664 SNPP VIIRS SDR Quality Flags Testing Methodology

Wednesday, 9 January 2013
Exhibit Hall 3 (Austin Convention Center)
Zi-Ping (Frank) Sun, Northrop Grumman Aerospace Systems, Redondo Beach, CA; and T. Ohnuki

The VIIRS SDR Quality Flags (QFs) were designed to flag data at pixel level (QF1), scan level (QF2 & QF3), scan line level (QF4), and detector level (QF5) for the 16 moderate resolution bands (768 x 3200 pixels per granule), 5 imagery resolution bands (1536 x 6400 pixels per granule), and 1 DND band (768 x 4064 pixels per granule). Each band has its own set of flags from QF1 to QF5. In addition, VIIRS SDR geolocation has its own quality flags at scan level (QF1) and pixel level (QF2). A methodology and a suite of software to test the complex VIIRS SDR and geolocation quality flags have been developed. The methodology and software were designed with two focuses. One is to broadly screen the data and calculate statistics and the other is to address deep dive case studies. The broad QF survey software goes through a given data set and generates a list of quality flags and flag combinations that are triggered and the number of times the flags are triggered. The software also keeps track of granules in which a particular flag was triggered. The deep dive case studies show whether flags were correctly triggered and whether the data were appropriately flagged. In this presentation, we will present (1) summarized tables of QF definitions and possible flags and flag combinations that can be appear in the VIIRS SDR data; (2) description of the broad QF survey tool and statistics generated by the tool; (3) case studies including dual gain bands dual gain anomaly flagging and M6 saturation rollover flagging; and (4) options to improve the quality flag design.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner