Scientific visualization - projection/georeferencing 3d, etc

Dear community, hello people,

in the end of last year I already described a little bit my project and asked for some first support. Meanwhile I started working for the first time with vvvv and run into some problems… that I somehow suspected. I deeply hope that there are already possibilities to solve the problems, or maybe challenging the professionals for new programming…
I have my vvvv and text file attached. As I described earlier I have to represent scientific data from an ongoing data base. To be accurate, I will represent earthquake data, with the attributes of position (x, y, -z, as lat/long/depth), event time and magnitude (energy).

First of all, it seems that I don´t get the data georeferenced e.g. projected correctly. But this is the basis for all further visual and statistic research.
When developing the concepts, I visualized (static) a set of test data with the open source GRASS GIS, region Indonesia (jpgs attached). GRASS GIS uses for projection the geodetic datum, GDAL (raster data) and OGR (vector data) libraries and EPSG Geodetic Parameter Dataset (a structured data set of Coordinate Reference Systems and Coordinate Transformations) to represent data on the uneven shape of the earth in 2d and 3d, and allows a scale exaggeration in -z (vertical ).
Is this spacial projection possible in vvvv? And how?

Prototype realisation

  1. After reading the data, I tried to built a chronological vector graph with the points2vector (3d) node. ( I need the cylinders to later also use its diameter and color as a variable). And the sphere node to represent the earthquake positions.
    I guess that the map/map range nodes and the earth radius are not the correct projection (this solution seems not scientificly exact and updated, but moreover it does not work, the vectors (cylinders) get all messed up (when entering a value for the cylinder size). Also navigation is difficult in this scale.

  2. In order to visualize the vector directions of the chronological graph in a single origin, with the vector tail placed in the origin (vector rose), I applied a vector translation and then used again a points2vector node.

  3. In order to observe the changing of directions in time of the vector rose, I used the counter node to animate the vectors in chronological order (about 60 vectors visible (sum about 200 data, but later should be above 1000 data).
    But how can I control the time, f.e. by changing/playing with the speed of the counter, how can I invert the time direction of the counter (going back to past) and how can I enter time intervals of the data, and insert sound (event sound)?

There are much more variables that I would like to insert, like the size of the vector diameters (cylinder), vector colors, extract statistics of vector directions in space, or add a dynamic envelope on the vector rose to better observe changes…
But the most important thing I have to resolve is the correct projection!!!

I really would appreciate help,
thanks sueniii

vvvv file (118.1 kB)
txt file (11.1 kB)

seems first image upload did not work, here the missing images in pdf

side view (182.6 kB)
top view (88.4 kB)

hej sueniii,

the way you converted your coordinates is probably taken from some vvvv girlpower examples. while this works well enough for quasi-georefeferencing it is not scientifically correct at all.

do you know exactly what kind of coordinates you have? the most common one would be mercator projection with a wsg84 ellipsoid.
furthermore you’d have to know how your jpg maps were projected. probably mercator too, but national maps are sometimes also projected with lambert conic projection.

once you’ve got all this info, it’s really easy to find the conversion formulas online get exact representation of positions on maps.

hi woei,

thanks for the fast reply!

Yes i took a lot from examples e.g. girlpower…

And yes I do have exact coordinates, you can have a look in the text file attached (earthquake catalog). The source data are point vector data in lat/long/depth coordinates.

I don´t want to project a map/raster data so far (maybe later a map of the region for orientation only)

Maybe you or somebody else has examples for scientificly correct 3d point data projection via lat/long/depth input…?

THANKS sueniii

a datum (lat/long) can be of different standards
that’s why you have to find out, what form the earthquake catalog is using to store the positions. so you can find the appropriate formula to convert that to cartesian coordinates

OK thanks, I will have a look!

Hi woei, hi everybody,

OK, for now I can say that the applied data set is from the geofon global seismic network at gfz and the EPSG-code 4326 is assigned to this data (geographic coordinates), that means the geodetic datum is WGS84 and ellipsoid WGS84.
So how to operate now with this information?!

Many thanks, sueniii

PS. Seems that my points2vector nodes in 3d are refering to some patch on my computer, cause its missing (red) in the attached file. How to resolve this problem when sharing files?

check wsg84 to cartesian conversion random google hit
you need to use this formula instead of Cartesian (3d) so your position data is the same unit as height

that looks interesting ;)
well implementing this formulas/script is way beyond my geo skills, have no idea how to realize this, do you have an example (patch)! or orientation…?
cheers sueniii

hello everybody, does anyone know how to deal with this above formula/script to get correct projection, don´t have programming backgrounds… many thanks sueniii