Home | Contact ST  

SAIDIN: A Tool to Access, Visualize and Distribute Satellite Data
Technology Provides Easy-to-Use Satellite Data, Allowing Manipulation and Enhancement of the Images


Òscar Chic
Coastal Ocean Observatory Manager
Roger Olivella
Ph.D. Student
Emilio Garcìa-Ladona
Department of Physical Oceanography
Institut de Ciències del Mar
Barcelona, Spain

Traditionally, satellite data has been difficult to work with due to a disparity in formats, the size of the files and the necessity of having proper software to process the data.

There are basically two sorts of remote sensing data users. One kind is mainly interested in doing qualitative studies, so it is enough for them to be able to browse images from data repositories and download the imagery. The other type of user, however, is interested in doing deeper analysis of the satellite data. These users need raw data.

This second group currently has the ability to download raw data from remote databases like the University Corporation for Atmospheric Research (Boulder, Colorado) Unidata program’s thematic real-time environmental distributed data services (THREDDS) catalogs in scientific formats, such as network common data form (netCDF) and hierarchical data format (HDF). Also, in the last two years, a powerful visualization tool for operational oceanography called Satellite Image Database Interface (SAIDIN), based on a synoptic view and a full-resolution window of a georeferenced image, has been developed to access, visualize and distribute data to a wide spectrum of users.

The aim of this project is to make it easy to work with satellite image data through an application that facilitates its use, allowing a certain level of manipulation and enhancement.

Data Acquisition System
A high-resolution picture transmission (HRPT) receiving station has been operational at the Institut de Ciències del Mar since 2001, allowing users to acquire data from several satellites. At present, the system is able to receive data coming from NOAA series satellites NOAA-17 and NOAA-18 and decode data from the Advanced Very-High-Resolution Radiometer (AVHRR), the Argos data collection system (DCS) and the Television Infrared Observation Satellite Operational Vertical Sounder sensors. In addition, sea-viewing wide field-of-view sensor (SeaWiFS) data from the satellite Orbview-2 have been decoded until the end of 2004, and data from the VEGETATION sensor on the satellites SPOT 4, FY-1C and FY-1D can be decoded using appropriate software.

The system is composed of a receiving subsystem, a computer subsystem of control and data archiving, and a processing and publishing subsystem. The receiving subsystem consists of a 2.4-meter-diameter antenna, a universal receiver for polar satellites and a tracking antenna system. The computer subsystem consists of a PC equipped with acquisition cards, a DVD writer and a digital linear tape/digital audio tape storage unit. The processing and publishing subsystem is a GNU/Linux machine with CentOS (community enterprise operating system) distribution.

In the years following the creation of the HRPT station, a near-real-time system was developed, allowing users to acquire data, process them to obtain temperature maps of the Western Mediterranean at its maximum resolution (nadir) of 1.1 kilometers and publish them to the Web in approximately one hour. Additionally, a bzipped netCDF file with AVHRR channels, latitude, longitude, land mask and multichannel sea surface temperature (MCSST) variables is created and added to a THREDDS catalog. It is also possible to process HRPT NOAA telemetry to obtain the position of the Argos DCS platforms and make this information available to users.

The processing layer executes routines of radiometric, geometric and atmospheric corrections automatically with scripts that use SeaSpace Corp.’s (Poway, California) TeraScan® software. The current routines for atmospheric correction are not sufficient, and sometimes under specific atmospheric conditions (like low clouds in a very dry atmosphere, sand advections or mists) do not output a product with the required quality. Therefore, it is recommended that users reprocess this data manually. After these corrections, maps of temperature are produced using McClain split-window equations and made available for the scientific community on the Web. During the initial processing, quicklooks and metadata subproducts are also generated, providing complementary information useful for gauging the quality of temperature maps and allowing users to detect processing errors.

Operation Products Database
At present, the following products are available: AVHRR NOAA MCSST near-real-time data (1.1-kilometer spatial resolution at nadir); AVHRR NOAA averaged SST data; quicklooks from AVHRR NOAA; AVHRR NOAA monthly SST animations; wind data from NASA’s QuikSCAT (Quick Scatterometer) satellite (0.25° spatial resolution: two meters per second, 20° accuracy); and daily, weekly and monthly SeaWiFS chlorophyll mean data (nine-kilometer spatial resolution) from the Western Mediterranean region (around [35N-46.5N][15W-16.5E]) with different temporal coverages.

Maps of average temperatures are obtained using the maximum pixel value algorithm, and visible and near-infrared raw radiance counts are collected (Level 1A data) and converted to geophysical parameters (Level 2 data). The Level 3 standard mapped image (SMI) data products are image representations of the binned data products. SMI files are generated for five geophysical parameters, including chlorophyll-a concentration.

The Instituto de Ciencias Marinas de Andalucía-Consejo Superior de Investigaciones Científicas (ICMAN-CSIC) obtains Level 3 SMI images from the NASA Goddard Space Flight Center repository and processes them using the SeaWiFS Data Analysis System v4.8.

The Level 3 data set of QuikSCAT/SeaWinds provides gridded values of meridional and zonal components of wind velocity and wind speed squared twice per day. QuikSCAT has been measuring ocean winds from July 1999 through the present.

ICMAN-CSIC downloads and processes QuikSCAT data from the NASA Jet Propulsion Laboratory in HDF format. These data are given on a regular 0.25° grid and separated into ascending and descending passes. Chlorophyll and wind products are projected in a Mercator projection using an M_Map toolbox written to The MathWorks Inc.’s (Natick, Massachusetts) MATLAB. Plans are in place for it to populate the chlorophyll database using moderate-resolution imaging spectroradiometer data from Aqua/Terra satellites.

