Cloud motion vectors from MISR using sub-pixel enhancements

REMOTE SENSING OF ENVIRONMENT

卷:107

Article

Davies, Roger ; Horvath, Akos ; Moroney, Catherine ; Zhang, Banglin ; Zhu, Yanqiu

关键词: cloud motion vector;    atmospheric motion vector;    multi-angle remote sensing;    height-resolved wind;    geometric cloud height;    stereo matching;    wind remote sensing;   

DOI  :  10.1016/j.rse.2006.09.023

来源: Elsevier

【 摘 要 】

The operational retrieval of height-resolved cloud motion vectors by the Multiangle Imaging SpectroRadiometer on the Terra satellite has been significantly improved by using sub-pixel approaches to co-registration and disparity assessment, and by imposing stronger quality control based on the agreement between independent forward and aft triplet retrievals. Analysis of the fore-aft differences indicates that CMVs pass the basic operational quality control 67% of the time, with rms differences - in speed of 2.4 m/s, in direction of 17 degrees, and in height assignment of 290 m. The use of enhanced quality control thresholds reduces these rms values to 1.5 m/s, 14 degrees and 165 m, respectively, at the cost of reduced coverage to 45%. Use of the enhanced thresholds also eliminates a tendency for the rms differences to increase with height. Comparison of CMVs from an earlier operational version that had slightly weaker quality control, with 6-hour forecast winds from the Global Modeling and Assimilation Office yielded very low bias values and an rms vector difference that ranged from 5 m/s for-low clouds to 10 m/s for high clouds. (c) 2007 Elsevier Inc. All rights reserved.

【 授权许可】

Free