Some strange errors reported by Hyp3-AutoRIFT

[wangshuaicumt]

Hello, professor. When I used Hyp3-autorift to track the pixel offset of the Turkish earthquake, some strange errors occurred. As shown in the figure, I used Sentinel-2 data and the command was: hyp3_autorift \ "S2A_MSIL1C_20230120T082241_N0509_R121_T37SCB_20230120T091424" \ "S2A_MSIL1C_20230209T082111_N0509_R121_T37SCB_20230209T091429" The error reported is as follows:

[jhkennedy]

Hi @wangshuaicumt, thanks for the error report. HyP3 autoRIFT currently uses a set of paramter files which are optimized for ice-sheet and glacier applications by default -- typically this error is seen when there is no ice coverage in the image.

For other applications, there are alternative parameter files it the ITS_LIVE data bucket, which you can find with this AWS CLI command:

aws s3 ls --no-sign-request s3://its-live-data/autorift_parameters/v001/ | grep ".shp"

and pass it to hyp3-autorift using the --parameter-file option. For Turkey,

Note:

• You'll have to transform the paths to something GDAL can understand (/vsis3 or /vsicurl path)
• These parameter files are extremely :exclamation: experimental

Because they are experimental, we haven't documented them well -- we've got a backlog item to explore these more, document them, and provide this option to the HyP3 On Demand service, but I don't have a timeline for that right now.

[wangshuaicumt]

你好@wangshuaicumt，感谢您的错误报告。HyP3 autoRIFT 目前使用一组默认针对冰盖和冰川应用进行了优化的参数文件——通常当图像中没有冰覆盖时会出现此错误。

对于其他应用程序，ITS_LIVE 数据存储桶中有备用参数文件，您可以使用以下 AWS CLI 命令找到它们：

aws s3 ls --no-sign-request s3://its-live-data/autorift_parameters/v001/ | grep ".shp"

并将其传递给hyp3-autorift使用--parameter-file选项。对于土耳其，

笔记：

• 您必须将路径转换为 ​​GDAL 可以理解的内容（/vsis3或/vsicurl路径）
• 这些参数文件极其❗实验性

因为它们是实验性的，所以我们没有很好地记录它们——我们有一个积压项目来进一步探索它们，记录它们，并为 HyP3 On Demand 服务提供此选项，但我现在还没有时间表。

Hello, professor! Thank you for your help. I have successfully run the corresponding results, but I have a few questions.

For large earthquakes, the fault zones formed are often long and narrow, but the coverage of the Sentinel-2 data is limited. I would like to ask the professor how to solve this problem. My idea is to stitch and then track the pixel offset. Does Hyp3-AutoRIFT support multi-view image stitching, or will the corresponding function be developed in the future?