Xin Qiao, Caiyun Zhang, Chaoyang Fang, Ge Chen
Ocean Remote Sensing Institute, Ocean University of China
Qingdao, Shandong, China, 266003
As a powerful tool in storage, processing, analysis and representation of geospatial information, Geographical Information System (GIS) has been applied in marine field by research institutes, government agencies and private sectors all over the world, and evolved to a series of Marine Geographical Information System (MGIS) since 1980s, especially the last decade. The rapid adoption of the GIS technology and the increasing number of science needs contribute greatly to the advancement of MGIS .
But up to now, there isn’t much progress in MGIS research and development due to the multi-dimensionality and dynamic nature of the ocean environment, which is different from the land. On the other hand, as an important data source of marine and atmospheric research, satellite supplies a brand-new dataset that has large-scale coverage, ideal spatio-temporal resolution, and contains many marine and atmospheric parameters . So it’s necessary to develop a marine GIS software which is specially aimed at the marine and atmospheric research works based on the huge satellite remote sensing dataset.
There are two key reasons that the global oceans and atmosphere are selected as the research object of MAGIS: ?.Oceans and atmosphere are two coupling systems with close connection; ?.Both of them are multidimensional and dynamic fluid systems, which means that the data properties, analysis and representation of them have many similarities. So, this study is proposed to design and develop a GIS software based on remote sensing data —- MAGIS.
1 Overall framework of the system
MAGIS is a GIS software which is designed to complete the data management, spatio-temporal analysis and visualization of marine and atmospheric remote sensing information. It is a seamless integration of three modules: data management, spatio-temporal analysis and visualization, as figure 1 shows.
Fig. 1 Functions of MAGIS (shadowed boxes are under construction)
2 Workflow of the system
Figure 2 is the workflow of MAGIS. It could be viewed as a process in which three modules manage and process kinds of “data”. MAGIS supplies two operation modes: menu-driven and Science Work Flow (SWF) driven. The data operation could only be executed step by step in menu-driven mode, while it could be executed automatically and continuously in SWF mode. Figure 3 is the interfaces of the system. Spatio-temporal interface and visualization interface are menu-driven, and data management function is also included in three interfaces.
Fig. 2 Workflow of MAGIS
Fig. 3 Typical interfaces of MAGIS
3 Major features of the system
Taking oceans and atmosphere as its research object, MAGIS is a satellite RS data based GIS software that could implement the management, analysis and representation of multidimensional dynamic data. Its major features are as follows:
- The system could recognize data with different formats (HDF, NETCDF, etc.). At the same time, it contains such functions as data pre-process, quality control, which enables the users to update MADB with new data. Collocation, merging and assimilation of multi-source data are also integrated into the system;
- Science work flow increases the automation of scientific data operations in MAGIS, and the SWF models could be saved and loaded by way of SWF files;
- The system is purely developed by Microsoft Visual C++ without any other GIS software, and realizes the integration of data retrieval, analysis and data visualization, especially the integration of oceanic dynamic models, thematic plots and advanced analysis methods that are commonly used in marine and atmospheric research.
4 Application Cases
TOPEX/POSEIDON (T/P) is the first space mission specifically designed and conducted for studying the circulation of the world oceans. The mission is jointly conducted by the United States National Aeronautics and Space Administration (NASA) and the French space agency, Centre National d’Etudes Spatiales (CNES). It was launched on August 10, 1992 and ended its mission in August 2002. During its operation, T/P covered the global oceans every 10 days, and made observations of the global oceans with unprecedented accuracy required for ocean circulation studies.
In the Science Work Flow (SWF) interface of MAGIS, we drag icons that represent data, analysis methods and visualization methods respectively to the right interface, and then draw lines that represents the data flow between icons, as figure 3 shows. Double-clicking the icons could modify the parameters of them.
In the present example, we use weekly gridded T/P dataset: maps of Sea Level Anomaly (MSLA) to make a preliminary analysis of the global oceans with MAGIS. The duration of data is from October 1992 to August 2002. As to the analysis methods, we select fundamental statistics, spectrum analysis and harmonic analysis to acquire the characteristics of global sea level variability.
After setting the parameters of each icon, we click the “Execute SWF” button on the toolbar, and then MAGIS could do the above works. After the execution, it will turn to the interfaces of results automatically.
4.1 Statistical analysis
The outputs of fundamental analysis are fundamental statistics (including minima, maxima, mean, range, variance and standard deviation) and (normalized) anomaly. We save the fundamental statistics results into a file, which we could use to make a plot. The normalized anomaly is set to the input of harmonic analysis.
