Representation of Time Serie of Assimilation Data and Its Application to Animation Content Creation and Representations

Method for Geography Markup Language: GML based representation of time series of assimilation data and its application to animation content creation and representations is proposed. It is validated the proposed method with NCEP/GDAS assimilation data. Also usefulness of the proposed interpolation method with tweening method is confirmed for unequal time interval of the time series data.


INTRODUCTION
Geography Markup Language: GML1 is proposed by Open GIS Consortium: OGC 2 and is International Standardization Organization: ISO standard of open Geographic Information System: GIS [1], [2].It is available to represent 3D GIS data with GML and also to represent time series of 3D imagery data.On the other hands, four dimensional assimilation data is available for four dimensional meteorological data representations such as NCEP/GDAS3 , NASA/DAO4 , etc.It allows visual perceptions and manipulations.Therefore, the proposed method may be useful for such purposes with the four dimensional assimilation data.GML based representation is based on client/server system essentially.Therefore, the required assimilation data have to be searched and downloaded.Assimilation data search starts with meta5 search followed by actual assimilation data.Therefore, data search procedure is required for implementations.Also interpolation method is required for equalization of time intervals between assimilation data because the assimilation data is unequal time interval of data.Tweening6 method is employed for interpolation.Thus the equal time interval of assimilation can be represented and manipulated smoothly.
The following section describes the proposed method followed by implementation and experiments.Then conclusion is described together with some discussions.

A. Open GIS
Open GIS is proposed by Open GIS Consortium: OGC.Open GIS allows collection of geographical data and information through internet in terms of distributed data servicing system based on client server model.Conceptual configuration of Open GIS is shown in Figure 1.Conceptual configuration of Open GIS system Meanwhile, abstracted specification of Open GIS is shown in Table 1 On the other hand, implementation specification is shown in Table 2. Implementation specification includes simple feature, catalog, grid coverage, web map service interface, Geography Makeup Language: GML, web feature service, etc. Essentially, Open GIS system is web based GIS www.ijarai.thesai.orgsystem.GML is XML 7 based data description language for geographical data.

B. Process Flow of the Proposed System
Figure 2 shows process flow of the proposed GIS animation and representation based n Open GIS.First, client makes search request for date, region, layer, altitude spatially and temporally.Meta data, then, is searched from the server.After that, search results are down loaded from the server.It is not always that the interval is constant.In order to demonstrate animation of assimilation data, it is better to show equal time interval of assimilation data periodically.Server provided assimilation data, however, are unequal time interval of time series data.Therefore, some interpolation is required for conversion from unequal time interval of data to equal time interval data.Tweening is utilized for the conversion.Tweening algorithm is as follows, Assuming two data point values, A and B, i-th tweening point, T i is expressed by equation ( 1) (1) where N denotes the number of tweening points.Tweening points of the simplest case are shown in Figure 2. In Figure 2, there are only two adjacent data (line in this case).The line 7 http://ja.wikipedia.org/wiki/Extensible_Markup_Languagewhich connects middle points of the two adjacent lines is called tweening (in between line).Thus interpolations can be done between two adjacent data.Fig. 2.
Tweening Algorithm Equal time interval of assimilation data file format is converted to GIF8 for animation.Then animated assimilation data can be represented.

C. Example of GIS Representation with NVIS of GRASS
Retrieved GIS data can be represented with NVIS9 of GRASS10 software.GRASS is Free Open Source Software: FOSS of GIS.In the GRASS, there is NVIS for GIS representations.Fig. 3.
Example of GIS representation with NVIS of GRASS software D. Assimilation Data NCEP (National Center for Environmental Prediction)/ GDAS (Global Data Assimilation Model) has 360 by 181 of meshed data of air temperature, relative humidity, cloud fraction, ozone mixing ratio, etc. at the number of altitudes which are predicted every 6 hours a day.NCEP/GDAS is described with GRIB format11 and can be read with wgrib12 of command.Example of NCEP/GDAS of air temperature on the surface is shown in Figure 4. Also example of NCEP/GDAS data of surface air temperature in GML expression is shown in Figure 5.

A. Meta Data Search and Assimilation Data Search
Web design is conducted for meta data and assimilation data search.Figure 6 shows the designed web interface of search engine.In this example, the entire query is input from the dialog box in the Figure 6.
The entire query is listed in Table 3.Meanwhile, the ranges for the query parameters are listed in Table 4. Through this web interface, the meta data and the assimilation data in concern can be retrieved with the entire query of attributed parameter for the assimilation data in concern.Then retrieved data is obtained as shown in Figure 7. Example of NCEP/GDAS data of surface air temperature in GML expression www.ijarai.thesai.orgExample of four dimensional assimilation data with equal space and time intervals which is derived from those with unequal space and time intervals through tweening algorithm of interpolation After that, these assimilation data format is converted to GIF format for animation representations.Thus animation of four dimensional assimilation data representation can be done as shown in Figure 10.IV.CONCLUSION Method for Geography Markup Language: GML based representation of time series of assimilation data and its application to animation content creation and representations is proposed.It is validated the proposed method with NCEP/GDAS assimilation data.Also usefulness of the proposed interpolation method with tweening method is confirmed for unequal time interval of the time series data.

Figure 2
Figure 2 Process flow of the proposed GIS animation and representation based on Open GIS

Fig
Fig. 5.Example of NCEP/GDAS data of surface air temperature in GML expression

Fig. 6 .
Fig. 6.Designed web interface for meta data search and assimilation data searchTABLE III.ALL THE QUERIES FOR SEARCHING META DATA AND ASSIMILATION DATA Title Value Date Year Month Day Hour Minite Second Region East end West end North end South end Coordinate System Latitude resolution Longitude resolution Unit of geographical coordinate Map projection Figure 8 (a) shows the retrieved surface relative humidity while Figure 8 (b) shows that of surface air temperature.Retrieved assimilation data are obtained by altitude by altitude and by time by time.Since the altitude interval and the time interval is not equal as shown in Figure 9, interpolation is required in space and time domains.Tweening algorithm is utilized for the interpolations.

Fig. 10 .
Fig. 10.example of the animation of four dimensional assimilation data representation

TABLE I .
SPECIFICATION OF OPEN GIS SYSTEM TABLE II.IMPLEMENTATION SPECIFICATION OF OPEN GIS

TABLE IV .
RANGES FOR THE QUERY PARAMETERS