Diving data loop closures altering depths
Graham Mullan
graham.mullan at coly.org.uk
Tue Jun 18 10:40:19 BST 2013
Hi Footleg
Serious point, though I am quite certain the answer won't be '9 metres':
Over the length of the Keld Head sumps, there will surely be a disparity
in height AOD caused by both the need to make a geodesic correction and
for the effect of friction backing up the water within the sump.
On the substantive point, yes, I agree that the loop closure algorithm
won't handle this point properly. However, the only direct experience
that I have of this was in the Gough's survey where we simply declared
that the sump data was of a lower grade than the overland data. The
latter had already closed sufficiently well to make the water level at
upstream and downstream ends of the sump the same. That won't apply in
this case, though a loop closure from Keld Head via Valley entrance to
the upstream end of the sump might be instructive
However, given that the points in para one imply that the depths may not
be geometrically identical anyway, an algorithm that assumes that they
are would also not be the solution in long linear caves. It is difficult
to envisage a loop closure algorithm that would be different for
different data sets.
I suspect that the answer to your conundrum will involve the word
'fudge' I'm afraid.
Graham
--
graham.mullan at coly.org.uk
0117 9502556 or 07887 637064 (or 0844 2327803)
On 18/06/2013 10:50, Footleg wrote:
> Hi Graham,
>
> Originally we thought the problem was being caused by the upstream end
> of Keld Head sumps being connected to the Valley Entrance survey at a
> different sump altitude than the resurgence GPS fix. Someone noticed
> that the current end of the Marble Steps Branch surveyed last week was
> coming out 9m above water level at the resurgence. So I took the X,Y
> coordinates of the sump tie in station in Kingsdale Master Cave and
> fixed it at the same plan position by adjusted the depth to the
> resurgence depth. But that did not fix the issue. So then I started
> looking for where the depths drifted off the survey data depths. This
> led me to notice that a few equates in the data were between different
> depths, so I adjusted them to see if that fixed the issue. Eventually I
> cut down the data to just this one loop which demonstrated that the loop
> closure between two points at the same depth caused the depths of other
> points to be adjusted.
>
> I can see that a geometric loop adjustment might not treat vertical and
> horizontal components independently. But in the case of diving data
> where the depths are absolute that would not be an appropriate
> algorithm. So perhaps Survex needs a new loop closure method for diving
> legs?
>
> Footleg
>
>
> On 18 June 2013 07:10, Graham Mullan <graham.mullan at coly.org.uk
> <mailto:graham.mullan at coly.org.uk>> wrote:
>
> Hum.
>
> I'd forgotten how little info the survex log file gives you. Suggest
> converting to therion format, shouldn't take long, running in that &
> looking at the log file which gives better stats.
>
> I admit it doesn't seem logical that it should invent a vertical
> misclosure where none existed before, but as closing the loop does alter
> the whole geometry of every leg & not just the two horizontal vectors,
> it is clearly feasible that this is mathematically the simplest way of
> bridging the gap. The engine won't 'know' your depths are absolute, only
> that this is one way of expressing the geometry.
>
> I note that the data already contains a couple of points where you've
> altered depths to match. What happens if you put back the original data
> as collected?
>
> Graham
> --
> graham.mullan at coly.org.uk <mailto:graham.mullan at coly.org.uk>
>
> 0117 9502556 or 07887 637064 (or 0844 2327803)
>
> On 17/06/2013 17:39, Footleg wrote:
> > OK, so this list strips off attachments. I guess it was set up to
> run on a
> > 386 CPU like Survex still aims to support ;)
> >
> > Both ends of the loop are at the same depth before the loop is
> closed. The
> > equate which closes the loop is commented out (line 4). All data
> is diving
> > format.
> >
> > My data, in plain text:
> > ------------------------------------------------------
> >
> > *begin Keld_Head
> >
> > *equate Entrance.029 EntranceLoopNorth.001
> > ;*equate Entrance.010 EntranceLoopNorth.019 ;Closing this loop
> alters the
> > depths of stations!
