The Data Processing And Error Analysis For Satellite Laser Altimeter

Space-borne laser altimeter system is an active measure system in remote sensing. The range between satellite and target is calculated by the precise time of flight of laser pulse, which is extracted by the weak received signal, and the accurate location and elevation of laser footprint is acquired by combining with satellite precise orbit and attitude data. Ultimately, the DEM model covering the surface of the earth is built through the consecutive flight of satellite.The transmitted laser could penetrate the upper vegetation to obtain the three-dimensional terrain, which is practically impossible using the method of traditional photogrammetry. The divergence angle of laser is arc-second magnitude, hence the position and resolution of spatial are much more accuracy than that of microwave radar approach. The1064nm wavelength of laser located in atmosphere transmittance window could be directly reflected by the ice sheet or sea surface, almost no penetration effect, and the elevation accuracy can reach15cm. All above advantages make it widely used in the polar environmental monitoring of ice sheet and sea-ice change, the interannual variability of vegetation, and the ocean environment, etc. If the laser footprints served as the ground control points for remote sensing images, the large scale topographic map could generated with1:10000or even1:5000.The ICESat carrying the GLAS system was launched by NASA in2003, and the GLAS, the only satellite laser altimeter system used for global surface observation by far, operated continually for7years. The measurement error of GLAS was not only brought by the device itself, also included the environmental effects, such as atmospheric scattering and refraction, solid tide induced by precession and nutation, and the surface charactors, such as slope and roughness. Therefore, it is crucial to establish the measuring theory, error model, as well as the entire data processing procedure for designing the parameters of laser altimeter system, assessing the error or accuracy and ensuring the validity of data product.This thesis conducts complete received waveform model and error analysis research on those altimeters used to earth observation and with full recording waveform, accomplishes the entire data processing procedure from original received pulse to final elevation product, and programs the software codes of data batch proceesing. In aspects of received waveform theory, the received waveform model was developed, and especially the model of ocean surface was deduced. In terms of error analysis, the error model of laser ranging was developed and the elevation error model of laser footprint was established. In data processing, the final elevation data products was calculated and generated from underlying original echo data and auxiliary engineering data of GLAS system including the precise attitude, position data of satellite and the pointing direction of laser, and some methods of data processing were innovated. In data applications, combined with the NASA GLAS altimeter data, WMO NCEP meteorological data and DTU ALS lidar data, the parameters of sea surface such as wind speed and weight height were inversed, the continual7years from2003to2009elevation changing of ice sheet above2000m in Greenland was mornited, the method to locate cross track position was developed, and the elevation precision of GLAS footprint was assessed by the DEM map generated by the airborne lidar data. All of above application results were compared to the measured data ultimately.The comparison of data processing results and GLAS results shows that the deviations of final elevation could be controlled within10cm in the ice sheet region of smaller slope and roughness. Utilizing ALS airborne lidar data with better spatial resolution and elevation accuracy, combined with the established error model, it is proved that on the flat surface of ice sheets the elevation precision of GLAS system can meet its design value of10cm. The stimulated waveforms and the actual measured echos on ocean surface have great similarity, with parameters error less than6%. Both the inversed wind speed above sea surface compared to the NCEP meteorological data and the calculated mean sea level compared to that of TOPEX/Poseidon data correspond well. With the crossover and repeated laser footprints of GLAS data, the changes monitoring shows that the ice sheet above2000m in Greenland increased by average of3.80cm with standard deviation of0.91cm, and the changing trend calculated by the crossover and repeated methods agreed well on March from2003to2009. The number of repeated footprints was4-15times larger than that of crossover ones, however, the distribution or location were very inhomogeneous on the map. Consquently, when using ICEsat data to analyze the icesheet changing, the repeated method was fit for small districts with diameter within several kilometers, while crossover was more suitable for the large regions.The standard data product of GLAS including the data storage format, the data classification and conversion relationship among different level of data were described in the appendix, which may be helpful for understanding the data processing and application appeared in this thesis. The software of data processing and application based on Visual Stidio2008were also showed simply at the end. The window form applications and executable files were coded in Visual Basic and C++respectively, and this software integrates data reading, processing, analyzing as well as result evaluation.The data processing methods and key techniques in this paper have strong versatility on the laser altimeter used to earth observation, and it may be significant to the future space-borne laser altimeter system of China. Also, the echo theory and error model of space-borne laser altimeter systems may be used to the optimize the design of the system parameters and assess the measuring precisions.