1335 The Benefits of the Bayesian Cloud Phase Approach

Wednesday, 25 January 2017
4E (Washington State Convention Center )
Denis Botambekov, CIMSS, Madison, WI; and A. Heidinger and A. Walther

The current Clouds from AVHRR Extended (CLAVR-x) Cloud Phase retrieval follows threshold based tree decision logic for multiple tests in visible and near infrared range of spectrum. The algorithm shows partly sensitivity to choice of thresholds and to order of tests.

    At the same time Bayesian Cloud Mask is a recognized, and accepted by many users, operational algorithm for NOAA AVHRR and GOES projects. The Bayesian approach employs in development phase the comparison of tests with proxy data from CALIOP. This resulted in an improved cloud detection   

   We are researching a new way to determine Cloud Type by using a Bayesian approach at the same moment with a Cloud Mask, which is considered to be more accurate at the difficult conditions, such as overlapping, etc.  It would be especially beneficial because the Cloud Phase uses essentially the same tests as the Cloud Mask.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner