Cooperative Institute for Climate & Satellites - Maryland

CICS-MD Staff: Submit your Research

Assimilation of VIIRS SSTs and Radiances into Level 4 Analyses

Research Topic: Data Fusion and Algorithm Development
Task Leader: Andy Harris
CICS Scientist: Andy Harris
Sponsor: JPSSO
Published Date: 7/25/2014



Two major SST analysis products generated by NOAA are the Real Time Global SST (RTG) and the Geo-Polar blended.  Both the RTG and Geo-Polar blended products have a wide range of users.  The RTG is one of the primary SST analyses used by the National Weather Service, and of importance to both their atmospheric and ocean forecasting applications.  The Geo-Polar blended product is a critical component of the Coral Reef Watch program (a program with substantial international visibility), the Oceanic Heat Content product (used in tropical storm prediction), the wide range of CoastWatch & OceanWatch users, and the NWS Ocean Prediction Center (for their high-seas forecasts).

This project involves the use of SST observations from NPP VIIRS as input to the two aforementioned analyses.  Improvements are anticipated to be significant given that VIIRS is a much better characterized and less noisy sensor than those currently used in these analyses.  After the VIIRS data has been incorporated into the analyses, the impact of the VIIRS data will be assessed.  As part of the overall project, the improvement to the RTG is being assessed by the RTG team by comparing with their current system.  If the resultant analysis is equal or superior to the current AVHRR and in situ data already in the system, it will be implemented as soon as possible as part of the NWS NCEP operational product suite.  


Code has been written and tested for the ingestion of the three different formats of NPP VIIRS SST produced by the ACSPO team, namely HDF, netCDF and L2P.  The latter is a specialized format that contains product layers specified by the Group for High Resolution SST (GHRSST), and is also in netCDF format.  Once ACSPO VIIRS becomes available in L2P format, we will be able to test the extra capability of our ingestion software to take account of the single sensor error statistics (SSES) provided for each pixel.

Figure 1 shows the bias correction for the VIIRS nighttime SST product with respect to the RTG_HR_SST for day 150 of 2013.  Note that these are the correction values, thus the prominent cool “bias” in the Southern Ocean is indicative of the VIIRS SST being warm cf. the RTG.  In this instance, however, it is quite likely that the bias is in the RTG rather than the VIIRS, since bias patterns similar to this have been observed for other L4 analysis products w.r.t. the RTG.


Figure 1: Bias correction for nighttime ACSPO VIIRS SST w.r.t. RTG for 2013 day 150.  The cool band of negative correction in the Southern Ocean is likely to be due at least in part to a bias in the RTG.

An example of the resolution of individual SST features in the VIIRS data can be seen in Figure 2.  Note that this is on the 0.05 degree input grid of the Geo-Polar SST Analysis, thus each pixel represents the quality-controlled “super-ob” of a number of VIIRS pixels, including the benefits of averaging after outlier rejection.  The feed-through to the final analysis can be discerned in Figure 3.


  • Testing of the SSES that should be provided in the GHRSST L2P version of the ACSPO product
  • Continue assessment of the impact of VIIRS on RTG SST Analysis (NCEP)
  • Investigate alternative bias correction estimates using VIIRS radiance data
  • Transition algorithm to operations
  • Assess impact of improved Geo-Polar SST analysis on Coral Reef Watch products




  • Tested VIIRS ingestion code transitioned to NOAA Operations;
  • Geo-Polar SST dataset provision to NOAA Coral Reef Watch


Figure 2: Nighttime ACSPO VIIRS SST after preprocessing onto the 0.05 degree grid in the vicinity of the Agulhas and Madagascar.


Figure 3: Final Geo-Polar SST analysis after incorporation of the VIIRS SST shown in Figure 2.  Note that the high-resolution features present in the VIIRS data are generally well-retained.









# of new or improved products developed following NOAA guidance


# of products or techniques transitioned from research to ops following NOAA guidance


# of new or improved products developed without NOAA guidance


# of products or techniques transitioned from research to ops without NOAA guidance


# of peer reviewed papers


# of non-peered reviewed papers


# of invited presentations


# of graduate students supported by a CICS task


# of graduate students formally advised


# of undergraduate students mentored during the year




The budget for year 2 has been split over Years 2 & 3, so a number of results will extend into a third year, as described above.

close (X)