Calibration and Validation
Various global scientific issues (like climate change, environment monitoring, and ecological security) are making more and more strict requirements on the accuracy of remote sensing information products, which put forward very high accuracy and stability demands on the on-orbit calibration of remote sensors. Nowadays, on-board calibration for spaceborne sensors cannot reach the level of actually traceable to the ground-based radiometric primary standard, whereas field vicarious calibration can only obtain limited calibration accuracy since it is likely to be influenced by scaling effect, atmospheric condition, environment variation, etc. In recent years, ESA, USA, and China have successively proposed the essential concept of spaceborne radiometric benchmark sensors. The main idea of radiometric calibration based on this benchmark sensor is: upload the SI-traceable radiometric benchmark instrument in a small number of radiation benchmark satellites, then transfer the traceable radiation values from the benchmark satellite to other satellites to be calibrated. However, as the high-resolution spaceborne sensor is concerned, the cross-points (between the benchmark satellite and monitored satellite) can hardly be found under the strict matching condition when performing cross-calibration, because of high-resolution satellite’s narrow swath. So, this project will propose a new method of benchmark transfer calibration for the high-resolution space-borne sensor, which uses RadCalNet site measurement as the ground reference value. The new method is to solve the problem of increasing of cross-calibration error due to unavoidable relaxation of matching constraints to improve cross-point opportunities between high-resolution satellites. In this project, Chinese and European researchers dedicated to radiometric calibration will collaborate in the transfer calibration technologies based on RadCalNet and the SI-traceable spaceborne reference instrument. On one side, new method of radiometric benchmark transfer calibration which adopts RadCalNet measurement as ground reference value will be cooperatively developed to break through important technical problems existed in the benchmark transfer chain (benchmark satellite -> standard TOA spectral reflectance provided by RadCalNet -> satellite to be calibrated (i.e., the monitored satellite)). On the other side, based on previous cooperative research on RadCalNet, both parties will make effort to further improve the accuracy of RadCalNet standard product and the inter-site product consistency, and incorporate more Chinese automated calibration sites into the RadCalNet framework if possible, to carry out demonstration applications of automated calibration & benchmark transfer calibration for Chinese and European high-resolution satellites. The main research contents of this project include constraint mechanism analysis of the radiometric benchmark transfer calibration; a new method of the radiometric benchmark transfer calibration; demonstration of radiometric calibration for Chinese and European high-resolution satellites. The expected achievements of this project include (1) constraint mechanism of the radiometric benchmark transfer calibration based on RadCalNet; (2) new method of the radiometric benchmark transfer calibration; (3) technical report on the demonstration of transfer calibration for high-resolution satellites; (4) academic papers and talent training. In the project executing process, a series of external projects can be available to effectively support the operation of this project. Relevant supporting projects include: (a) Spaceborne radiometric benchmark transfer calibration and its ground-based validation; (b) Global automated radiometric calibration network; (c) Land satellite calibration network; (d) Radiometric re-calibration of thermal infrared band of land satellite and surface temperature retrieval.
