AMPR IPHEX Calibrated Brightness Temperatures - Level 2B

Data within this directory were acquired by the Advanced Microwave Precipitation Radiometer (AMPR) 
during the Integrated Precipitation & Hydrology Experiment (IPHEX) field campaign in May-June of 2015. 

These files include the Level 2B calibrated and georeferenced brightness temperature for the four AMPR-observed
frequencies (10, 19, 37, 85 GHz). These data are archived in a netCDF-4 format that contains the calibrated
brightness temperatures in addition to ER-2 aircraft navigation and instrument scene georectification variables.

A set of Python software has been developed for reading, plotting, and providing some additional analysis 
capabilities. This software is available from: https://github.com/nasa/PyAMPR


--- V2 Updates ---
In addition to the calibrated brightness temperature, and objectively determined quality control
metric is also provided. The quality control metric is estimated based on the brightness temperature
difference of a pixel within a 9x9 kernel of neighboring brightness temperatures. The QC metric is 
a discretized indicator of the difference within 5 Kelvin increments. Typical scene values fall in the 
QC 1 & 2 bins. However, very noisy scenes -- generally indicative of instrument issues or potential
scene contamination or interference from another instrument --- are isolated to values >= 4. 
As with any objective measure based on thresholding, however, there is a gray area in the 3 & 4 bins
where some of the data is of high quality but physical phenomena are generating sharp, local
features that are flagged as suspect. This, in an of itself, could be useful for those wanting to
isolate features (e.g. the edges of a strong convective cell). 

An incidence angle flag has also been included for quickly identifying pixels associated with
large incidence angles typically encountered during aircraft roll maneuvers. During a roll, often the
edge pixels began to see very large incidence angles or may even contain off-earth sidelobe 
contamination. But, non-edge pixels may still be receiving observations from a typical/moderate (say -45 to 45 degree)
incidence angle. Thus, we have opted for use of incidence angle flagging directly versus simply
eliminated entire scans when the abs(roll angle) is greater than a threshold. 

New pixel field of view (FOV) water fractions are now included.  A 1-km gridded land/water fraction
dataset -- constructed from 250m MODIS Land/Water Mask (https://lpdaac.usgs.gov/products/modis_products_table/mod44w) --
has been used, together with the instrument FOV beamwidths to estimate the percent FOV that
contains surface-water features. These data can be used to quickly identify (and eliminate if desired) 
those pixels originating from a mixed-surface (land and water) scene. 
** Note ** 
The 250m MODIS Land/Water Mask includes water flagging for inland water bodies. However, no land-water mask is perfect,
and it is possible that some smaller inland water bodies are missed. If so, then our FOV estimates will also
be missing the water fraction contributions in such cases. 

********
As an EXAMPLE to quickly identify typical good data, a series of flagging based on the following
conditions may be used:

QC Incidence Angle = 1
Pixel FOV < 0.1  or Pixel FOV > 0.9 , (i.e. mostly land or mostly water)
QC Flag Value <= 3
********

A set of quick-look imagery (in .jpg) is also now included. Both the raw brightness temperatures and an example application
of QC are included for each channel for all flights. Imagery is located in the /quick_look   directory.


------------------

HISTORY
03.23.2015, Release V1
06.01.2015, Release V2

Contacts:
Brent Roberts, jason.b.roberts@nasa.gov , AMPR PI during IPHEX
Timothy Lang,  timothy.j.lang@nasa.gov  , Current AMPR PI