> > *equate Entrance.029 EasterAven.001
> > ;*equate Entrance.001 kh12.017
> > *equate EntranceLoopNorth.015 kh12.001
> > *equate kh12.012 kh16.001
> > ;*equate EntranceLoopNorth.001 kh16.032
> >
> >
> > *begin Entrance
> > *export 001 010 019 029 062
> >
> > ;Cave: Keld Head Entrance series-From entrance peg belay (0.45m
> above
> > water level) going ; upstream in main passage, to 305.83m from base in
> > Kingsdale Passage.
> > ;Surveyors: JNC
> > ;Date: 24/05/97
> > *date 1997.05.24
> > ;vis = 1.5m
> >
> > *fix 001 69573.00 76576.00 0.45 ;temp fix of depth as water depth
> 260.0
> > ;GPS
> > *entrance 001
> >
> > *calibrate compass 4.83 ;was +5.17
> >
> > *data diving from to bearing length fromdepth todepth
> >
> > 001 002 237.0 05.10 00.45 -02.0 ;st001 = peg belay for diving line
> at Keld
> > Head ent. pool
> > 002 003 246.0 04.50 -02.0 -01.5
> > 003 004 266.0 06.00 -01.5 -03.5
> > 004 005 269.0 01.88 -03.5 -03.2
> > 005 006 271.0 06.40 -03.2 -03.8
> > 006 007 274.0 08.36 -03.8 -01.3
> > 007 008 264.0 05.30 -01.3 -01.3
> > 008 009 274.0 01.13 -01.3 -01.3 ;st009 = joint above 1st airbell
> > 009 010 297.0 02.92 -01.3 -00.9 ;st010 = line junc. main orange to R;
> > centre blue = ahead;(this survey); ascending white plastic
> (airbell?) to L
> > 010 011 276.0 01.60 -00.9 -01.1
> > 011 012 314.0 02.17 -01.1 -02.4
> > 012 013 303.0 06.75 -02.4 -03.6
> > 013 014 282.0 06.32 -03.6 -04.0
> > 014 015 299.0 02.80 -04.0 -04.6
> > 015 016 303.0 06.50 -04.6 -04.6
> > 016 017 293.0 03.07 -04.6 -04.6
> > 017 018 317.0 05.43 -04.6 -05.2
> > 018 019 356.0 00.68 -05.2 -05.3 ;00.30m before st019 = white line
> off on L
> > to Woodings Pass.
> > 019 020 353.0 02.10 -05.3 -05.2
> > 020 021 326.0 03.78 -05.2 -05.3
> > 021 022 333.0 02.72 -05.3 -05.2
> > 022 023 357.0 02.95 -05.2 -04.7
> > 023 024 027.0 01.66 -04.7 -04.3
> > 024 025 038.0 02.28 -04.30 -03.5
> > 025 026 045.0 04.57 -03.5 -02.6
> > 026 027 035.0 04.56 -02.6 -02.8
> > 027 028 015.5 01.70 -02.8 -03.4
> > 028 029 051.0 02.47 -03.4 -03.7 ;st029 = line junction: white,
> up(="100m"
> > airbell) orange, right, "out" orange, left (=main survey)=Kingsdale
> > Passage. straight on, yellow (=Crossland's ?)
> > 029 030 338.0 06.14 -03.7 -05.2
> > 030 031 292.0 10.48 -05.2 -05.9
> > 031 032 277.0 11.83 -05.9 -05.7
> > 032 033 287.0 08.58 -05.7 -05.9
> > 033 034 305.0 17.25 -05.9 -05.8
> > 034 035 302.0 06.15 -05.8 -05.8
> > 035 036 286.0 11.92 -05.8 -05.8
> > 036 037 317.0 04.24 -05.8 -05.8
> > 037 038 333.0 03.50 -05.8 -05.9
> > 038 039 354.0 05.40 -05.9 -06.1
> > 039 040 340.0 02.35 -06.1 -06.0
> > 040 041 323.0 04.80 -06.0 -05.9
> > 041 042 315.0 03.10 -05.9 -05.8
> > 042 043 298.0 03.35 -05.8 -05.6
> > 043 044 296.0 04.97 -05.6 -06.0
> > 044 045 259.0 07.98 -06.0 -05.8 ;st045=line clove hitched around
> flat slab
> > 045 046 268.0 08.83 -05.8 -05.9
> > 046 047 268.0 04.28 -05.9 -04.8 ;st047=boulders, white prong
> > 047 048 316.0 02.00 -04.8 -04.5
> > 048 049 314.0 05.03 -04.5 -05.4
> > 049 050 335.0 01.08 -05.4 -05.4
> > 050 051 350.0 01.28 -05.4 -05.4
> > 051 052 315.0 04.42 -05.4 -05.4 ;st052=thin white string jump line
> up (to
> > airbell ?), eyehole belay in boulder
> > 052 053 318.0 04.15 -05.4 -04.8
> > 053 054 290.0 02.83 -04.8 -05.0
> > 054 055 298.0 05.20 -05.0 -04.3 ;frogspawn !!!
