Assessment of the SNPP VIIRS Cloud Optical Properties and Cloud Top Parameters Using the Calipso Lidar Measurements

In this paper we will present the results in the comparison of the cloud optical thickness and cloud top height between the SNPP VIIRS cloud products and Calipso cloud layer products. Both the SNPP and Calipso are essentially in the same polar orbit plane, but they differ in altitude. Due to the altitude difference, nearly-simultaneous nadir overpass (SNO) between the two spacecrafts occur approximately every 64 hours. In this comparison study, SNPP granules are selected based on SNO covering several climate regions of interest. Specifically, we have chosen the high-, mid-, and tropical- latitude regions, in the months of February, May, and August 2012. In some comparisons global coverage is used.

From various Calipso cloud products, we have chosen to use the 1-km cloud layer lidar products (CAL-LID-L2-01kmClay-xxxx) so that SNPP Cloud Top Intermediate Product (IP), which has 750-meter spatial resolution at nadir, can be directly compared without introducing errors from aggregation. For the comparison of cloud optical thickness (COT), however, we chose to use the 5-km cloud layer COT from Calipso product (CAL-LID-L2-05kmClay-xxxx) and the 6-km SNPP aggregated EDR COT product. A nearest neighbor approach rather than spatial interpolation is used in cloud property comparisons between the data-point match-up based on geolocation. For the 5-km Calispo product COT, values are arrived from aggregation along the flight path while the SNPP COT EDR products result from aggregation along track and scan directions. Due to this difference, the COT performance comparison is expected to yield less accuracy than in the CTH.

To accurately assess the performance of SNPP cloud products, only the “high quality” pixels are used in the comparison. High quality pixels are defined for which the retrieved values are derived from converged algorithm solutions. In addition, both SNPP and Calipso data must be “confidently cloudy” and having the same cloud phase. With mutual agreements, the input cloud mask errors in cloud detection and cloud phase can be eliminated.

In conclusion we shall describe some of the updates to the CTP algorithm and the expected performance improvements from these updates.