Read Me

GUISDAP 8.2

This release of guisdap contains several new features and bugfixes,
but users everything is made compatible with older versions (hopefully).
It has been developed and tested using Matlab 6.5, so there may (or may
not) be some problems with older Matlab versions.


Installing

Guisdap-8.2 is somewhat easier to install than the previous version.
It has been succesfully installed on Solaris (Forte 7) and linux (gcc 3).
Other architectures may have to change the makefiles, if you want speed,
othervise it should run without the compiled routines.
Please report any changes needed for other systems.

untar the distribution at your favourite place
cd into the guisdap directory
if you want the speed improvements the mex utility provides:
 make "lib" and "mex6" directories
 cd models/iri2001
 "make" (should be gnu-make) and "make install"
 cd ../nrlmsise00
 "make" and "make install"
 cd ../../mexsources
 "make"
make a link of bin/guisdap to a place on your path (ex /usr/local/bin)

Integration of historic KST data is done using the Nigel Wade integration
package which you have to install separately. Provided are paths to
NW/mex6 for the mex files and NW/lib for the libraries.

Provided are also a searchpath to ~/gup/mygup where your personal
files should go.


Running

Running analysis is simplier than before.
At the prompt, type "guisdap --help" to see all the switches available.
Normal analysis is done running "guisdap -a" which will bring up a
window with some entries to fill in. (Hopefully selfexplained)
This is the same as "guisdap" and then "analyse". When finished hit "GO".

As usual, you are on your own initialising experiments. If someone writes
or have a manual for this, I gladly debug and include it. In any case,
the initialisation scripts are avaiable with "guisdap -i".


New features

-The built-in integration have been improved with a satellite
 checking routine (normally disabled) ignoring dumps flagged to contain
 point targets. See anal/satch.m for controlling parameters.
-An alternate way of measuring variances on the data has been developed,
 making it possible to analyse a few or single integrated dumps, without
 the need of slow theoretical variance calculations. (Only for the built-in
 integration)
-The use of the parameter "analysis_code" has been extended to control
 what parts of the data to be analysed, which is useful in splitted and
 sliced experiments. Can also be used for experiments run with several
 modes, for example cp1.
-Machinery for correction of gain variation to experiments containing
 gain channels.
-Numerous visible and hidden improvements and refinements


Default parameters

In the analysis there are a number of parameters extracted from the data.
Which parameters depends on the experiment, altitude and radar. The default
settings given below can be overridden by the "fit_altitude" parameter.

			UHF (930 MHz)	VHF (224 MHz)	ESR (500 MHz)
Electron density	All heights	All heights	All heights
Ion temperature		Above 80 km	Above 100 km	Above 90 km
Temperature ratio	107-1500 km	120-1500 km	113-1500 km
Collision frequency	90-107 km	Never		Never
Ion drift velocity	All heights	All heights	All heights
Ion composition		Never		Never		Never


Further information and bugfixes can be found at www.eiscat.com

Good gup!