Procedures to Validate S-NPP Sounding Products using Conventional and Reference/Dedicated Observations,

The NOAA Products Validation System (NPROVS) supported by JPSS and operated at NOAA NESDIS Office of SaTellite Applications and Research (STAR) provides ongoing data access, collocation and inter-comparison of multiple satellite sounding product suites against global conventional radiosondes (RAOBs). Satellite product suites include NOAA legacy operational products such as ATOVS, MiRS, IASI (MetOp-A), AIRS (Aqua), GOES, COSMIC and more recently (2012) the CrIMMS and NUCAPS from S-NPP and latest products from MeTop- B (2013). NPROVS has operated since 2008 and particularly over the past year has undergone modifications specifically in support of cal/val and algorithm development goals as defined within JPSS/STAR sounding product project.

The following paper overviews the basic element of the S-NPP cal/val program and associated modifications and expansion of NPROVS to meet CrIMSS cal/val objectives. The two basic elements of the program can be separated into the Collocation Dataset and Validation components. The datasets consist of two elements, those containing conventional observations and the special sets containing Reference (ie, GCOS reference Upper Air Network (GRUAN) and Dedicated (ie, satellite overpass synchronized and funded by JPSS) observations. Conventional datasets are global with relatively large samples whereas as the reference/dedicated are sparse but highly reliable and traceable. Respective strategies for compiling collocated observations within each set are presented. Respective validation strategies in the context of each dataset are also presented. For example, opportunities for up-scaled validation and a direct interface to algorithm development activities are both justified and exercised using the reference and dedicated observations compared to conventional observations. However, the large sample and more complete global coverage of conventional data also offer distinct validation advantages. The utilization and interplay of these two approaches, referred to as NPROVS (conventional) and NPROVS+ (reference/dedicated), to provide a thorough ca/val program for S-NNP comprise the main theme of this paper. The paper concludes with examples of validation results using GRUAN observations including the direct use of GRUAN uncertainty estimates in satellite product assessment and also as a source of feedback on GRUAN performance.