RSD marit's home projects

The HARLIE code is being modified to move HARLIE data off the cone using radiosonde winds and create isosurfaces of volume generatde.

First the HARLIE code developed last summer is being modified to apply some standard correction factors. One problem so far appears to be that the scan angle correction makes the data worse not better. This page has some plots showing the scan angle correction.

Further analysis of the data shows that some sort of scan angle correction needs to be applied.

Initial tests applying the radiosonde winds to the data show that the wind direction at different altitude is different enough that the data volume generated with wind-blown data will be skewed such that isosurface generation will be problematic.

I would suggest the approach of using one average wind direction is chosen for all altitudes to generate a more aligned volume.

This is an IDL program that plots perspective views of Harlie data. A tar file of IDL code can be downloaded and run to create various perspective views.

This tar file contains the IDL source code.

run_j_process_harlie is the name of the IDL procedure that calls the main Harlie plotting program. You can set the perspective view x,y and z rotation and zoom in run_j_process_harlie and also specify which files should be processed. j_plot_harlie_persp is the subroutine that does the perspective plotting. The IDL procedure MESH_OBJ is used to define a conical object to texture the harlie data onto. I actually define two mesh objects, each with normals pointing in the opposite direction so that if one object has all its normals pointing away from the viewpoint the other object will have normals pointing towards the viewpoint (if an object's normals are pointing away from the viewpoint it is invisible.) The procedure cleanplot clears all system graphics variables, while the IDL procedure CREATE_VIEW sets up the perspective view. Most of the routines I've written start with 'j_' so there won't be naming conflict with your local subroutines, two exceptions are findmanyfiles and harlie_struct__define.

I took out the code that created vis5d files because it didn't seem that useful given the size of the vis5d data files.

Here are some example quicktimes made from the perspective views generated by the IDL code. They show the effects of varying ax, ay and az. I haven't tried to make particularly nice images here - I am just trying to show how to get the data onto a cone and viewed from any angle.