Hi

How I can type leveling standard? I need to put value of 1mm/km of 2mm/km
if you see picture there is oa, ob(sqr(d)) and oc(d) so in which area I need to put 1 or 2, or I need to put something else?

also I want to ask for datum points. what is difference between reference and datum point?
if I want datum on all points I need to upload file only in datum points?

if I want for example datum on one point, I need to upload file with one point in datum points and file with rest of points in new points?

if is it correct for wat I can use reference points?

Thanks and best regards!

Hello Tasa,

How I can type leveling standard? I need to put value of 1mm/km of 2mm/km
if you see picture there is oa, ob(sqr(d)) and oc(d) so in which area I need to put 1 or 2, or I need to put something else?

The distance depended uncertainties for leveling networks are denoted by σb. You should set σb = 1 mm (or σb = 2 mm) and σc = 0. Since JAG3D does not support σa = 0, you have to select a small number that does not effect the stochastic model, e.g., σa = 0.001 mm

if I want datum on all points I need to upload file only in datum points?

Yes, that is correct.

if I want for example datum on one point, I need to upload file with one point in datum points and file with rest of points in new points?

Yes, that is correct, too.

if is it correct for wat I can use reference points?

You can use reference points, if you have, for instance, some kind of official coordinates that should be fixed.

Kind regards
Micha

Thanks for answer Micha, I appreciate it.

Now I want to ask same questions but for 2D.
If I want to put 2 + 2ppm for distance then I need to put in distance: in oa = 2 (mm)
, oc = 2 (ppm) and ob(sqr(d))=0? is it correct?

What I need to put for angle(direction), if I have instrument standard of 1'' - 0.31mgon, what to put in oa = (mgon), oc = (mm) and ob(sqr(d))= (mm)?

Also for example if I have 4(x, y) points can I put datum on point 1 (both X and Y) and on point 2 but to put datum only on Y? Is this possible?

Thanks,
Best regards.

Hello,

the equations defining the stochastic model are given in the Wiki. Even if the content is in German, the equations are interpretable.

Now I want to ask same questions but for 2D.
If I want to put 2 + 2ppm for distance then I need to put in distance: in oa = 2 (mm)
, oc = 2 (ppm) and ob(sqr(d))=0? is it correct?

Yes, this is correct.

What I need to put for angle(direction), if I have instrument standard of 1'' - 0.31mgon, what to put in oa = (mgon), oc = (mm) and ob(sqr(d))= (mm)?

You have to set σa = 0,3 mgon. If you don't have any further uncertainty, the σb and σc are zeros. Usually, a centring uncertainty has to taken into account. Let say, you are able to point the centre of the target with 1 mm, in this case, you can specify this further uncertainty by σc = 1 mm.

Also for example if I have 4(x, y) points can I put datum on point 1 (both X and Y) and on point 2 but to put datum only on Y? Is this possible?

Yes, but this is a academic construction not used in any practical application. The stochastic points can be diverted from its intended use to reach such a configuration. For instance, you have the points A, B, C and D and xA, yA and yB should be used for defining the datum, xB, C and D are parameters to be estimated. C and D are new points but A and B are stochastic points. Create a file like

A <yA> <xA>  0.00001  0.00001
B <yB> <xB>  0.00001  100.0

The three components that define the datum are fixed (due to the large weight) and xB is a (more or less) free parameter.

kind regards
Micha

Hello Micha

Thanks for answer.

Now I have questions for deformation analysis.

I have 1D measurements in two epoch. I lodaded datum points, new points in zero measurement, new points of last measurements, leveling data of zero measurement and of last measurement, then I load congurence analysis with points names for zero and last measurements (new points in both series)...

First question is how I can do deformation analysis for datum points?

I tried to do deformation analysis of datum points with gaus-helmert and unimodal transformation(uncheck Scale Mz in gaus-helmert)

For unimodal transformation(uncheck Scale Mz in gaus-helmert) after processing my results I always get that is unstable points who has bigger deviation, but positive deviation!.

for example if I calculate deviation in excel the biggest negative deviation between points is - 3mm and the biggest positive is +2.4mm.
But in deformation analysis(unimodal transformation(uncheck Scale Mz in gaus-helmert) I get that is point with +2.4mm unstable and -3 is stable. That is not logically?
if I turn off point with +2.4 another unstable point is usually some positive results like 1.8 which is still better deviation compared with -3..

If is only problem in method which working on this way, what you suggest me, how to do deformation analysis for network points?

Thanks,
Best regards.

Hello tasa,

Now I have questions for deformation analysis.

For congruence analysis an example is given in the documentation of the wiki. Maybe you can translate this tutorial by e.g. google-translate or deeple.

