DEM Accuracy Derived from ASTER Data
5. Estimation of DEM Accuracy Using ASTER Simulation Data
The accuracy of DEM is estimated the two steps.
In first step, the estimation is carried out using simulation data. The second step, the estimation is caried out using geometric model in additonal to noise data.
5.1 ASTER Simulation Data Generation
ASTER simulation data is generated using JERS1 OPS data and DEM data derived from topographic map. In this paper, it is used that 50m grid DEM provided from Geogrpahical Survey Institute of Japan.
-
Original DEM is used to elevation data to be gnerated from : 1:2500 topographic may by Geographic Survey Institute of Japan.
- Image data is resample fromJERS1 data.
- Orientation image data is calculated from JERS1 data.
- ASTER backward image data is calculated from DEM and orthoimage data.
Figure 3 shows the generated ASTER simulation image data. The characteristics are follows.
- nadir and backward image stereo pair data are generted.
- B/H is 0.6 on along track.
- ground resolution is approximately 15m.
- orbit data are adopt the nominal value.
Figure 3 ASTER simulation data
5.2 Results of DEM Generation
The result of stereo matching shows in table 5 and figure 4.
Figure 4 Shaded image of DEM derived from ASTER Simulation Data
Figure 5 Shaded image of DEM derived from ASTER Simulation Data
Theoretical accuracy of DEM is defined by next equation.
Accuracy = (Ground Resolution)/(base and height ratio)* (matching) Accuracy)
In this case, if matching accuracy is one pixel, DEM accuracy is 25m. And if matching accuracy is half pixel, close to theoretical value of half pixel precision.
Table 5 Accuracy of Elevation
| Positive Maximum Error (m) |
Negative Maximum Error (m) |
Average Error (m) |
R.M.SE.(m) |
| 175 |
-199 |
0.87 |
12.31 |
5.3 Estimation of DEM Accuracy
We estimated the accuracy of ASTER DEM data through ASTER simulation data were processed. When there is no error at orientation elements and stereo matching, the precision of DEM is decided by B/H and ground resolution.
As this case has no error in orientation data, the cause of error seems to be mismatching.
The orientation elements are decided by system configuration. The position knowledge of EOS-AM1 spacecraft becomes high to utilize data of TONES satellite compared with others spacecraft. Therefore orientation error is not mainly cause in elevation error.
And it is easy to be matching on stereoscopic image in along track direction, compared with one on cross track direction. Therefore the accuracy of DEM using real ASTER data will be not change one of simulation data.
5.4 The Cause of Mismatching
The area where mismatching are occurred are estimated that JERS-1 data were processed about 30 scenes include desert zone tropical rain forest zone, a grassy plain steep mountain zone. As a result, the area where mismatching are occurred had the next characteristic.
-
when the same intensity level
- when resemblance pattern continues
- when saturation data exist
- when area of the sea cloud level exists
- shadow department of clouds
JERS-1 data has the capability that are B/H = 0.3 ground resolution = 18 m and 24m. In case of JERS-1 data, the accuracy of DEM is settled 603m-80m. it's average is estimated with 30m-40m.
The results of experiment were able to get several characters near by a theory value, showed in the table.
On the other hand, ASTER data has the capability that are B/H = 0.6, the ground resolution = 15 m. in case of ASTER data the accuracy of DEM is settled 12.31 it's near a theory value.
5.5 Result of Cloud and Water Area Extraction
The result of water area detection shows in figure 6. Black color area is water area. The take area is extracted. Well
Figure 6: Extraction Water Area
Estimation DEM Accuracy using Aster Model Data
STER geometric model is designed. After that, The DEM curacy is estimated by addition of orientation error.
Geometric parameter is used nominal ASTER orbital data.
1. Experiments Method
Orbital data error is applied for each orientation parameters such as location data (X,Y,Z) and orientation data (w,F,k). The maximum of error is applied from the orbital stability of EOS-AM1. The orbital position is added the error by 30m steps until +-150m. And orbital orientation is added on the error by 15 arcsec until- arcsec. Also
error value is added is added on nadir observation and /or ackward observation.
2. Results
The results show the next tables.
