Estimating lake ice thickness in Central Ontario
Estimating lake ice thickness in Central Ontario
复制标题
估算安大略省中部的湖冰厚度
DOI：
10.1371/journal.pone.0208519
复制
发表时间：
2018
期刊：
影响因子：
3.7
通讯作者：
S. Howell
中科院分区：
文献类型：
--
作者：
Justin Murfitt;L. Brown;S. Howell
湖泊是加拿大的一个重要地理特征，对区域气候有影响。在冬季，它们是雪上摩托车和冰钓等娱乐活动的重要场所，也是北方社区重要的供应路线的一部分。准确报告湖冰特性（如厚度）的能力至关重要，然而，加拿大的报告不足，并且缺乏温带地区（如安大略中部）的湖冰厚度记录。在这里，我们评估应用先前开发的温度模型和雷达卫星-2估计湖冰厚度在中央安大略，并提供洞察该地区长期冰厚度的变化。ALS环境科学浅水冰剖面仪（SWIP）用于验证温度和雷达模型。结果表明，传统的方法，使用温度来预测冰的厚度在冰的生长具有较低的RMSE值为2.3厘米和大于0.9的相关性。用雷达卫星2号估算冰厚的R2值为0.6（p < 0.01），但RMSE值高达11.7 cm。雷达卫星2号方法的不确定性可能与散射机制和雷达信号与中纬度湖冰的相互作用等尚未探索的问题有关。将优化的温度模型应用于长期温度记录显示，冰盖每十年变薄0.81厘米。
Lakes are a key geographical feature in Canada and have an impact on the regional climate. In the winter, they are important for recreational activities such as snowmobiling and ice fishing and act as part of an important supply route for northern communities. The ability to accurately report lake ice characteristics such as thickness is vital, however, it is underreported in Canada and there is a lack of lake ice thickness records for temperate latitude areas such as Central Ontario. Here, we evaluate the application of previously developed temperature models and RADARSAT-2 for estimating lake ice thickness in Central Ontario and provide insight into the regions long term ice thickness variability. The ALS Environmental Science Shallow Water Ice Profiler (SWIP) was used for validation of both temperature and radar-based models. Results indicate that the traditional approach that uses temperatures to predict ice thickness during ice growth has low RMSE values of 2.3 cm and correlations of greater than 0.9. For ice decay, similar low RMSE values of 2.1 cm and high correlations of 0.97 were found. Using RADARSAT-2 to estimate ice thickness results in R2 values of 0.6 (p < 0.01) but high RMSE values of 11.7 cm. Uncertainty in the RADARSAT-2 approach may be linked to unexplored questions about scattering mechanisms and the interaction of radar signal with mid-latitude lake ice. The application of optimized temperature models to a long-term temperature record revealed a thinning of ice cover by 0.81 cm per decade.