I tried to do deformation analysis of datum points with gaus-helmert and unimodal transformation(uncheck Scale Mz in gaus-helmert)

You don't need the coordinate transformation module. JAG3D provides such a function to compare two epochs in a deformation analysis, cf. zfv 1/2017

/Micha

Hello Micha

Thanks for answer.

I have tried with tutorial which is on website. But I couldn't get information of deformation in datum points.

I uplodaded in datum points (point_1d_datum), new points in epoch_0, new points of epoch_1,
leveling data of epoch_0(turned on reference epoch button) and of epoch_1(turned on control epoch button), then in point nexus 1d I loaded vector_epoche0_1
I adjusted a and b for leveling data and did adjust network.

There on website in tutorial explanation it sees that is the point 3 from datum points group unstable. But in my adjustment I couldn't get same results.

In my adjustment in datum points (in congurence of points tab), there are only zeros..
although on picture which is on website are different results

Also in reports I don't get information of deformations of datum points

Here is picture how I loaded data maybe you can see what is problem.

One method of deformation analysis is gaus-helmert. I wanted to tell you that I get illogical results for 1D transformation. For 2D I get good results. Maybe is my mistake with 1D, but better check is everything OK.

Thanks a lot!

Hi,

I have tried with tutorial which is on website. But I couldn't get information of deformation in datum points.

You have to enable the option congruence analysis in the least squares setting dialog.

/Micha

Hello Micha,

Thanks for your answer.

Can you tell me which method is used for deformation analysis and congruence points? Is Karlsruhe, Pelzer..., I have read on website that is B method but I am not sure what is that.
I need that information to put in my reports which method is used.

Thanks a lot.

Hello tasa,

I have read on website that is B method but I am not sure what is that.

The B-method (or β-method) was introduced by the pioneer work of Baarda, cf. A testing procedure for use in geodetic networks. It describes a procedure to adjust the α-level (type-I error) based on a joint test power (type-II error). JAG3D supports three methods to adjust the significance level. If you don't like to use the B-method, just select your preferred method, cf. Main menu --> Properties --> Test statistic.

Can you tell me which method is used for deformation analysis and congruence points? Is Karlsruhe, Pelzer...,

The method implemented in JAG3D is described in detail in Kongruenzanalyse auf der Basis originärer Beobachtungen. In contrast to the Karlsruhe or Hannover (Pelzers) approach, which based on preliminary results of (at least) two independent free network adjustments, JAG3D combines the original (raw) observations of both epochs in a joined network adjustment. Like the Karlsruhe approach, JAG3D uses explicit hypothesis tests to detect and to identify possible deformations. From this point of view, it is more related to the Karlsruhe approach.

Kind regards
Micha

Thank a lot Micha, I appreciate it.

I have noticed that in adjustment and deformation analysis for example for 1D height of zero epoch or (Z0 like import data in adjustment) of some point for example is Z0 = 100.00 and after processing is Z = 100.15 (calculated). In Δz is value of -0.15, Why software sees that point has settlement because there is minus 0.15 but actually point have uplift because new height is 100.15 that mean it is higher for 0.15. Is that only problem in settings, do I need somewhere to put for deviation calculation to be Z-Z0? Not to be like now Z0-Z because that is little confusing. (Z is currently calculated in software, Z0 is imported as known)

Is possible with that software to do pre-processing of network, if is yes do you have some tutorials?

Thanks a lot!

Hej Tasa,

I have noticed that in adjustment and deformation analysis for example for 1D height of zero epoch or (Z0 like import data in adjustment) of some point for example is Z0 = 100.00 and after processing is Z = 100.15 (calculated).

Yes, this is possible. Please note, datum points are unknown (new) points, too. These points are estimated within the adjustment process. If the geodetic datum is defined as an average value of several points (instead of some kind of minimal configuration), the datum points get (small) corrections. The size of the correction depends on the quality of the datum definition. In a leveling network, only a single point is needed to define the datum. If so, this (single) datum point will not corrected, i.e., Δz ≡ 0.

In Δz is value of -0.15, Why software sees that point has settlement because there is minus 0.15 but actually point have uplift because new height is 100.15 that mean it is higher for 0.15.

In geodesy (at least in Germany, maybe, you can give me some feedback), we distinguish between an error ε and a correction v. The absolut value is identical, i.e., |ε| ≡ |v|, but the sign is differ: $\hat{l} = l + v = l - \epsilon$.

Points that define the datum are not part of the observation space. Thus, no error can be estimated. However, some users asked me to add some kind of metric that indicates the differences between the a-priori values and the estimated a-posteriori values. Since JAG3D provides the error ε for the observations, the same definition is used for the estimated differences Δ. It seems to be more consistent.

Is that only problem in settings

No, it is just a convention like for the error and the correction. If you need the correction instead of the error, please swap the signs mentally.

Is possible with that software to do pre-processing of network, if is yes do you have some tutorials?

Can you explain, what pre-processing means in your context.

kind regards
Micha