Abstract:
Sea Surface Temperature Imagery Overview:
The University of Rhode Island (URI) Oceanographic Remote
Sensing Laboratory maintains an archive of processed, geographically
rectified and atmospherically corrected (for 5 channel AVHRR sensors)
images of sea surface temperature (SST) for selected areas of the
eastern seaboard and offshore of the United States. The images are
classified by region and
... are created from all available Level 1b AVHRR
at URI. Since data from all of the NOAA Series AVHRR Platforms (TIROS-N,
NOAA-6, NOAA-7, NOAA-8, NOAA-9, and NOAA-10) are compiled in these data
sets a generic term, NOAAXX, is used to specify the source platform.
URI generates 512x512x1 byte images of 6 regions from Level 1b
-data. The names of the geographic regions, the latitude and longitude
range of the regions and the pixel resolution defined for their
processing are:
Region Lat/Lon Resolution
Gulf of Maine 40-45N 64-72W 1.1km
New York Bight 37-42N 69-76W 1.1km
Cape Hatteras 34-39N 70-76W 1.1km
Northeast Area 34-44N 63-76W 2.0km
Sargasso Sea 30-38N 64-74W 1.7km
Sub-Tropical Convergence 22-33N 61-73W 2.0km
All images are generated under a standardized processing procedure
described below.
SST Image Processing
The image display and processing system used at URI is DSP,
a system developed by the University of Miami Rosenstiel School of
Marine and Atmospheric Sciences. The procedure described is
followed for processing Level 1b data to a 512x512 rectified and
corrected image format.
1/ Ingesting of Level 1b data from magnetic tape to magnetic hard disk.
In the case of TIROS-N, NOAA-6, NOAA-8 and NOAA-10 channel 4 is
ingested. For NOAA-7 and NOAA-9 both channels 4 and 5 are ingested.
2/ Precision renavigation and update of separate navigation database
records. Operators correct spacecraft position errors resulting
from inaccurate ephemeris information stored with the Level 1b
data records by the platform software. Through changes in time and
platform orientation parameters the operator corrects the
spacecraft ephemeris using the known position of landmarks
visible in the satellite passes. This procedure provides
consistent navigational accuracy to within 1km of true.
Corrected values for ephemeris and platform orientation are
permanently stored, on-line, in the DSP processing system navigation
database. This database is referenced by the processing system
when it creates a corrected version of the Level 1b data at
full resolution.
3/ Processing of data to produce corrected temperature or sensor
radiances. For NOAA-7 and NOAA-9 accurate values of SST which have
have been corrected for atmospheric attenuation can be obtained
using two channel split window method with channels 4 and 5.
(McClain et al 1983, McMillin and Crosby 1984). The form of the
equation for SST is,
Tss = A0 + A1*T1 + A2*(T2 - T1)
where Tss is the desired SST value, T1 is the channel 5 brightness
temperature and T2 is the channel 4 brightness temperature. A0, A1
and A2 are the regression coefficients for the equation. The
following coefficients have been derived from comparison with
in situ measurements obtained from drifting buoys (NOAA/NESDIS 1982,
1985)and were used to generate SSTs from the NOAA-7 and NOAA-9
digital data.
NOAA-7 A0 A1 A2
DAY -283.93 1.0351 -3.046
NIGHT -288.23 1.0527 -2.627
NOAA-9 A0 A1 A2
DAY -268.92 .9864 -2.670
NIGHT -268.41 .9855 -2.668
For TIROS-N, NOAA-6, NOAA-8 and NOAA-10 sensors radiances are
computed directly from channel 4 of the the Level 1b data.
While it is not possible to remove the atmospheric attenuation
from single channel data, images generated using sensor radiance
values contain detailed information on the thermal structure
the surface waters and features.
The processing to SST values and sensor radiances produces an
intermediate full resolution version of the data that is used
to generate the final data product, the standardized remapped
images.
4/ Geographically rectify and remap 512x512 standard images. The
end data product is the rectified image. These images are
generated from the corrected version of the Level 1b data at a
specified resolution and location. A pseudo-mercator
projection is used with the pixel scaling defined by LATSIZ = R,
and LONSIZ = R/cos(LAT). R is the specified resolution.
Navigation and processing information are stored along with the
remapped image as an image header at the time of creation.
URI SST Archive
Over 1800 separate images for each of the six regions are presently
stored on magnetic tape. The archive is constantly being updated
with new as well as historical data.
URI SST Satellite Image Archive available on the World Wide Web (WWW)
The URI Graduate School of Oceanography has provided access to the
AVHRR sea surface temperature satellite image archive through the
World Wide Web. Each image covers the area from latitude 60.575N
longitude 96.222W to latitude 9.399N longitude 33.768W and contains
1024 x 1024 pixels. The resolution is 5 km/pixel in equirectangular
projection. The archive consists of over 20,000 images.
There are two ways to browse the images using a web browser or
XBrowse.
To view the images using a forms-based WWW browser, just use the forms
to get the desired image. XBrowse is an interactive tool for
satellite image browse and retrieval and must be built onto the user's
workstation.
Related Data Sets:
URI NOAA-6 Level 1b
URI NOAA-7 Level 1b
URI NOAA-8 Level 1b
URI NOAA-9 Level 1b
URI NOAA-10 Level 1b 
