Abstract
The advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) is one of instruments boards on EOS-AM1 spacecraft, will be launched by NASA in USA on June 1998. The ASTER instrument has developed by Ministry of International Trade and Industry of Japan, based on the requirement of ASTER science team constituted with Japanese and American scientists.
DEM working group of Japan's ASTER science team defined the general specification of DEM products, and has been developing the algorithm of DEM generation. The DEM products generated by the algorithm will be distributed through ASTER ground data system (ASTER GDS) that has been development by Earth Remote Sensing Data Analysis Center (ERSDAC) of Janan, In this paper, We disclose the specification, contents and accuracy of the DEM products.
ASTER DEM products will be provided to 2 type products such as XYZ set and Z set. 'XYA set is consisted of header, relative DEM and Quality flage (correlation value, abnormal value, water and cloud area). Z set is consisted of header and the 30m grid DEM made from XYZ set and Quality flage. Relative DEM means elevation data derived from ASTER stereo pair images using data are genrated from JERS1/OPS data and DEM derived from topographic maps.
Using the simulation data, DEM are generated by stereo matching on area correlation.
As a result, the relative DEM was generated as the accuarcy of 12.5m R.M.S.E.
1. Aster instrument
1.1 Spececraft Parameters
The Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) is one of instruments boarded on EOS-AM1 spacecraft, will be launched by NASA, USA on June, 1998.
The spacecraft parameters shows in table 1.
Table 1 Orbit Parameters
| Orbit |
Sun Synchronous |
| Cycle |
16 days |
| Period a orbit |
98.9 min. |
| Altitude |
705m (at equator) |
| Inclination |
98.2 degree |
| Orbit Position Knowledge |
150m (3s |
1.2 ASTER Instrument
the ASTER instrument was developed by Ministry of International Trade and Industry of Japan and Japan Resources Observation System Organization based on he requirements ASTER science team constituted with Japanese and American scientists.
The ASTER is consisted of three sensors such as the Visible and Near Infrared Radiometer (VNIR), The Short Wavelength Infrared Radiometer (SWIR) and Thermal Infrared Radiometer (TIR).
Table 2 ASTER Instruments Characteristic
| |
VNIR |
SWIR |
TIR |
| Range (mm) |
Band 1
0.52-0.60 |
Band 4
1.600-1.700 |
Band 10
8.125 -8.475 |
Band 2
0.63-0.69 |
Band 5
2.145-2.185 |
Band 11
8.475 -8.825 |
Band 3 N
0.76-0.86 |
Band 7
2.235-2.285 |
Band 12
8.925 -8.725 |
| |
Band 8
2.295-2.365 |
Band 14
10.95-11.65 |
| |
Band 9
2.360-2.430 |
|
| Ground Resolution |
15 m |
30m |
90m |
| Cross-track Pointing (deg.) |
24 |
8.55 |
8.55 |
| Swath Width(km) |
60 |
60 |
60 |
VNIR has the following characterists for DEV generation.
-
the stereoscopic capability in the along track direction
- the stereoscopic viewers are observed are observed by nadir and backward
- Base to Height Ratio (B/H) is fixed 0.6.
- Ground resolution is approximately 15 m.
- the backward viewing detector has tilt angle on the degree from along track direction. Because the
corresponding area with nadir image and backward image is made sure of maximu, by compensation for the earth rotation.
2. Polity of DEM Generation
We have had the following policies on DEM generation.
-
DEM are generated operationally, if it is possible.
- GCP aren't utilized with a basis.
- Coarse DEM are utilized with a basis.
- DEM of 30m grid are generated.
- DEM generation of Southeast Asia area is given high priority.
- DEM are revised to higher accuracy data using multiple observation data.
By these policies, many relative DEM products will be possible generated. But some problems are caused.
- As ASTER is optical sensor, it is impossible to generate DD DEM in cloud area, etc. Therefore, these area has to be detected automatically.
- For assure to quality of product, abnormal DEM data that caused by mismatching have to be detected automatically.
- The accuracy of elevation data depends on pointing knowledge of sparcecraft because GCP isn't utilized.
