l-MTED NATIONS
ECONOMIC AND SOCIAL COUNCIL
GENERAL
SCA/NRD/CART/48 5 Nbveirfcer 1986 Originals ENGLISH
BDCNCMEC COMMISSION FOR AFRICA
Sixth United NaUons Regional Cartographic
Conference for Africa
Addis Ababa, Ethiopia, 10-21 November 1986
EAST AFRICAN GBCDETIC NETWORKS - WffiT NEXT?
B.Ee F0K4STCN, DIRECTOR OF OVERSEAS SURVEYS ORCNANCE SURVEY OF GREAT BRimiN, UNITED KINOXM
BCA/lffiD/CaRP/48
B E FURMSTON, Director of Overseas Surveys Ordnance Survey of Great Britain, United Kingdom
I. The African Doppler Survey (ADOS) has received wide support from African ind donor nations and there are now a considerable number of stations in the continent with coordinates derived from Precise Ephemeris Transit satellite
fixations.
>. At the 3rd Symposium on Geodesy in Africa, held in April 1986 in Yamassoukro, several papers turned attention beyond the ADOS campaign itself to the development, 3f a datum and figure of the earth which may fit the geoid in Africa more cJnsely than those currently in use - and to the possibility cf carrying out a readjustment :>f the existing geodetic data to provide a homogeneous, and more accurate, set 3f coordinated data for existing stations of national frameworks.
CURRENT ADJUSTED VALUES
3. The British Directorate of Overseas Surveys (DOS) was responsible for a great deal of the observation, and adjustment, of the geodetic framework of East and Central Africa. No comprehensive report of the field operations was produced but details of progress are contained in DOS Annual Reports. A pape* setting out the sequence of adjustments and their results was presented to the Symposium
Geodesy m Africa' hold in Khartoum in 197*1(1) No major adjustments of data concerning East or Central Africa have been carried out by DOS or, it is beli-ved by any other organisation, since that time.
«• „ t"he AfriCan adjustments of DOS are certainly capable of improvement.
Windsor s paper points out sowe of the constraints and comments:-
"Were these adjustments to be undertaken today (i.e.197*0 with the large electronic computers available, they would be done very differently. The results overall would be better, particularly if the secondary breakdowns between the
mam networks were included in the overall adjustment".(1)
The major constraint, which was particularly ssvere in the earliest work, wa*
the sheer voxume of tin* consuming effort in large adjustments vhich forced work to be broken down into stages. It is difficult now to recall, o- to imagine, the application and determination necessary to compute, with a hand operated mechanical calculator, the long strings of cross-multiplications necessary to enter correctly each figure in a desk sized matrix of squared paper. There werr often four or wore computers working -ound one -.uch sheet and there was certainly no question of repeating trial solutions in search of an optimum solution. Even after electronic computers were in use the limited memory and processing speed
of those available to DOS in the I96o's still limited the size'and comity of
Lh r w obsen'atl0nS and the vulnerability o.f their tap., readers sometimes
led to the repetition of many hours of work.
DCA/NRD/CART/48
?age 2
. The result, »f course, was the -Q-Uai ad3ust»ent »f £
.riangulation (later incorporating ^"^^^^a -preliminary work on
.hrough the 30th Arc, Uganda, Kenya and Norther" ^"^ J to m very,he remaining networks in Tanzan^ ^o«ed that ^ey «ere no g g disappointing ..
comfortably into the space that had been left ^e^d^tfflents of compiex
£■£. Le^vefus - ^wild^o - ways L, which errors adulate
n survey networks.
..