Figure 4 is the mean global sea level anomaly of Oct. 1992 to Aug. 2002. The whole Atlantic, northwest and west parts of Pacific, and the east Indian Ocean have relatively high sea level anomaly.
Fig. 4 Global mean sea level anomaly of Oct. 1992 to Aug. 2002 derived from T/P altimeter
4.2 Spectrum analysis
We conduct a spectrum analysis on each of the 360*164 points of global oceans, and use time series plot to show the frequency-amplitude plot at 3°N, 200°E and classed post map to represent the global oceans’ variability period.
Figure 5 is the frequency-amplitude plot at 3°N, 200°E. It shows us that the primary energy of sea level anomaly variability concentrates on frequency 1, 0.1 and 0.6 year/cycle, which represent annual variability, 1.2 month variability and semiannual variability respectively.
Fig. 5 Sea level anomaly spectrums at 3°N, 200°E
We then use the primary frequency of each point in global oceans to make a 2-D classed post map, as figure 6 shows. About 60% of the global oceans are dominated by annual variability, then 1.2 month variability (25%), and the remaining is seasonal variability and interannual variability. The result is consistent with the fact: the major factors that affect sea level are some marine and atmospheric phenomena like sun radiation, atmospheric motion and ocean currents that have annual features, so the annual variability contributes most to the sea level variability. And secondly, the sea level is also affected by the tides, which has a monthly period. The remaining 15% variability is mainly caused by other marine and atmospheric phenomena like El Nino.
Fig. 6 Geographical distribution of primary frequency in global sea level variability
After the major period of global sea level variability is obtained, a harmonic analysis is performed to extract the annual, semiannual and monthly components of the global sea level variability. We set the harmonic numbers to 4, and period to 52 (weeks). Figure 7 shows the annual and semiannual amplitudes of sea level variability.
Fig. 7 Harmonics of global SLA: (a) Annual component; (b) semiannual component
Obviously, the sea level anomaly of the Northern Hemisphere is higher than that of the Southern Hemisphere, especially the northwest Pacific and the northwest Atlantic. Maxima are found at 10°N and 40°N, where it takes on a banded structure. And the area north of the Indian Ocean also has a relatively high sea level anomaly. In contrast, the Southern Hemisphere is lower.
Higher frequency harmonics have less banded preference. As to the semiannual component, the maximum is found at the northern Indian Ocean, the second maximum at the equator Pacific.
Figure 8 is the spatial distribution of the month in which maximum of annual sea level anomaly occur. The Northern Hemisphere reaches a maximum in JJA, while the Southern Hemisphere reaches maxima in DJF, which is obviously shown in the Atlantic and the Indian Ocean, but not very compellent in the Pacific. This is consistent with the fact: the Northern Hemisphere receives much more sun radiation in summer than in winter, and this leads to the expansion of sea water, and the sea level rise accordingly.
Fig. 8 Month in which annual sea level anomaly maxima occur 5 Perspective of MAGIS
MAGIS is a self-developed Marine GIS software in combination of satellite RS technology and GIS technology. Fully considering the spatio-temporal distribution, sampling and storage format of marine environmental data, MAGIS integrates spatio-temporal analysis methods that are commonly used in marine and atmospheric research, and lays a solid foundation for GIS application in marine field.
In the next development phase, more kinds of satellite data, more efficient and more analysis models and much advanced representation of multidimensional dynamic data will be integrated into MAGIS. On the other hand, the following aspects also need to improve:
- Modification to the Browser/Server mode;
- SWF need to be perfected;
- Introduce VR (Virtual Reality) to MAGIS
- Dawn Wright Darius Bartlett. Marine and Coastal Geographical Information Systems,[M]London: Taylor & Francis, 2000.
- Wei(Wayne) Ji, Rongxing Li. Marine and Coastal GIS: Science or Technology Driven?. [J] Marine Geodesy, 2003, (26), 1-4.
- Feng Shizuo, Li Fengqi, Li Shaojing. An introduction to marine science. [M]. Higher Education Press, 2000.
- Chen Ge, etc. A Satellite Remote Sensing Based Marine and Atmospheric Geographical Information System —- MAGIS. [J] Journal of Ocean University of China, accepted.
- Lee-Lueng Fu, Edward J. Christensen, and Charles A. Yamarone Jr. TOPEX/POSEIDON mission overview. [J] Journal of Geophysical Research, Vol.99, No. C12, 24,369-24,381, Dec. 15, 1994