NSA/AERI/C1 - Hatch status in datastream suspect

The AERI has an external hatch that covers the foreoptics in precipitating conditions to 
minimize the exposure of these foreoptics (and in particular the scene mirror) to rain.  
Physical limit switches are used to detect if the hatch is open or closed, and the hatch's 
status is determined by the hatch controller and included in the AERI datastream.  
During this period of time, the hatch was determined to be indicated as open too frequently.  
Analysis of several case studies suggested that the hatch was actually opening and 
closing properly but that its status was not being correctly reported by the hatch controller 
to the datastream.

This is a problem because the hatch may be indicated as open in the datastream when in 
reality it was closed; this would result in biases and misinterpretation of the data.

In late-2008, the AERI mentors were able to figure out the problem with the hatch 
controller and resolve this issue. 

An algorithm was written to determine the hatch status from the observed spectrum.  This 
algorithm was applied to all AERI spectra at the NSA site from Jan 2000 to August 2010; 
monthly netCDF files with the update flags are available as a PI product.  Details of the 
method, and monthly distributions of the hatch condition, are in a report that is included 
with the data.  The report demonstrates that the seasonal distribution of hatch 
open/closed conditions is much more uniform year-to-year, and much more realistic based upon the 
known seasonal dependence of precipitation at the NSA site.

The data are available at http://iop.archive.arm.gov/arm-iop/0pi-data/turner/

Please view the publication of  Using a Neural Network to Determine 
the Hatch Status  for detecting when the hatch is in error.

null

• Mean of radiance spectra ensemble(mean_rad)

• Mean of radiance spectra ensemble(mean_rad)

• Mean of radiance spectra ensemble(mean_rad)

• Mean of radiance spectra ensemble(mean_rad)