.0 Tanzania. Host agree, in broad terms that be - value, .houl^ ^ ^Several studies have been aade of the DOS -£*££ SoulfresuTt "of^
, large simultaneous adjust.e^t but no =«^rete ^tepS ^^ ^ ^^^ Yanoussoukro
,nd. A recent paper presented to the Jra ajrapo d fi 1(J obserVatxonS>
:arries criticism a good deal further, =1»™ment model(s)"(2). Although the
;rroneous reductions and even "adequate adjustment^elW ^ ^ ^ ^^
■.tatistical tests set out apply fully only to ver> xa 6 £ calculations
confusion
^iperie^Lrff^^
corrections to observed quantities and
. Windsor explains that all the later af ^^^
iven in her paper were completed by varxat*™° measured observations ixed points held as known Thus both Uk ^^ dlstortionS caused by this
ouS reasonably be allotted there «as no other
. it is worth noting that the role * n 19*6, was to produce topographic maps
otential and to support the -Pi—at
ntirely due to be foresight of Brigadier
hat the establishment of a 5^/^/
dequate, homogeneous »»PPin*
Hotted for this purpose, ay be deficient for modern
s still adequate as a control ^ o.nmonwealth countries of East and
The scale and
hose in the framewor. of rror of the accepted values
urposes is some 1:20,000 in places
^uSTms^ fxrst Director,
Hotxne D^ ^ uisite to^^Argument that funds were
currently accepted DOS adjusted values
ourren > w ^^ adequate and
h^ogeneous .apping th.-oughout the
icaE It has als0 proved satisfactoryri ^ £caie fcy ^^ surveyors,
n errors are, * ^ here the scaie scientific
GWARDS A HEW ADJUSTMENT?
derived positions
or the computation of "ie?1"08"^' BOr"
he survey framework certaxnly exxsts It a retrograde step, however,a retrog ^ ^
Bra/NKD/CAKr/48
Page 3
networks and cadastral values could be iustified rh ^,
aim for a set of scientific valUPi tJ v! e better oourse would be <-°
Europe. Ordnance Survey adjusters o5 th^R ^T " Satisfact<"y solution in have improved scale errors to something t ^ "etWOrk (0SSN7° and 0SSN80^
values obtained from these adiusSnt^happr°aohlnS Part Per million and the needed and for projectssuch as thT **" "^ Hhere ereat accuracy is
for North Sea oL fnttlxtTtlZ ^u^i^d^T"0"*^ ^ V3lU6S
blocks, the periphery of each block LiZ?,^ f ™ WaS 0OI°Puted in
readjustments show significant etan»~ ?* °°ntr0;1 the next" Becent
Spain, Northern Norwafan^^ Italy du1 to'"f^fusl^^^pted values in Portugal,
which passed systematic errors Into the outivL Of/he 6arUer c™Put^ion?
Europe is, however, still usually based o^ the olTTl Natl°nal nappl"S ^
so. iy Dasea °" the old daturas and may well remain
:ostofa neTh^o^neius^^ °f HOTk and "» very
established by DOS, let alone tt ^'"'^f' °f £Ven those f^^works
technology it is not now ttecomputatxon ofT °f ^"^ Wlth mode- computing which is the major problem Sere are holt S °f V6ry large adjustments involved all of which wlll re™^ anaiv"^ In T* ** qUantities »f data
t
of which wlll re^ anaiv^ In T
f d hs »f data
not an exciting task, the ultimate 'f f" a"d che<*inS- Although
pains taken at this stage. ° the Wh°le ^^ Hln d*P<«d on the
r^^ ^re^ ^T
secondary work providing croL'connect ons CTn^ oT"* °f ^
measurements than the wntetir* ^ ■ yns' Ine^ often contain more EDM
later, after EDM cLe fn'oienerL'use'^TheT' TT the "°rk "aS «"««-t8ken
p^^SL^f^oiva^te^ °f °b—ions will mean
purpose, it is SU8gestJ thSgt triL adjuit^ntf^ L^ D°S obse^a«ons for this traverse between Doppler station^%!nf ?f °Uld be aade of chai"= and/or
means that the qua!??,VtL ^r"^^' "L™88"1?" Ik iS ^ ** *™ -ch
positional closure on previously f[™li stations Eeparated "•« "e bias effect of
pabl1^\L1iT^rcan:ndBurO " •» a «*- *«
concepts and the computer ^ ^sf £"• re-adJustfflent^ the technical
are still working on an adjustment, which'started ?h- ^ taSk " EuroPea" nations
considerable cost. In considerin, ju^t^tLI ^^^^^^^^
SOS - SAVE THE OLD
s^vey f^r^ IITIT TnTeT ^ ^ "^ °f the °" s^«ons of the
engaged on mapping projects have encounte^/^M WS *"* °SD SUrVe^ parti-
destruction of, both pillars and ™TstaWols ^ °f ""^^ ^^> and
ECA/NRD/GART/48 Page 4
15 In North East Tanzania, between June 1983 and April 198*, in the course of reconnaissance for the establishment of ground control for 1:50,000 mapping 82
trigs were visited.
