. However, he only "finished" the FLOW-net files (including slight shifting of the transponder positions), but we only had his "preliminary" files for FLOC and SOURCE dives (which only had 1 fix per minute). At first we converted all these from AlvinX/Y to UTMX/Y and then to Lat/Lng. These were the first generation of files: FLOW SITE: [NAV.NAV93.ALVIN.2670]2670.LAT (FLOW - Embley/Lilley) [NAV.NAV93.ALVIN.2671]2671.LAT (FLOW - Chadwick/VanDover) [NAV.NAV93.ALVIN.2672]2672.LAT (FLOW - Tivey/PIT) [NAV.NAV93.ALVIN.2677]2677.LAT (FLOW - Johnson/McDuff) FLOC SITE: [NAV.NAV93.ALVIN.2673]2673.LAT (FLOC - Koski/Lilley) [NAV.NAV93.ALVIN.2676]2676.LAT (FLOC - Chadwick/Lilley) [NAV.NAV93.ALVIN.2679]2679.LAT (FLOC - McDuff/PIT) SOURCE SITE: [NAV.NAV93.ALVIN.2674]2674.LAT (SOURCE - Embley/Delaney - ABORTED) [NAV.NAV93.ALVIN.2675]2675.LAT (SOURCE - Embley/Delaney) [NAV.NAV93.ALVIN.2681]2681.LAT (SOURCE - Delaney/Broad) The following dives had major problems with the in-sub navigation It MIGHT be possible to get some ACNAV navigation for them later: 2678 (SOURCE - Delaney/Kelley - nav box was down most of dive) 2680 (SOURCE - Embley/Koski - out of range of all but one xp most of dive) In addition, there were major problems with the SOURCE files - the preliminary nav for dives 2674, 2675, and 2681 was obviously misplaced relative to the bathymetry, so I applied a rigid shift to all of them, based on the fact that dive 2675 visited the RIDGE expendable transponder "X" at the beginning of the dive. I shifted the nav position at that time (46 09.554'/-129 48.064'), to the calibrated position of the transponder (46 09.489'/-129 47.940'), which amounted to a shift of -0.065'lat and +0.124'long. Ok, those were the OLD files. After we did all that we decided that converting from AlvinX/Y -> UTMX/Y -> Lat/Lng introduced an unnecessary error and that we really should just go straight from AlvinX/Y -> Lat/Lng, using WHOI's "makelat" or "xy2ll" UNIX programs. So, I just did all that. In addition, while on the AII in 1994, we reprocessed navigation ourselves for dives 2673, 2676, 2679, and 2681 (so we now have every fix, instead of just 1 per minute for these dives). Some notes: Dives 2674 and 2675 still only have 1 fix/minute because we have never processed these ourselves. All the others have all the available fixes. Dives 2678 and 2680 are still in limbo. The conversion from UTM-lat/lng to Alvin-lat/lng for the FLOW site dives introduced a small error (< 1 m), because I started from the lat/lng files and their precision is only that good. For the other sites I worked from AlvinX/Y data (recorded to 0.1 m), so they don't have that problem. This whole process moved lat/lng fixes a variable amount, but about -30m in y and +40m in x. What next? More??? Oh yes!!! These nav files now more accurately reflected the 1993 Alvin dives based on the AII transponder calibrations. However, these aren't very good as we know, and there are substantial offsets between 1994 and 1993 dives that visited the same markers. SOOOOO ... I did one more thing. I applied a rigid shift to all the dives - a different shift for each site. Here are the shifts for each site: FLOW SITE: 93->94 shift = +15m in X/-9 m in Y based on the average difference between fixes at markers 0, 7A, and 6A on dives 2670 (93: the McDuff replay, not the original nav!) and 2788 (94). (the 93->94 shift from the original 1993 nav = +54m in X/-67 m in Y) FLOC SITE: 93->94 shift = +8m in X/-11 m in Y based on the average difference between fixes at markers 3A, and 11 on dives 2676 (93) and 2789 (94) SOURCE SITE: 93->94 shift = +27m in X/-56 m in Y based on the average difference between fixes at Beard and Church vents on dives 2681 (93) and 2790 (94) All the 1993 dives were put in the same reference frame as the 1994 dives. SO FINALLY ... here are the NEW files: FLOW SITE: [NAV.NAV93.FINAL]ALV2670.GPS (FLOW - Embley/Lilley) [NAV.NAV93.FINAL]ALV2671.GPS (FLOW - Chadwick/VanDover) [NAV.NAV93.FINAL]ALV2672.GPS (FLOW - Tivey/PIT) [NAV.NAV93.FINAL]ALV2677.GPS (FLOW - Johnson/McDuff) FLOC SITE: [NAV.NAV93.FINAL]ALV2673.GPS (FLOC - Koski/Lilley) [NAV.NAV93.FINAL]ALV2676.GPS (FLOC - Chadwick/Lilley) [NAV.NAV93.FINAL]ALV2679.GPS (FLOC - McDuff/PIT) SOURCE SITE: [NAV.NAV93.FINAL]ALV2674.GPS (SOURCE - Embley/Delaney - ABORTED) [NAV.NAV93.FINAL]ALV2675.GPS (SOURCE - Embley/Delaney) [NAV.NAV93.FINAL]ALV2681.GPS (SOURCE - Delaney/Broad) For now, there are 2 versions - the new version is the shifted file; the old version is the unshifted file. The morals of the story? 1) Don't let anyone else process your navigation - they never finish it and it just gets too confusing; 2) Alvin calibrations from year-to-year don't take you back to the same spot - navigated X/Y positions can be off typically 30-40 m, but as much as 100 m. This is beyond GONZO, isn't it? It's MONDO-GONZO! -Bill Chadwick (chadwick@new.pmel.noaa.gov) ------------------------------------------------------------------------- Update 8/2/95 After the 1995 Alvin dive series, it was decided to shift both the 1993 and the 1994 Alvin dive navigation to be compatible with the new 1995 navigation at the FLOC SITE (where 2 of Paul Johnson's expendible transponders were deployed and will be there for ~5 years) AND for the SOURCE site. Since the 1993 nav had previously been shifted to match the 1994, they both needed the same shift to match the 1995 nav. -------- The FLOC shift was determined by comparing nav fixes from 1994 and 1995 at marker locations: Marker 1994 nav 1995 nav difference (94->95) 11 6756/7218 6749/7194 -7x/-24y 3A 6460/6710 6463/6684 +3x/-26y 2 6369/6440 6372/6424 +3x/-16y 10 6207/6088 6224/6066 +17x/-22y --------- average +4x/-22y So this average shift was applied to 1993 dives 2673, 2676, 2679 and 1994 dives 2789, 2791, 2793. ---------- The SOURCE shift was determined similarly: Vent/Marker 1994 nav 1995 nav difference (94->95) Beard 4379/6587 4372/6608 -7x/+21y Church/8A 4385/6610 4391/6633 +6x/+23y Mongo 4409/6658 4404/6674 -5x/+16y 1 4402/6651 4395/6669 -7x/+18y ---------- average -3x/+20y So this average shift was applied to 1993 dives 2674, 2675, 2681 and 1994 dives 2787, 2790. -Bill Chadwick (chadwick@new.pmel.noaa.gov) & Julia Getsiv (getsiv@new.pmel.noaa.gov)