Rejected! Update to Langley VAP Dismisses Outliers

Published: 5 February 2013

An example of a clear day where most values made it through the cloud screening algorithm.
A Langley plot is a regression of log (signal) versus airmass. Under appropriate conditions, the Langley regression yields values that represent the response of an instrument in absence of atmosphere (i.e., at the top of atmosphere), and these values are ultimately useful for instrument calibration. The existing LANGLEY VAP (available through special request from the ARM Data Archive) uses input from the multifilter rotating shadowband radiometer (MFRSR). This new addition to LANGLEY uses the enhanced wavelength resolution and range of the Shortwave Array Spectroradiometer-Hemispheric (SASHE).

An example of a partly cloudy day where many values did not make it through the cloud screening algorithm, but there were still enough good values to determine an I-naught value.
SASHE is a ground-based instrument that measures both direct and diffuse shortwave irradiance. SASHE provides hyperspectral measurements from about 350 nm to 1700 nm at a wavelength resolution from 1 to several nanometers. An automated process couples a Langley regression with an iterative cloud-screening algorithm that rejects statistical outliers. The outlier rejection is implemented at 500 nm, excluding rejected times from the Langley regression for all other wavelengths as well. In addition, wavelength-dependent thresholds are applied to further restrict the airmass range for spectral regions where signal is weak. Finally, a Langley regression over those times passing the acceptance tests is computed at each reported wavelength.

The new SASHE Langley VAP is currently in the evaluation stage, and any comments on the data set are welcome. Please send feedback to Connor Flynn. Evaluation data are available from March 1, 2012, through January 1, 2013, for the Southern Great Plains (SGP) site, and from June 27, 2012, through January 2013 for the Cape Cod, Massachusetts, mobile facility deployment (PVC). The VAP will run autonomously when the evaluation is completed. Resultant data will be used in the Aerosol Optical Depths from SAS-He VAP.

To access these data, log in to the Data Archive. (Go here to request an account.)