Etude de la précision du satellite lidar GLAS-ICESat pour l'altimétrie des eaux continentales

N. Lemarquand (1) , N. Baghdadi (1) , Jean-Stéphane Bailly (1)

1 UMR TETIS - Territoires, Environnement, Télédétection et Information Spatiale

For the coming century, the control of water resources will be certainly the key of all the stakes for billions of human beings. Unfortunately a reduction in the number of stations is observed with a decline of measurements quality. Remote sensing, which saw the development of numerous satellite radar altimeters and more recently the launch of the satellite lidar ICESat, could be an interesting alternative for the study of the hydrological networks. The objective of this study is to estimate the potential of ICESat for monitoring continental water missions through the cases of the Lake Geneva (Switzerland and France) and rivers of Metropolitan France. Our first axis of study concerned the satellite-based assessment of ICESat on the Lake Geneva by comparing laser data to hydrological gauge water levels. Two hydrological stations (Chillon and Saint Prex) were used to evaluate the accuracy of ICESat elevations. First it was necessary that all data was in the same datum to conduct a consistent comparison. ICESat elevations, which are referenced in the Topex ellipsoid, were converted into orthometric elevations by a translation between Topex ellipsoid and WGS84 and then into the vertical reference IGN69 (RGF93) with the grid RAF98. The shots of water alone were then extracted track by track and the mean elevation calculated for each track was used for the comparison with reference elevations (hydrological gauges). The error RMS is 33 cm (-0.20 cm ± 0.21 cm) without any saturation correction. When the saturation correction is supplied and different from -999.000, the quality of water elevation data is improved : the error RMS is 14 cm (0.01 cm ± 0.10 cm). However GLAS temporal profiles show a slow progressive adaptation of GLAS sensor before proposing correct elevations. On the passage of ICESat from the land to water, the first spots elevations are higher than reference elevation and the following spots from 30 cm to 50 cm. The progressive return to the normal can last 0.2 s. It corresponds to 8 measurements and an adaptation distance of 1.360 km. When the transition footprints are excluded, the accuracy for the ICESat elevation measurements is 5 cm. Besides hydrological objects with a small size (small lakes, small rivers), which can not apply a margin of 1.5 km to remove transition footprints, could not be monitoring using ICESat with a good accuracy. Next the accuracy of ICESat was investigated on French rivers with a width larger than the size of ICESat footprint (about 55 m for the laser 3). The error RMS is 1.15 m (0.03 m ± 1.17 m) due to the time of ICESat adaptation on the passage from land to water. ICESat is not adapted for the monitoring of the continental water resource.

Mots clés: LAC; COURS D'EAU; DONNEE DE SATELLITE; ALTIMETRIE

Domaines: Sciences de l'environnement

N. Lemarquand, N. Baghdadi, Jean-Stéphane Bailly. Etude de la précision du satellite lidar GLAS-ICESat pour l'altimétrie des eaux continentales. Revue Française de Photogrammétrie et de Télédétection, 2011, 2 (194), p. 45 - p. 52. ⟨hal-00599265⟩