首页 | 本学科首页   官方微博 | 高级检索  
     


Quantitative mapping of arid alluvial fan surfaces using field spectrometer and hyperspectral remote sensing
Authors:Onn Crouvi  Eyal Ben-Dor  Dov Avigad
Affiliation:a The Geological Survey of Israel, 30 Malkhe Israel Street, Jerusalem 95501, Israel
b The Institute of Earth Sciences, The Hebrew University, Givat Ram, Jerusalem 91904, Israel
c The Department of Geography and the Human Environment, Tel-Aviv University, Tel-Aviv 69978, Israel
Abstract:Mapping and dating of arid and semi-arid alluvial fans are of great importance in many Quaternary studies. Yet the most common mapping method of these features is based on visual, qualitative interpretation of air-photos. In this study we examine the feasibility of mapping arid alluvial surfaces by using airborne hyperspectral reflective remote sensing methodology. This technique was tested on Late Pleistocene to Holocene alluvial fan surfaces located in the hyperarid southern Arava valley, Israel. Results of spectral field measurements showed that the surface reflectance is controlled by two main surficial processes, which are used as relative age criteria: the degree of desert pavement development (gravel coverage %) controls the absorption feature depths, while the rock coating development influences significantly the overall reflectance of the surface, but its effect on the absorption feature depths is limited. We show that as the percent of the surface covered by gravels increases, the absorption feature depth of the common gravels, in this case carbonate at 2.33 μm, increases as well; whereas the absorption features depth of the fine particle in-between the gravels, decrease (hydroxyl and ferric absorption features at 2.21 μm, and 0.87 μm, respectively), as the fines are removed from the surface. Using these correlations we were able to map the surface gravel coverage (%) on the entire alluvial fan, by calculating the gravel coverage (%) in each pixel of the hyperspectral image. The prediction of gravel coverage (%) is with accuracy of ± 15% (e.g. gravel coverage of 50% can be predicted to be 35% to 65%). Using extensive accuracy assessment data, we show that the spectral based mapping maintained high accuracy degree (R2 = 0.57 to 0.83). The quantitative methodology developed in this study for mapping alluvial surfaces can be adapted for other surfaces and piedmonts throughout the arid regions of the world.
Keywords:Quantitative hyperspectral remote sensing mapping  Alluvial fan surfaces  Reflectance spectra  Desert pavement  Rock coating
本文献已被 ScienceDirect 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号