station during the morning of February 23rd, only 
36 hours after launch. The image telemetry 
recording equipment and data processing facilities in 
Toulouse performed extremely well: the first images 
could be displayed and film recorded within a few 
hours. 
Panchromatic images from HRV 1 and Multispectral 
images from HRV 2, acquired while the spacecraft 
was on a descending pass across Europe to North 
Africa, immediately confirmed the high quality of 
the data: no visible geometric distorsion, high 
resolution, perfect registration of the three bands 
in the multispectral mode and excellent signal to 
noise ratios. 
It was thus possible to present to the 
international press on February 26th a 
representative set of panchromatic images (Piedmont 
area in Italy, just south of the city of Turin, 
Nice-Monaco area on the Riviera coast in France) 
and multispectral images (particularly an excellent 
image of the Djebel Amour area, 200 km south of 
the city of Algiers). 
The check out phase of the SPOT 1 spacecraft 
started immediately and lasted for two and a half 
months. Among the various operations which took 
place during this period, the following should be 
noted: 
- operation of the two redundent tape recorders 
which allow SPOT to acquire images from any part 
of the world, 
- commissioning of the second central receiving 
station in Kiruna (Sweden) together with its 
associated archiving and preprocessing center, 
- collection of in flight calibration data for the 
24 000 individual detectors in each HRV (6 000 
detectors for each band), 
- verification of attitude control accuracy, 
- measurement of image geometric quality, 
including absolute location accuracy, internal 
distorsion, band to band registration, Modulation 
Transfer Function (MTF), etc., 
- fine-tuning of the orbit to make it coincide with 
the a-priori SPOT worldwide grid used for scene 
referencing. 
2 - Image quality assessment. 
One of the main objectives of the check-out 
phase was the detailed assessment of the quality of 
HRV's images, both radiometrically and 
geometrically. Although other civilian remote 
sensing instruments based on CCD linear array 
technology have been experimented in the recent 
years, such as the german MOMS, little experience 
is available on the caracteristics of images 
produced by such sensors. 
A detailed description of the method used for 
image quality verification was presented by G. 
BEGNI (BEGNI et al, 1986). The main results are 
presented here. 
2.1. Radiometric image quality 
- Signal to noise ratio was specified at 200 in 
each of the spectral band panchromatic, XS1 and 
XS2 and 260 for spectral band, XS3 (near 
infrared). Measured signal to noise ratios are equal 
or better than specified in each band for both 
HRVs with the exception of the panchromatic band 
of HRV2 where it stands at 110. This seems to be 
due to a partially structured noise appearing as an 
horizontal stripping of the images for which no 
explanation is available yet. 
- Detector equalization was planned to be 
performed with the on-board calibration unit. This 
proved to be difficult to use in a quantitative 
manner because of spikes related to defaults in the 
calibration unit focal plane. This unit is therefore 
used only for monitoring the detectors and 
associated electronic stability. Quantitative 
measurement of detector equalization (sometimes 
called normalization) was performed on the basis of 
analysis of wide, homogeneous, snow-covered scenes 
in the Antartic and Greenland, where averaging 
over many lines of each scene provides an excellent 
technique for detector equalization. This method has 
proved to be extremely powerfull, although it has a 
significant impact on operations because a large 
number of scenes over these areas needs to be 
collected; 
- Absolute calibration of the detectors is performed 
over the White Sands area in New Mexico (USA) 
with the help of Prof. SLATER' team of the 
University of Arizona. Results obtained in March 
1986 are shown below with a comparison with 
measurements performed in 1985 with the NOAA 
provided integrating sphere: 
HRV 1 
HRV 2 
.Integrating 
sphere 
White Sands 
Integrating 
sphere 
White Sands 
Panchromatic 
.68 
.0.61 
.61 
60 
XS i 
.60 
.36 
.62 
.55 
XS 2 
.47 
.0.41 
.50 
.44 
XS 3 
.70 
.0.56 
.70 
.58 
- Modulation Transfer Function is more difficult to 
measure quantitatively. A set of digitized aerial 
photographies of various urban sites had been 
established with various parametrized MTF before 
the launch. Comparison with actual SPOT 
panchromatic images over the same sites showed 
that the HRVl's MTF is better than specified, 
while HRV2's MTF is just as specified. The 
specifications are recalled below in the 
panchromatic mode at half the sampling frequency: 
PA 
line 0.27 
column 0.16 
2.2. Geometric image quality 
- Location accuracy. 
Each SPOT scene is supplied with geographical 
location parameters. For scenes processed at level 
1A and IB, the location parameters are estimated 
using orbit determination and attitude control. The 
r.m.s. value obtained is 860 m against the 
specifications of 1 500 m in the vertical viewing 
mode. 
For level 2, precision processed images using 
ground control points, the measured r.m.s. is 30 m 
against the specification of 50 m. 
- Image distorsion. 
For a given scene, image distortion can be 
characterized by scale variation (comparison between 
known and calculated distances between 2 points far^ 
apart within the image). The specification was 10 
for level IB images. The measured scale distorsion 
on actual images processed at this level (system 
corrected images) is 1.4 iO*" 3 . The intrinsic 
geometric quality of SPOT images is therefore 
excellent and confirm the inherent stability of CCD 
array detector imaging systems. 
- Band to band registration. 
For multispectral images, it is essential to have 
very accurate registration of images in the various 
spectral bands. Here again, the specification is 
easily met: 0.10 to 0.15 sampling intervals against 
0.30 specified. 
- Altitude determination in the stereoscopic mode. 
One of the most interesting applications of 
SPOT data relates to topographic mapping using 
stereo pairs of images taken from different orbits. 
Test sites in South Eastern France where very 
accurate ground survey is available are used to 
assess the ability of SPOT stereo images to 
properly derive altitude and horizontal coordinates