A SAIDIN applet showing an info window at the bottom left, which displays temperature information from Puertos del Estado data buoys and the buoys’ position overlaid on a NOAA temperature map in the middle window.

Quality Control
A separate application is devoted to add and update news, links and some documentation of the Web site. It also allows Web site maintenance and quality control procedures.

During prediction tracking calculations, a first filter based on the pass selection by the criterion of maximum angle of elevation is applied. This also chooses a region of interest. An additional supervised quality control is operated on a monthly basis to prevent transmission errors. Images that do not have the required quality can be erased or simply marked, since images not suitable to contribute to the temperature average could still provide useful information.

A statistical study was conducted during the RV Hespérides‘ Omega-97 campaign, comparing the in-situ data from a thermosalinograph with the satellite data. It obtained a statistical root mean squared error of 0.7, which is within the error range accepted for NOAA SST data.

SAIDIN Features
The SAIDIN Web site provides access to daily SST temperature averages; monthly animations and historical data from AVHRR NOAA; historical wind fields from QuikScat/SeaWinds; and daily, weekly and monthly averaged SeaWIFS chlorophyll maps. Moreover, quicklooks and metadata of the NOAA images can be analyzed.

Recently, SAIDIN has been implemented as a new THREDDS catalog viewer, in addition to the Integrated Data Viewer and netCDF/THREDDS ToolsUI. Some problems still must be solved, and it remains under development. The application provides geographical information and temperature pixel values and allows the user to choose between different color palettes (black and white, color or zebra) to enhance visualization. It offers the capability to browse the database by time and to download single images as well as a set of images.

For marine operational and monitoring activities, the SAIDIN interface also allows the user to overlap and combine information coming from other data sources, when available. Recently, the trajectories of Lagrangian drifters (like those of the Argos DCS tracked system) have been implemented, and it is also possible to overlay temperature information from a network of environmental buoys (i.e., coastal buoys from Puertos del Estado, Spain).

SAIDIN is implemented in JavaScript as a Web application. Client layer is an applet that works like a multiapplication; that is, changing the front-end depending on parameters the user chooses. The back-end is composed of several servlets that interoperate with the applet and carry out tasks such as the requests of the client and searching within the database. The management of imagery is based on a Java2D image library. The applet is signed to allow it to read and write files in the local file system.

The applet is divided into three panels. The left one has a search utility that allows users to sort the images within the database by date and also shows a small quicklook of the selected image. The center panel has a cropped version of the original image at full resolution presented with two sliders. Users can drag the mouse over the quicklook on the left and a zoomed-in image is simultaneously shown in the central panel. This window contains a contextual menu that provides other functionalities. The panel on the right shows information and allows the user to manipulate the brightness and contrast of the images. In SST mode, the pixel info on the top window shows information about the pixel selected in the central panel. It changes dynamically as the user moves the mouse over the image. It provides latitude, longitude, temperature, pixel (row and column) and digital number.

The application, from the server side, is based on Java server pages and servlets technology. The application is entirely contained in one Web application archive file that can be easily deployed in a servlet container (e.g., Apache Tomcat). The whole configuration is codified in one XML file facilitating the Web site’s deployment and maintenance.

As a result, the application can be used by other data centers to distribute their own data with respect to file name conventions and image formats.

From the user’s point of view, there are two ways to use the application—either to run the a2pplet embedded in the Web site pages through any browser or to download the applet from the Web and execute it locally (provided the user has previously installed a Java virtual machine). In this way, the user does not need to open the browser to run the application and access the server database remotely.

The SAIDIN interface is under continuous development to provide more functionality, and it has recently been updated with a European Organisation for the Exploitation of Meteorological Satellites Broadcast System for Environmental Data, which can receive data based on standard digital video broadcast technology using commercial telecommunication geostationary satellites.

This allows a low cost and very simple system (compared with an HRPT) to obtain similar data with only a slight delay. It is intended to provide improved near-real-time quality control using independent information, such as that supplied by an environmental array of buoys. Additional improvements include migrating the Web site to a content management system to facilitate Web updating and decoupling the applet application using Web site supplementary information.

New implementations will allow users to better integrate SAIDIN into the THREDDS catalog server as a viewer and increase SAIDIN’s capabilities as a light geographic information system, allowing users to overlay data from several sources and to implement new products, such as ocean surface velocity fields derived from SST maps.

From a technical point of view, SAIDIN will provide flexible file name conventions, image formats and database server dependency.

Òscar Chic heads a coastal ocean observatory based in Barcelona, Spain, that consists of a meteorological station, a current and wave profiler, a satellite receiving station and a video monitoring system. His research interests include developing Web applications to distribute scientific data that are compatible with Open Geospatial Consortium geographic information system protocols.

Roger Olivella received a B.Sc. in physics from the Autonomous University of Barcelona in Spain in 2001 and an M.S. in remote sensing and geographic information systems in 2003. He is currently a Ph.D. student at the Institut de Ciències del Mar in Barcelona. His research interests include meridional overturning circulation, quaternary glacial-interglacial changes and ocean physiology.

Emilio García-Ladona is the head of the department of physical oceanography at the Institut de Ciències del Mar in Barcelona, Spain. His research interests include studying ocean mesoscale dynamics, mainly through remote sensing, and pattern recognition techniques based on wavelet analysis and singularity analysis to exploit satellite image data for a range of topics from ocean dynamics studies and Lagrangian dynamics to ecological modeling related to harmful algal blooms.

-back to top-

-back to to Features Index-

Sea Technology is read worldwide in more than 110 countries by management, engineers, scientists and technical personnel working in industry, government and educational research institutions. Readers are involved with oceanographic research, fisheries management, offshore oil and gas exploration and production, undersea defense including antisubmarine warfare, ocean mining and commercial diving.