Therefore, the performance of pointing knowledge had be clear.
- Coarse DEM is desired one of high accuracy. As the best DEM product is DTED made by Defense Mapping Agency of USA, it isn't available. So, next better product is GLOBE product of DCW product. As the GLOBE product is 1km Grid data, GLOBE will be selected as coarse DEM.
3. The Processing Algorithm
The general flow chart shows in figure 1 and figure 2.
Figure 1 General Flow Chart (1)
Figure 2 General Flow Chart (2)
This algorithm will be implement in ASTER Ground Data System.
3.1 Orientation Parameter
Orientation data are defined using TDRSS On-Board Navigation System (TONS) data. Therefore, knowledge of Satellite position is very well than one of ordinary satellite. The performance of EOS-AM1 shows follow.
Table 3 TONS Performance
| |
normal |
Worst |
| Radial Position (m) |
5 |
15 |
| Intrack Position (m) |
30 |
100 |
| X-Track Position (m) |
20 |
50 |
| X-Track Velocity (m/s) |
0.022 |
0.056 |
Table 4 Pointing Accuracy at Instrument Interface
| Worst Case (arcsec, 3s) |
Roll |
Pitch |
Yaw |
| Dynamic |
12.1 |
14.2 |
12.9 |
| Static |
55.6 |
56.0 |
85.1 |
3.2 Stereo Matching Method
We adopted the stereo matching method using area correlation technique. In a search of correspondence point, We utilized stereo matching method on three stages. Because it was a purpose that process time is saved and mismatching is reduced. So, First correspondence of tie point is examined using SSDA method. Second and Third correspondence is examined using correlation method.
3.3 Cloud and Water Area Extraction
The area where might be occurred mismatching such as cloud and water will be extracted using 3 data (Visible and Near Infrared). Cloud area is extracted using threshold.
Water area extracted using threshold, variation in specific window and coarse DEM.
- Land-water check are carried out using coarse DEM.
- If a coarse DEM is marked as water, calculate the
variation in specific window. While variation is less than the threshold value, the area is extracted as water area.
3.4 Abnormal Value Extraction
Abnormal elevation data is checked using calculated DEM. If land inclination calculated from neighborhood elevation data is larger than a threshold, the DEM is defined as abnormal value. Also if the elevation data is out of between the specified minimum elevation and the specified maximum elevation, the DEM is defined as abnormal value.
3.5 DEM Browse Generation
The shaded image I generated from DEM data. Then the shaded data is compressed by JPED compression.
3.6 Data Conversion
DEM generated using stereo matching method is located on every other pixels of ASTER Leavel 1Adata. It's unit is the earth fixed coordination. The Z set is converted from the earth fixed coordination to map projection such as LAT/LON, UTM, Merchator, and so on.
If GCP are available, DEM is corrected using GCP. The orientation data is not corrected, and DEM is corrected directly by bias data calculated from GCP and DEM.
4. ASTER DEM Product
Two kinds of DEM products will be distributed.
-
XYZ set product
- Z set product
4.1 XYZ Set
XYZ set product is the Data Set that consisted of relative elevation data, correlation value and quality flage.
The elevation is calculated by stereo matching, and tie point are corresponded with the locate of every other pixels of input image data (ASTER level 1A data)
the coordinate system is the earth fixation coordinate system. Therefore, it isn't data but random data on the earth fixation coordinate system.
The quality flags are the index data which are indicated the area where elevation was abnormal, matching impossibility area and so on.
The correlation value is also one of quality index. If correlation value is low value, probably the mismatching is occurred.
XYZ set DEM is planed to be processed 30 products per day on ASTER GDS.
4.2 Z set
Z set product consists of elevation data made from XYZ set DEM and quality flage.
Elevation data is converted grid data on map projection.
Such as UTM, polarstereo, latitude longitude. When GCP exists, we can make the absolute DEM.
By using quality flags, the no-calculated DEM data is able to be interpolated from calculated DEM data of neighbor data, or converted into a fixed value.
DEM provided by observation of multiple times is possible to be upgraded to high accuracy data. Also DEM mosicing is available using the continuous scene data.
