A web-based land surface remote sensing products validation system (LAPVAS): application to albedo product

摘要

Quantitative remote sensing product (QRSP) validation is a complex process to assess the accuracy and uncertainty independently using reference data with multiple land cover types and long-time series. A web-based system named as LAnd surface remote sensing Product VAlidation system (LAPVAS) is described in this paper, which is used to implement the QRSPs validation process automatically. The LAPAVS has two subsystems, the Validation Databases Subsystem and the Accuracy Evaluation Subsystem. Three functions have been implemented by the two subsystems for a comprehensive QRSP validation: (1) a standardized processing of reference data and storage of these data in validation databases; (2) a consistent and comprehensive validation procedure to assess the QRSPs' accuracy and uncertainty; and (3) a visual process customization tool with which the users can register new validation data, host new reference data, and readjust the validation workflows for the QRSP accuracy assessment. In LAPVAS, more than 100 GB of reference data warehoused in validation databases for 13 types of QRSPs' validation. One of the key QRSPs, land surface albedo, is selected as an example to illustrate the application of LAPVAS. It is demonstrated that the LAPVAS has a good performance in the land surface remote sensing product validation.