Calibration Result of JAXA standard products (As of September 28, 2007)
PRISM Level 1B2 data products
- Radiometric Accuracy
1) Relative Accuracy
Better than 0.4 % (better than 1DN) (RMS): Vertical streaking stripes may appear in some of the images with similar characteristics.
2) Absolute Accuracy
Better than 4.6 % (RMS)
* Evaluation method: Compared with calibrated AVNIR-2 over deserts, ocean etc.
- Geometric Accuracy
1) Absolute Accuracy
The evaluations were carried out about 1,390 GCPs (64 image scenes) for each radiometer.
| | Pixel direction
(cross track) | Line direction
(along track) |
Distance |
| Nadir view (RMS) |
6.5 m |
7.3 m |
9.8 m |
| Forward view (RMS) |
8.0 m |
14.7 m |
16.7 m |
| Backward view (RMS) |
7.4 m |
16.6 m |
18.1 m |
* Evaluation method: Compared with the GPS measured geolocation of GCPs
after projected onto the GRS 80 in correcting the height.
2) Relative Accuracy
| |
Pixel direction
(cross track) |
Line direction
(along track) |
Distance |
| Std. dev. in a scene(1σ) |
1.9 m |
2.3 m |
3.0 m |
AVNIR-2 Level 1B2 data products
- Radiometric Accuracy
1) Relative Accuracy
Better than 0.4% (better than 1DN) (RMS)
2) Absolute Accuracy
Band 1: better than 3.8% (RMS)
Band 2: better than 4.6% (RMS)
Band 3: better than 2.2% (RMS)
Band 4: better than 15.6% (RMS): Atmospheric influence accounts for the error as much as 50%.
* Evaluation method: Compared with MODIS sensors
onboard Terra/Aqua satellites over deserts, ocean etc.
- Geometric Accuracy (for all pointing angles)
1) Absolute Accuracy
| |
Pixel direction
(cross track) |
Line direction
(along track) |
Distance |
| RMS |
106 m |
19 m |
108 m |
* Evaluation method: Compared with the GPS measured geolocation of GCPs
after projected onto the GRS 80 in correcting the height.
2) Relative Accuracy
| |
Pixel direction
(cross track) |
Line direction
(along track) |
Distance |
| Std. dev. in a scene(1σ) |
4 m |
4 m |
6 m |
PALSAR Level 1.1/1.5 data products
- Radiometric Accuracy (for all off-nadir angles)
| Absolute Gain | 0.64dB(1σ) |
| Noise Equivalent Backscattering Coefficient | -34dB |
| VV/HH Gain Ratio (PLR) | 0.02dB(0.004dB:1σ) |
| VV/HH Phase Difference (PLR) | 0.32deg(1.01deg:1σ) |
| Cross talk (PLR) |
more than 31dB |
| Resolution |
Azimuth Direction (single look) |
4.49m |
| Range Direction (FBD, PLR, and DSN) |
9.6m(FBD, PLR, DSN) |
| Range Direction (FBS) |
4.8m(FBS) |
| Sidelobe |
Azimuth Direction |
-16dB |
| Range Direction |
-12.5dB |
| Two-Dimensional Direction |
-8.6dB |
* Evaluation method: Analyzed data from
Corner Reflectors and uniform forests over Amazon.
- Ambiguity
| Range Direction |
23dB |
| Azimuth Direction |
unmeasurable |
- Geometric Accuracy (for all off-nadir angles)
| 9.3 m (RMS) |
FBS, FBD, PLR |
| 70 m (RMS) |
WB1, WB2 |
* Evaluation method: Compared with the GPS measured geolocation of Corner
Reflectors after projected onto the GRS 80 in correcting the height.
<Future Improvement>
- Examination on PRISM stripe noises.
We have been developed an algorithm to reduce stripe noises of PRISM imagery, and it is now preparing as the updated processing software.
- Aging property of geometric accuracy evaluations of PRISM and AVNIR-2
- Aging property of absolute radiometric accuracy evaluation
Monitor the variation of absolute radiance to update the calibration coefficients.
<Previous Calibration Result>
Calibration results of JAXA standard products (As of March 29, 2007)
Initial calibration result of JAXA standard products (As of October 23, 2006)
|