QUANTITATIVE ASSESSMENT OF THE SAMPLING PROPERTIES OF A SPACEBORNE LIDAR (ATLID)

The European Space Agency plans to install the backscatter lidar syste m ATLID (atmospheric lidar) on a polar-orbiting platform at the beginn ing of the next century. This kind of active remote sensing will provi de highly accurate information about cloud-top height, which, in addit ion to collocated passive sounder's measurements of brightness tempera ture, might improve retrieved vertical temperature profiles and serve as a supplementation of present cloud climatologies. Due to technical constraints, ATLID will not provide spatially continuous information a bout cloud-top height. The representativeness of the lidar measurement s for the whole cloud field constitutes the sampling problem and is in vestigated in two steps: first, a scan mode for ATLID is developed, wh ich on the assumption that the cloud field is a two-dimensional random variable gives an equal pixel spacing along and across the flight tra ck of the orbiter. Second, the simulated lidar measurements given by t he elaborated scan mode are contributed to a spatially continuous clou d field represented by Advanced Very High Resolution Radiometer images . From the dispersed lidar measurements with a footprint diameter of a bout 1 km the cloud field is restored by a spatial interpolation schem e and compared with the original cloud field by a linear regression an alysis. It turns out that the sampling error and hence the benefits of ATLID strongly depend on the meteorological situation: if the require d vertical accuracy of the lidar measurement is about 250 m correspond ing approximately to half of the vertical resolution of present retrie val schemes, the probability for a meaningful ATLID information is bet ween 40% and 70%. Since an imager cannot provide a useful brightness t emperature in case of multilayered or broken clouds within one imager pixel, the synergism of ATLID with a passive instrument also depends o n the homogeneity of cloud-top height within the range of 1 km. To che ck this small-scale variability of cloud tops, data from the European Lidar Airborne Campaign 1990 are evaluated. Results show that for opti cally thick clouds the variability exceeds in 3% to 38% of all conside red cases a threshold of 250 m. Additionally, power-spectrum analyses confirm the result of the sampling analyses.