A method for correcting seal-borne oceanographic data and application to the estimation of regional sea ice thickness
A method for correcting seal-borne oceanographic data and application to the estimation of regional sea ice thickness
复制标题
DOI：
10.1016/j.jmarsys.2018.08.002
复制
发表时间：
2018-11-01
影响因子：
2.8
通讯作者：
Costa, Daniel P.
中科院分区：
文献类型：
--
作者：
Frazer, Eamon K.;Langhorne, Pat J.;Costa, Daniel P.
与低纬度地区相比，南极洲周围高纬度地区的海洋样本数量明显不足。基于哺乳动物的仪器，如电导率-温度-深度卫星中继数据记录仪（ctd - srdl）提供了一种可能的解决方案。不幸的是，这些都受到仪器的绝对盐度的影响。本研究调查了2011年在罗斯海西南部安装在威德尔海豹（Leptonychotes weddellii）上的ctd - srdl收集的一组卫星传输数据。发现装置之间未经校正的盐度偏移高达1.4 g kg(-1)，使数据不适合某些未经校正的海洋学研究。在这里，开发了一种校正方法，该方法使用被认为位于同一位置的成对ctd - srdl的剖面，如果它们在规定的时间和空间窗口内发生，则对同一水体进行采样。利用最小二乘法，找到了共置对偏移量矩阵的最佳拟合解，减少了ctd - srdl之间的盐度偏移量。这些偏移量比原始偏移量小10倍。校准后的参考仪器与一些设备放在一起，进一步提高了所有ctd - srdl的绝对精度。利用校正后的CTD-SRDL资料，估算了海冰形成对海水中盐分的抑制作用，并推导了罗斯海西南部海冰厚度的时间演化。我们对区域海冰厚度的估计与2011年11月在有限区域内直接测量的海冰厚度一致，进一步证实了我们的方法。
The high-latitude oceans surrounding Antarctica are substantially undersampled compared to lower latitudes. Mammal based instruments such as Conductivity-Temperature-Depth Satellite Relay Data Loggers (CTD-SRDLs) present one possible solution. Unfortunately, these are subject to instrument-dependent offsets in absolute salinity. This study investigates a set of satellite-transmitted data collected by CTD-SRDLs mounted on Weddell seals (Leptonychotes weddellii) in the South-western Ross Sea in 2011. The uncorrected salinity offset between devices was found to be up to 1.4 g kg(-1), making the data unsuitable for some oceanographic studies without correction. Here, a correction method was developed that uses profiles from pairs of CTD-SRDLs that are considered to be co-located and to sample the same body of water if they occur within defined time and space windows. Using least squares, a best-fit solution to the matrix of offsets in co-located pairs was found that reduces salinity offsets between the CTD-SRDLs. These offsets are smaller than the original offsets by a factor of 10. A calibrated reference instrument, that was co-located with some of the devices, provided further improvement in the absolute accuracy of all the CTD-SRDLs. Using the corrected CTD-SRDL data we estimate the rejection of salt into the water column by sea ice formation, and derived the time evolution of sea ice thickness in the South-western Ross Sea. Our estimates of regional sea ice thickness are in agreement with direct sea ice thickness measurements taken over a limited area in November 2011, providing further affirmation of our method.