Signal cutoff to improve retrieval for Raman method #317
Conversation
|
Hi @yZenox , your "patch" has severe implication for all lidars. Also I think the statement that the first 10 bins are unphysical is not valid. THus, we need to disucss about it. Currently I would not like to merge it in dev. |
|
@ulysses78 @martin-rdz What do you think? I guess we cannot do it in such a hard coded way, right? Maybe to use the config file instead? E.g. starting with range bin 262 and then allocating height 41.25 m to it? |
|
Yes, those cutoffs definitly have to be added to the config file for the single polly-device, not in the python file itself. |
…ut partial bins. "slope" parameter should be tested. This replaces the default "smooth" function. Behaviour should get tested besides the pollyFernald module.
… data. flag387OC/flag607OC is missing. Much repetetive copy-pasted code, which eventually should get replaced by loops.
…ut files. The run is splitted into calibrated and uncalibrated mode (=usage of mean lidar constants). The parameter should be put into config file. Run_picasso_test can probably be replaced by an extended picassoProcTodolist.
| RCS = smooth(RCS, window_size, 'moving'); | ||
| % RCS = smooth(RCS, window_size, 'moving'); %old function | ||
| %RCS = dynamic_smooth(RCS, window_size, 'moving'); %new function | ||
| RCS = dynamic_smooth(RCS, window_size, 'dynamic_window'); %new method (full bins) |
There was a problem hiding this comment.
The option whoch smoothing to be applied, should be transfered to the config file instead of beeing hard-coded here. Maybe in future, some other smoothing shall be applied and then it need to be set and traceable....
|
|
||
| %here the lower end of the extinction profiles is set to contant values | ||
| %better be done via config file | ||
| sig532(1:10)=sig532(10); |
There was a problem hiding this comment.
should be added to config file
| %again lower end of the extinction profiles is set to contant values | ||
| %better be done via config file | ||
| sig1064(1:10)=sig1064(10); | ||
| sig607(1:10)=sig607(10); |
There was a problem hiding this comment.
same , value needs to be trasnfered to config file...
| LidarLCexport(4,:)=data.LCUsed.LCUsed607NR; | ||
| LidarLCexport(5,:)=data.LCUsed.LCUsed1064; | ||
| end | ||
|
|
There was a problem hiding this comment.
For what is this needed`? Why are not all lidar constant exported? The lidar constants are usually exported to a data base. What is th eidea behind here?
The signal data in the first bins behaves unusual / non physical. This affects smoothing and differentiation for the Raman retrieval methods. So the first 10 bins are set equal to the value at the 10th bin. This is done for 532 and 1064nm. For 355nm this might be beneficial as well.