NORTH EAST TANZANIA: JUNE 83 TO APRIL 84
TRIGS VISITED BADLY DAMAGED BUT POSSIBLY RECOVERABLE
16 or 19-5%
DESTROYED AND BEYOND RECOVERY
18 or 22.0*
TOTAL DESTROYED OR AT RISK
Over one fifth were totally destroyed and beyond recovery of any accurate position (it was sometimes possible to use the position, estimated to a metre or so, for 1:50,000 mapping purposes only). Nearly another fifth were badly damaged or missing but could, if action Is taken quickly, be rebuilt accurately in their former position from buried and/or witness marks.
16. In Kenya, in July 1985, in the course of a 10 day reconnaissance for the establishment and identification of control for the air triangulation of a block o*
photography HI trigs, were visited. Nearly 30* were destroyed; over half of the
tota/were badly damaged, destroyed or missing although these could prob*** * ^
rebuilt, accurately if witness marks still exist, sometimes to within a few centimetres even if witness marks are destroyed because of their particular state.
KENYA: SOUTH NYAKZA: JULY 1985
TRIGS VISITED BADLY DAMAGED BUT POSSIBLY RECOVERABLE
21 or 51-2*
DESTROYED AND BEYOND RECOVERY
12 or 29-5*
TOTAL DESTROYED OR AT RISK
80.7*
17 In Uganda, OSD has no concrete statistics but have been told that a great
number of survey stations of all classifications have been destroyed. It is understood
that help is being sought for their replacement.
18 Most of this loss has been caused by deliberate action although this may have been based on innocent misinterpretation and/or ignorance of the ^tion of thestations. Considerable labour and ingenuity has been used to ensure that hidden, as well as surface, marks have been destroyed. In one instance a large hole 0 metres deep had been excavated in which the remains of a 1.2 metre concrete pillar
now lie.
Page 5
cannot be recovered. If S rt fC £ ?UrVe? stations ""eh do not exist
necessary to Info™ govern^ d « 1 u T^"1 aCtiO" does 3eem to »*
for infra structure, that" are ^ t f"™67 StaU°nS are Valust>le ite^
their destruction inevitably leads told r-LTl^ "* tO COOTdin^e and that Government needs to be convinced of th^hPl°n °f exPensiv^ »°rk.
before it becomes te^inaT i" to inspect %5?° "^ "^ tO repair daraa«e
regular inspection is also Jhe test ^v t, regularly and because such^
of the value placed on the concrete Mocks tnTZ ""* ^°Ple ^ally are aware
Funds invested in this wav a™T h the Way ln whioh they are used
repeating a raajor ^^'^""^J^.r'-i Pr°P°rU°n °f the °»" »^
Party offices, local councils and the "fSs^ c^t^, " ™a"Ple thrOUSh 5cbools'
ordinary people, especially those living!! central government must reach
are to understand toeir *£^£lSp'^?ZZ£X™* »»•««.. if they
" TTlTt ^^ ln the »»telllteS, the
Salaxy of receivers using several technio^f 7 f f°r flxati™ ««« a new
concept of a »U pack^e from Lxch coordin.f aV3lla"e and «" be refined.