> > 055 056 301.0 02.00 -04.3 -04.5
> > 056 057 306.0 07.36 -04.5 -04.6
> > 057 058 313.0 05.75 -04.6 -06.5 ;st058=start of CaCo3 veins
> > 058 059 310.0 08.18 -06.5 -06.8 ;st059=end of CaCo3 veins
> > 059 060 316.0 09.20 -06.8 -06.2
> > 060 061 306.0 07.98 -06.2 -07.2
> > 061 062 305.0 08.52 -07.2 -06.9
> >
> > *end Entrance
> >
> > *begin EntranceLoopNorth
> > *export 001 015 019
> >
> > ;Cave: Keld Head- North side of Entrance Loop- starting at five
> way juntion
> > below 100m airbell 105.2m from base kh01 st.029 and rejoining main
> line
> > 41.59m from base at kh01 st.010
> > ;Surveyors: JNC
> > ;Date: 27/05/98
> > *date 1998.05.27
> > ;vis = ???m
> >
> > *calibrate compass 4.63 ;was +5.45
> >
> > *data diving from to bearing length fromdepth todepth
> >
> > 001 002 109.0 01.20 -03.70 -04.10
> > 002 003 112.0 01.34 -04.10 -04.00
> > 003 004 138.5 03.42 -04.00 -05.50
> > 004 005 095.0 02.41 -05.50 -05.70
> > 005 006 093.0 01.23 -05.70 -06.60
> > 006 007 136.0 00.97 -06.60 -06.70
> > 007 008 165.0 02.87 -06.70 -06.50
> > 008 009 146.0 06.17 -06.50 -06.20
> > 009 010 178.0 03.90 -06.20 -05.40
> > 010 011 130.0 06.27 -05.40 -05.80
> > 011 012 127.0 02.75 -05.80 -05.50
> > 012 013 140.0 02.53 -05.50 -04.70
> > 013 014 139.5 05.62 -04.70 -04.40
> > 014 015 145.0 01.83 -04.40 -04.40
> > 015 016 185.0 00.34 -04.40 -04.40
> > 016 017 213.0 07.38 -04.40 -02.20
> > 017 018 178.0 01.87 -02.20 -01.20
> > 018 019 170.0 01.50 -01.20 -0.9 ;Original depth was -01.10, but
> changed
> > to match depth of tie in station on main entrance line. Footleg
> > 17-Jun-2013 ;st.019=kh01
> > st.010
> >
> > *end EntranceLoopNorth
> >
> > *begin EasterAven
> > *export 001
> >
> > ;Cave: Keld Head - 100m airbell and Easter Aven
> > ;Surveyors: JNC
> > ;Date: 17/04/03
> > *date 2003.04.17
> > ;Instruments: JNC standard
> >
> > *calibrate compass 3.82 ;was +4.29
> >
> > *data diving from to bearing length fromdepth todepth
> >
> > 001 002 054.0 05.60 -03.7 00.0
> > 002 003 007.0 03.50 00.0 00.0
> > 003 004 028.0 06.00 00.0 00.0
> > 004 005 070.0 01.50 00.0 00.0
> > 005 006 320.0 03.00 00.0 00.0
> > 006 007 320.0 05.00 00.0 00.0
> >
> > *data normal from to length bearing gradient
> >
> > 007 008 05.00 - UP
> >
> > *end EasterAven
> >
> >
> > ;Cave: Keld Head - The Open Road and Part of Crassland's Passage
> - starts
> > from line juction at end of leg 14 along
> > ; north east side of entrance loop
> > ;Surveyors: JNC
> > ;Date: 05/04/00
> > ;Instruments: JNC
> >
> > *begin kh12
> >
> > *export 001 012 017
> >
> > *calibrate compass +5.17
> > *data diving from to bearing length fromdepth todepth
> > 001 002 063.0 13.36 -04.40 -04.20
> > 002 003 036.0 02.62 -04.20 -03.30
> > 003 004 068.0 01.31 -03.30 -02.90
> > 004 005 074.0 02.66 -02.90 -02.80
> > 005 006 052.0 02.75 -02.80 -02.60
> > 006 007 063.0 02.50 -02.60 -02.40
> > 007 008 072.0 00.85 -02.40 -02.50
> > 008 009 084.0 02.48 -02.50 -03.50
> > 009 010 117.0 04.09 -03.50 -03.60
> > 010 011 119.0 07.17 -03.60 -04.40
> > 011 012 098.0 05.78 -04.40 -04.70
> >
> > ;st012=line junction at end of Open Road with Crossland's Passage
> - this
> > continues further into Crossland's pass. to the left ( on 39 deg.
