[GRASSLIST:3541] Re: Geophysical/Potential field modules for GRASS?

[Hamish]

> This is part of what is needed, but not all of it. What happens in 
> geophysical survey is that someone walks up and down transects within
> a grid; the data recorder ticks off the readings at each point; it may
> let the operator record the transect. As so...
[...]
> 
> 1.1  2.5  3.1  4.5  5.1
> 1.2  2.4  3.2  4.4  5.2
> 1.3  2.3  3.3  4.3  5.3
> 1.4  2.2  3.4  4.2  5.4
> 1.4  2.1  3.5  4.1  5.5
>
> The data recorder outputs a data file more or less like this.....
> 
> 1 time date value
> 2 time date value
> 3 time date value
> ...
> 
> These need to be translated into xy coordinates for each point given 
> the (x) spacing between transects and the (y) spacing between data 
> reading along transects.


Sounds like a job for Matlab/Octave to me .. We don't get to lay nice
grids over our study sites at sea (or manage to get the ship to go in a
straight line anyway), so what I've done in the past is to log the NMEA
output from a GPS into a laptop while our instruments spit out
"1,time,date,value" style data which gets recorded into another file[*].
A Matlab postprocessing script does a linear interpolation of position
from the NMEA log for each data point, and directly writes a GRASS sites
file. Any spatial interpolations of the data can then be done from
there in the GIS. [It's not as hard as it might sound]

[*] all clocks sycronized to GPS time before hand of course

If you already have a grid set out, you can skip all the GPS stuff & the
script is all the easier.

I don't see how this could be implemented as a generic GIS module.
Each job is custom..


Note you can deal with no-data points (NULL) in the GRASS 5.0/5.3 sites
format by using a numerical value of 'nan'. e.g.:
easting|northing|#1 %nan @"date/time"

This probably won't work with most sites modules, but it works with more
than you might think as 'nan' is a valid floating point number on many
systems. e.g. 'd.site.labels attr=double' doesn't mind it.



> I do think it would be handy for other reasons to add a -g flag to 
> r.transect

I think so too; done.
also fixed a bug (G_find_cell() was stripping the mapset off the map name).




Hamish