refined, if necessary, by later recalculation of "^ 5e,read ins'»taneously science fiction dream but near realty even if ft^'6" d3ta> ls "° lonser a
watch. eamy even if its size ls not yet thafc Qf a w The nd
ls not yet thafc Qf
x^has nll'ZloTnlc^Vry I'TJlteZTl0' ^V^01' °r °f —observation
are any longer required at all L'.?'* coordinated sround station"
developing countries. It may not vet L *• *' P°Se particula'- problems to
deserves debate in this forum A nwtor^f * a"SWerable but " certainly
be considered in this paper ^ °f 1SSUes are involved - only three will
foreseeable that satellite fixat.nnfn ?rfer °f ':10'°°0 «ale. It ,s certainlv
to 1:2,500 or even u^f. f tlTrell'XTxTt "" ^ ^^^ZT^
-ed of fiction and a full GPS cc^L^^~^ if
-J- Cadastral surveys have tra.-n ti™,11
Ascription of shape and relative di't.n^ C°n°entrated °" the depiction and
As a result it nas'often ZenteM^TZ s^st^ "T °°—<linat-. or position.
reproduce the observation and calculation .7 f measurement which cannot
>f 10 seconds of arc and distance correct to °rle"taUon to southing of the order
regardless of the lengtb of the ^"^^TTg i^T
ECVNRD/GVRr/43 Page 6
expediency - the practicality of extending cadas^{«££to ™« "«
(or perhaps 'readings' might be a better term snee the surveyo ^
i^i^^^^srr r/^iS r™ aPPlied *,. *
measurements within a small area.
, . t ,, *.,-. m9t,v tvn^s of normal engineering
2*1. Much the same considerations appljr to many type feasible with
survey. The setting out of roads and **"« "^«th£ LeTe'ui always be once assoc^atefwith geodetic^framework survey will be needed.
Availability
25. A question which -ust be of concern is ^ber^U of us as ^ ^ providers of land information to our countries can «to^ jr^^ .& &
system, lodged in space, which might c°^e*^* for information may not be able stile environment and many ca-is cstand by if there is even a short radius„ ,, . m,,^v, o c,h/-,r't hiatus in satellite signax&»n. ,,.,„ <a-i»nsi*? for whateveri^^ »»ii«
cause -
They should be maintained, for the cost oi UT*^ ± accuracy, cadasThere will, of
Affordability
27. At present, at least, the trend is that high tecfol^ instr«nt^
that would include satellite f^o^^ul^^^e^ive" «iet»?ve to p
they are utilised ^ten^^^' / *yhi_h[Y educated and skilled manpower- The
and their use is often dependent on nignxy e ^f ^ craftsmen and
manpower is limited, and considerably »"e£ consideration, but those unskilled labour of older techniques, willc^ x® ^^^^ ing countries. The
must come from a very limited SUPW includes a very nigh proportion of installing high ^chnology c™ ^ ^ foreSeeable future.
28. creat strides have been .ade in —.of ^^SSSL^S^l
to protect systems from outside climate and ^"d^°ns; N* t0 dust is hard on tropical climate of Africa with extra.es of_^xdity from ^> r essential electronic based systems. Mechanisms such as tape ™s
for portability, cannot be ^^/^^J^fbe needed and that this is both
equipment suggests that repair and maintenance wi exchange costs, unless tLe consuming and expensive, again with a bias to foreign * wcrkillg facilities are made available reasonably close to, and ace
areas.
Page 7
29.
use.
?!ADJtJSTMENT?
carrv
™) ^ frl°a" (There are already „- L^aineworks cf «t least ro.i. h " un"kelJ that this cost ct h d"contin»«ies on the
wholly to sa^"f.fe^JUStffient irrelevant. A Stnh f b««Wow, from
SCA/NRD/CART/48
References
1.
Overseas Surveys","The Adjustment of Main Survey Networks in Africa by the Directorate
Kiss L M Windsor directorate Paper presented to the 1st Symposium on Geodesy in Africa Khartouni 1974 Subsequently published in African Geodetic Journal Vol 1 No 2.2- "Testing of Homogeneity in Geodetic Networks" S B Mattey Paper presented to the 3rd Symposium on Geodesy in Africa
Yamoussoukro 1986
3- "The African Arc of the 30th Meridian". K F Rain^ford
Empire Survey Review Vol 11 No 82 (1951)