> mag.) and
> > leads back to the entrance belay to the right. The rest of this
> survey
> > goes to the right (ie. back to the entrance)
> >
> > 012 013 183.0 04.56 -04.70 -03.50
> > 013 014 225.0 03.62 -03.50 -03.30
> > 014 015 205.0 01.78 -03.30 -01.70
> > 015 016 131.0 01.18 -01.70 -00.90
> > 016 017 212.0 03.55 -00.90 +00.45 ;st017=peg belay at entrance.
> >
> > *end kh12
> >
> > ;Cave: Keld Head - Crosslands Passage
> > ;Surveyors: John Cordingley
> > ;Date: 27/02/03 & 29/03/03
> > ;Instruments: JNC
> >
> > *begin kh16
> >
> > *export 001 032
> >
> > *calibrate compass +5.17
> >
> > *data diving from to bearing length fromdepth todepth
> >
> > 001 002 035.0 17.40 -04.70 -05.00 ; 001 depth in notes was -3.7,
> but tie in
> > station is at -4.7, so changed to match. Footleg 17-Jun-2013
> > 002 003 043.5 03.20 -05.00 -05.30
> > 003 004 047.0 10.70 -05.30 -05.90
> > 004 005 030.0 03.90 -05.90 -05.80
> > 005 006 359.0 10.05 -05.80 -06.40
> > 006 007 028.0 07.60 -06.40 -06.60
> > 007 008 044.0 05.10 -06.60 -06.80
> > 008 009 027.0 08.90 -06.80 -07.40
> > 009 010 004.0 03.40 -07.40 -07.40
> > 010 011 261.0 15.50 -07.40 -07.30
> > 011 012 286.0 12.40 -07.30 -07.00
> > 012 013 330.0 05.00 -07.00 -07.30
> > 013 014 343.0 01.50 -07.30 -06.10
> > 014 015 223.0 01.50 -06.10 -06.00
> > 015 016 273.0 04.80 -06.00 -05.50
> > 016 017 226.0 04.45 -05.50 -05.50
> > 017 018 191.0 05.60 -05.50 -05.20
> > 018 019 272.0 04.85 -05.20 -06.60
> > 019 020 267.0 04.10 -06.60 -07.40
> > 020 021 220.0 04.15 -07.40 -07.80
> > 021 022 206.0 19.80 -07.80 -07.00
> > 022 023 249.0 06.80 -07.00 -07.00
> > 023 024 244.0 07.55 -07.00 -06.80
> > 024 025 274.0 06.62 -06.80 -06.80
> > 025 026 300.0 04.82 -06.80 -06.90
> > 026 027 315.0 03.82 -06.90 -06.20
> > 027 028 279.0 06.80 -06.20 -03.50
> > 028 029 237.0 03.80 -03.50 -02.90
> > 029 030 196.0 07.80 -02.90 -03.90
> > 030 031 242.0 05.00 -03.90 -04.20
> > 031 032 221.0 01.35 -04.20 -3.7 ; 032 depth in notes was -4.20,
> but tie
> > in station is at -3.7, so changed to match. Footleg 17-Jun-2013
> >
> > *end kh16
> >
> > *end Keld_Head
> >
>
>
More information about the Survex
mailing list