Table 6 addition Error on Nadir Data
| |
Along Track (m) |
Cross Track (m) |
Elevation (m) |
| X axis (-150m) |
-134.73 |
11.55 |
-223.86 |
| Y axis (+150m) |
134.74 |
-11.55 |
223.88 |
| X axis (-150m) |
3.91 |
-68.05 |
15.42 |
| Y axis (+150m) |
-3.91 |
68.05 |
-15.42 |
| Z axis (-150m) |
-0.22 |
3.25 |
-0.81 |
| Z axis (+150m) |
0.22 |
-3.25 |
36.05 |
| Roll (-90arcsec) |
9.14 |
-156.92 |
-35.17 |
| Roll (+90arcsec) |
-8.91 |
156.89 |
35.17 |
| Ritch (+90arcsec) |
310.03 |
-26.58 |
515.26 |
| Ritch (-90arcsec) |
-310.48 |
26.52 |
-515.26 |
| Yaw (-90arcsec) |
-13.55 |
-0.91 |
-22.50 |
| Yaw (+90arcsec) |
13.55 |
0.91 |
22.49 |
Table 7 addition Error on Backward Data
| |
Along Track (m) |
Cross Track (m) |
Elevation (m) |
| X axis (-150m) |
-0.21 |
11.25 |
223.86 |
| Y axis (+150m) |
0.21 |
-11.25 |
-223.88 |
| X axis (-150m) |
-3.91 |
-67.80 |
-15.53 |
| Y axis (+150m) |
3.90 |
67.80 |
+15.53 |
| Z axis (-150m) |
0.19 |
2.35 |
-135.28 |
| Z axis (+150m) |
-0.19 |
-.2.35 |
135.28 |
| Roll (-90arcsec) |
-9.06 |
-158.79 |
-42.23 |
| Roll (+90arcsec) |
9.30 |
158.80 |
43.13 |
| Ritch (-90arcsec) |
0.69 |
29.27 |
-691.15 |
| Ritch (+90arcsec) |
-0.69 |
-29.22 |
690.15 |
| Yaw (-90arcsec) |
4.85 |
83.49 |
26.51 |
| Yaw (+90arcsec) |
4.79 |
-83.49 |
-26.43 |
Table 8 Addition Error on nadir and Backward Data
| |
Along Track (m) |
Cross Track (m) |
Elevation (m) |
| X axis (-150m) |
-134.93 |
0.30 |
0.10 |
| Y axis (+150m) |
134.93 |
-0.31 |
-.11.88 |
| X axis (-150m) |
-0.01 |
-134.83 |
0.95 |
| Y axis (+150m) |
-0.01 |
134.82 |
0.96 |
| Z axis (-150m) |
0.04 |
0.90 |
-134.82 |
| Z axis (+150m) |
-0.04 |
-313.39 |
134.82 |
| Roll (-90arcsec) |
-0.15 |
313.87 |
-17.87 |
| Roll (+90arcsec) |
0.16 |
-3.79 |
17.99 |
| Ritch (-90arcsec) |
311.02 |
3.79 |
-174.88 |
| Ritch (+90arcsec) |
-310.87 |
82.67 |
175.15 |
| Yaw (-90arcsec) |
15.42 |
-82.66 |
4.02 |
| Yaw (+90arcsec) |
15.36 |
-82.66 |
-3.93 |
7. Future
ASTER DEM product's accuracy will be 12.5m that has the potentiality of drawing 1:100,000 scale topographic map.
The Alogorithm will be upgradiong, especially searching method of stero tie point, extraction method of mismatching area.
8. Acknowledgment
This projects has been supported by ERSDAC of Japan. The author thanks members DEM working group of Japanese ASTER science team for his comments and advise.
9. Reference
-
ERSDAC, 1995 Algorthm theorithm Basis Document for ASTER Level-1 data Processing.
- KUDVA, EOS-AMI Pointing performance spacecraft instrument interface, ASTER science team meeting in pasadena, 1996.
- KUDVA, TONS performance, ASTER, Science Team Meting in Pasadena, 1996.
- neeck, pointing stability performance, ASTER Science Team meeting in Pasadena, 1996.
- Tokunaga, Overview of DEM Product Generated by Using ASTER Data, ISPRS Archives XXXI B4,1996.
