Evaluation of LiDAR-Derived Snow Depth Estimates From the iPhone 12 Pro
Evaluation of LiDAR-Derived Snow Depth Estimates From the iPhone 12 Pro
复制标题
对 iPhone 12 Pro 的 LiDAR 雪深估计的评估
DOI：
--
复制
发表时间：
2022
影响因子：
4.8
通讯作者：
C. Fletcher
中科院分区：
文献类型：
--
作者：
Fraser King;R. Kelly;C. Fletcher
雪是全球水能收支的重要贡献者，对春季洪水和水资源管理实践产生影响。激光测高[光探测和测距（LiDAR）]是一种遥感技术，已证明在监测积雪深度方面具有技能，但购买和运输传统LiDAR设备的费用限制了其业务使用。在这项工作中，我们证明了安装在Apple iPhone 12 Pro消费智能手机上的LiDAR传感器是一种实时手持测量仪器，可以准确观察雪深的变化。在加拿大南部安大略进行的两项独立的实地实验发现，与<italic>现场</italic>雪尺测量相比，iPhone LiDAR能够准确捕捉雪深的每日变化。<italic>在现场</italic>和激光雷达的比较xs<inline-formula><tex-math notation="LaTeX">$n=75$</tex-math></inline-formula>天在测量站点A表现出的相关性<inline-formula><tex-math notation="LaTeX">$r > 0.99$</tex-math></inline-formula>，平均绝对偏差小于1毫米，均方根误差（RMSE）约为6毫米。一个类似的正协议也注意到在第二个现场研究站点的<inline-formula><tex-math notation="LaTeX">$n=16$</tex-math></inline-formula>在同一时期的测量。激光雷达传感器的高精度表明，可以开发一种移动的应用程序，使用户能够在降雪事件之前和之后快速扫描积雪覆盖的区域，从而使用这些数据，通过基于公民科学的方法来测量积雪深度的变化，帮助填补目前的观测空白。
Snow is a critical contributor to the global water-energy budget with impacts on springtime flooding and water resource management practices. Laser altimetry [light detection and ranging (LiDAR)] is a remote-sensing technique that has demonstrated skill in monitoring snow depth, but the expense of purchasing and transporting traditional LiDAR equipment limits their operational use. In this work, we demonstrate that the LiDAR sensor installed on the Apple iPhone 12 Pro consumer smartphone is a real-time, handheld measurement instrument for accurately observing changes in snow depth. Two independent field experiments in Southern Ontario, Canada, found that the iPhone LiDAR was able to accurately capture daily changes in snow depth when compared to <italic>in situ</italic> snow ruler measurements. <italic>In situ</italic> and LiDAR comparisons of xs<inline-formula> <tex-math notation="LaTeX">$n=75$ </tex-math></inline-formula> days at measurement site A exhibit a correlation of <inline-formula> <tex-math notation="LaTeX">$r > 0.99$ </tex-math></inline-formula>, mean absolute bias less than 1 mm, and a root mean squared error (RMSE) of approximately 6 mm. A similar positive agreement was also noted at the second field study site for <inline-formula> <tex-math notation="LaTeX">$n=16$ </tex-math></inline-formula> measurements over the same period. The high accuracy of the LiDAR sensor suggests that a mobile application could be developed which allows users to quickly scan a snow-covered area before and after a snowfall event and consequently use this data to aid in filling current observational gaps through a citizen-science-based approach to measuring changes in snow depth.