New Generation SST for Open Ocean Real-time Demonstration Operation

New Generation Sea Surface Temperature
for Open Ocean Ver.1.0
Real-time Demonstration Operation

Since 2000, the New Generation Sea Surface Temperature (NGSST) Development Group (Leader: Hiroshi Kawamura) has been working on a new satellite-based SST product, which utilizes benefits of the modern satellite/in situ-based ocean observing systems and overcomes weaknesses of the present operational SST products (Literatures, Sample products). Real-time generation and distribution of the new SST products for open oceans has started. The demonstration operation of NGSST-O Ver.1.0 is carried out at the following WEB/FTP sites.

JAXA is providing daily AMSR-E and MODIS data in real time. AVHRR and GOES data are directly received at Tohoku University.

This real-time operation is a part of the GODAE High Resolution Sea Surface Temperature Pilot Project (GHRSST-PP). Further, we are developing the NGSST for Coastal Ocean (NGSST-C with 1km spatial resolution and generated four times in a day) being supported by the MEXT (Ministry of Education, Culture, Sports, Science and Technology) Special Coordination Funds for Promoting Science and Technology ∪The Ocean Environment of East Asia and New Generation Sea Surface Temperature∩ and cooperating with Ocean Remote Sensing Programme of UNESCO/IOC/WESTPAC.

New SST Image

2011/10/08

MENU

All View
FTP Site
Format
Sea Ice Binaries

News:

[2008/09/09]
 NGSST-O server is now in bad condition.  Please note that the NGSST-O data distribution service may be suspended without warning.
[2008/07/29]
 The NGSST-O server will be stopped on Aug.2 and 3 due to electric power cut.
[2007/11/27]
 The NGSST-O server will be stopped on Dec.1 and 2 due to the electric power cut. 
 
[2007/07/30]
 The NGSST-O server will be stopped on Aug.4 and 5 due to the electric power cut. 
[2007/03/07]
 NGSST-O server can not be used on Mar.9 and 10 due to the electric power failure. 
[2007/02/15]
 NGSST-O server can not be used on Feb.16 and 17 due to the electric power failure.
[2006/08/01]
 NGSST-O server can not be used on August 5 and 6 due to the electric power failure.
[2006/03/15]
 The AVHRR SST data are not available during 3/15-22 because the Tohoku University computer center is out of service.
[2005/08/02]
 The data processing will stop from Aug.6 to Aug.8 due to the electric device check.
[2004/10/26]
 The NGSST-O data from Sept.26-Oct.23, 2004 were produced again.  The old data were produced without the AVHRR data because of a mistake in the NGSST-O production system.
[2004/08/05]
 The data production will stop from Aug.7 to Aug.8 due to the electric device check. 
[2004/07/24]
 The data production is resumed at 2004/07/23.  The data from 6/30-7/23 are re-produced.
[2004/07/16]
 The machine trouble stops the data renewal from July 10.  
[2004/06/04]
 Erroneous SST pattern appearred near Shantar Is. in the images of June 1 and 2.  This error is caused by the error of the GOES SST.  Therefore we stop using the GOES SST in the NGSST production from June 1.  The data of June 1 and 2 have been already re-produced.
[2004/03/16]
 From today, we start to open the sea ice mask data from February 2003 to the present.  All the SST images on our home pages have been changed by superimposing the sea ice mask over the SST image (Note: the images are changed, but the contents in the SST data files are not changed).  The sea ice mask data will be obtained through FTP.  Please note that the present sea ice mask data are tentative and the data might be changed in the near future.
   The SST data from Feb. to Sept. 2003 are also added.
[2004/01/19]
 The GOES SST data is used from 1/18 NGSST-O. 
 However, MODIS and AVHRR are doing stop to even 1/13 by 1/15 as for the AMSR-E data due to the trouble now. 
[2004/01/19]
 The NGSST-O data creation stops from 2004/1/16 due to the trouble of AVHRR, MODIS, and the AMSR-E data. 
 The cause is being investigated just now. 
[2003/11/19]
 Notice : The data creation of "11/21 (Friday)" does stop for the network construction in the laboratory. 
[2003/11/10]
 
[2003/11/06]

Outline of the NGSST-O products
Satellite SST observations from infrared radiometers (AVHRR, MODIS) and a microwave radiometer (AMSR-E) are objectively merged to generate the SST product, which is quality-controlled, cloud-free, high-spatial resolution (0.05 degree-grided), wide-covering (50 degree x 50 degree), and daily SST digital map. Using the satellite-derived SSTs, a first guess (mean SST weighted by the auto-correlations in an observational window) is calculated, and then the grid SSTs are produced through an optimum interpolation scheme with de-correlation scales of 200km in latitude/longitude directions and 5 days in time. In order to represent the small-scale and high-frequency SST variations, these processes have been tested and established.

The analyzed area of new SST∈13-63N, 116-166E∷ covers the whole areas of the Okhotk Sea, Bohai/Yellow Sea and East China Sea as well as the western North Pacific Ocean. Since the grid size is 0.05 degree, the pixel and line numbers of SST map are both 1000. The today∏s SST product is generated at around 16:00 LST (7:00 UT) in the next day. Digital values of the mapped SST can be obtained through the FTP. For processing of the digital data, computing information is given in Format.

Quality control of the satellite SSTs prior to the merging process is essentially important. Abnormal values of the infrared 0.01-degree grided SSTs mostly affected by clouds are eliminated through comparison with the cloud-free AMSR-E SSTs, which are the merged SSTs formed by the 10-days AMSR-E observations. The quality control of the infrared measurements in coastal seas, where no AMSR-E observation is available, is conducted using the merged SST product of previous day and the spatially-interpolated coastal SSTs.

In order to produce the today∏s SST map, the satellite-derived SSTs of today and of previous four days (five days in total) are used. Since we need one day for collecting the today∏s SSTs, the generation of today∏s SST maps is delayed one day. If, for a given grid point, there is no SST observation in the 200km spatial and 5day temporal windows, the grid is treated as observational blank.

Validation of the new SST products was carried out using the in situ SSTs of buoy drifting in the analyzed area for August 2003. Match-ups of 18,845 points show that the bias of SST products is -0.31 K against the buoy observations and the root-mean square error 0.79 K.

Index

Feedback : webmaster@ocean.caos.tohoku.ac.jp