首页> 外文期刊>Remote Sensing of Environment: An Interdisciplinary Journal >Stem volume retrieval in boreal forests from ERS-1/2 interferometry
【24h】

Stem volume retrieval in boreal forests from ERS-1/2 interferometry

机译:ERS-1 / 2干涉法在北方森林中提取茎量

获取原文
获取原文并翻译 | 示例
           

摘要

C-band repeat-pass interferometry, in particular, the coherence, has been shown to be of great potential for stem volume retrieval. For boreal forests, we have investigated a stem volume retrieval method based on inversion of FRS- 1/2 coherence measurements by means of a semiempirical model. A multitemporal combination of several stem volume estimates has been used in order to reduce errors in the estimation. The retrieval procedure was first applied in a forest estate located in Kattbole, Sweden, where accurate in situ measurements were taken. Stem volume was determined both at the stand level (between 2 and 14 ha) and at the pixel level (25 x 25 m). A multitemporal combination of coherence data acquired in stable winter-type conditions gave the most accurate results. Based on the results obtained in Kattbole, the retrieval procedure was extended to a large area of 4235 km{sup}2 around Kattbole. Retrieval was performed in all forested areas on a pixel basis (25 × 25 m), generating stem volume maps. In Kattbole, at the stand level, stem volume up to 350 m{sup}3/ha was estimated with an error comparable to the ground truth, i.e. 10 m{sup}3/ha. At the pixel level, the error reached the value of 55 and 71 m{sup}3/ha in the forest estate and in the large area, respectively. Compared to the results from the stand analysis, the higher error is believed to be mainly due to the higher uncertainty of coherence estimation at high stem volume and to geometric mismatch between field data and coherence data. Moreover, over large areas, spatial variation of the parameters in the model should be considered.
机译:C波段重复通过干涉测量法,特别是相干性,已被证明在茎体积检索中具有巨大潜力。对于北方森林,我们研究了基于半经验模型的基于FRS-1 / 2相干性测量值反演的茎体积检索方法。为了减少估计中的误差,已使用几个茎体积估计的多时间组合。取回程序首先应用于瑞典卡特博勒的一个森林庄园,在那里进行了准确的原位测量。在支架水平(2到14公顷之间)和像素水平(25 x 25 m)上都确定了茎的体积。在稳定的冬季类型条件下获得的多时相干数据组合给出了最准确的结果。根据在Kattbole中获得的结果,将检索过程扩展到Kattbole周围4235 km {sup} 2的大区域。在所有林区中以像素为基础(25×25 m)进行检索,生成茎体积图。在Kattbole的展位上,估计的茎干高达350 m {sup} 3 / ha,其误差可与地面实况相媲美,即10 m {sup} 3 / ha。在像素水平上,在林区和大面积地区,误差分别达到55和71 m {sup} 3 / ha。与林分分析的结果相比,认为较高的误差主要是由于在高茎体积下相干估计的较高不确定性以及田间数据和相干数据之间的几何不匹配。此外,在大面积上,应考虑模型中参数的空间变化。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号