Manual for the HAMSR real time Neural Network Retrieval
--------------------------------------------------------
Version Date : 2022/03/11
Author : Mathias Schreier



This document is as short description of real time retrieval system for the HAMSR instrument.
The retrieval system is developed by Shannon Brown and creates real time output during thew time of observation.
The description here is just a short summary of the output files and output variables.

In a nutshell: 
The output are netcdf-files and should be readable by any nc-compatible system. 
HAMSR files cover usually 1 hour of observations and can be relative large because of the larg amount of observations per hour.
The files contain Level 1 data (brightness temperatures) as well as airplane information (lat, lon, ...) and retrieved atmospheric variables (level 2).


Most of the variables are self explaining by their names.
- TB describes the observed brightness temperatures for all 25 channels.
- With respect to pressure levels, we have temperature, water vapor and relative humidity at 42 pressure levels.
- The pressure for each level is given with the variable "hamsr_press_levels".
- Additionally there are column information of total precipitable water (TPW) and cloud liquid water (CLW), as well as rain information.
- Airplane information is given with altitude and AC-variables.
- A reflectivity, similar to radar reflectivity is calculated, based on the information of the channels.


Regarding possible filtering of this data:
The neural network does not really have a error flag to discriminate good and bad data.
However, a few things can help to filter the data and make sure only reasonable values are used: 
- The neural network retrieval is mainly validated at low scan angles. Variables like temperature, water vapor and relative humidity are therefore best for scan angles (variable "inc") less than 10 degrees.
- The retrieval will provide profiles at all levels, no matter where the airplane is located. To ensure validity, the profiles should be cut off at airplane height. 
- Lastly: the rain flag should be used to filter data. Data with rain > 0 should be filtered before any process.

For any further questions please contact Shannon Brown or Mathias Schreier at JPL.



The dimensions of the variables:
---------------------------------
channel(25)		
cross_track(varies)
along_track(varies)
HAMSR_levels(42)
HAMSR_dBz_levels(3)

time and geolocation variables:
--------------------------------
time(along_track);
units = "seconds" ;
comment = "seconds since 2000-01-01 00:00:00.0" ;

lat(along_track, cross_track) ;
units = "degrees_north" ;
scale_factor = 0.001 ;

lon(along_track, cross_track) ;
units = "degrees_east" ;
scale_factor = 0.001 ;

inc(along_track, cross_track) ;
units = "degrees" ;
comment = "Pixel Incidence Angle [-180:180]" ;
scale_factor = 0.01 ;



Level 1:
---------
TB(along_track, cross_track, channel) ;
units = "K" ;
scale_factor = 0.001 ;
long_name = "Brightness Temperature" ;


airplane information:
---------------------
int altitude(along_track) ;
units = "m" ;

AClat(along_track) ;
units = "degrees_north" ;
comment = "Airplane Latitude [-90:90]" ;
scale_factor = 0.001 ;

AClon(along_track) ;
units = "degrees_east" ;
comment = "Airplane Longitude [-180:180]" ;
scale_factor = 0.001 ;

ACroll(along_track) ;
units = "degrees" ;
comment = "Airplane Roll [-180:180]" ;
scale_factor = 0.01 ;

ACpitch(along_track) ;
units = "degrees" ;
comment = "Airplane Pitch [-180:180]" ;
scale_factor = 0.01 ;

ACheading(along_track) ;
units = "degrees" ;
comment = "Airplane Heading [-180:180]" ;

scale_factor = 0.01 ;

Level 2 variables (total column):
---------------------------------
PWV(along_track, cross_track) ;
units = "cm" ;
comment = "HAMSR Precipitable Water Vapor - Regresssion Algorithm" ;
scale_factor = 0.001 ;

CLW(along_track, cross_track) ;
units = "mm" ;
comment = "HAMSR Integrated Cloud Liquid Water - Regression Algorithm" ;
scale_factor = 0.0001 ;

rain_flag(along_track, cross_track) ;
units = "/" ;
comment = "Rain flag (0-no rain, >0 rain)" ;
scale_factor = 1. ;
long_name = "HAMSR Rain Flag" ;


Level 2 pressure variables:
------------------------------
ham_airT(along_track, cross_track, HAMSR_levels) ;
units = "K" ;
comment = "HAMSR Vertical Air Temperature" ;
scale_factor = 0.1 ;

ham_airQ(along_track, cross_track, HAMSR_levels) ;
units = "g/m^3" ;
comment = "HAMSR Vertical Absolute Humidity" ;
scale_factor = 0.001 ;

ham_airRH(along_track, cross_track, HAMSR_levels) ;
units = "%" ;
comment = "HAMSR Vertical Relative Humidity" ;
scale_factor = 0.01 ;

ham_pres_levels(HAMSR_levels) ;
units = "mb" ;
comment = "HAMSR Profile Pressure Levels" ;
scale_factor = 0.1 ;

Reflectivity profiles:
-----------------------
ham_dBz(along_track, cross_track, HAMSR_dBz_levels) ;
units = "dBz" ;
comment = "HAMSR Reflectivity Profile" ;
scale_factor = 0.01 ;

ham_dBz_heights(HAMSR_dBz_levels) ;
units = "m" ;
comment = "HAMSR Height of Reflectivity Profile" ;
scale_factor = 1. ;