Science Data for the article “More Frequent Spaceborne Sampling of XCO2 Improves Detectability of Carbon Cycle Seasonal Transitions in Arctic-Boreal Ecosystems

These datasets are results from modeling experiments to test for the impact of spaceborne sampling of atmospheric CO2 on estimates of land-atmospheric CO2 exchange in the Arctic region. Specifically, we use flux inversion observing system sensitivity experiments (OSSE) to test the relative impact of low frequency sampling from low earth orbit (LEO) vs high frequency sampling from highly elliptical orbit (HEO). LEO sampling is based on the Orbiting Carbon Observatory 2 (OCO-2) mission, which has been in orbit since 2014. HEO sampling is based on the Arctic Fourier Transform Spectrometer Investigation (AURORA) missions, which is a new mission concept introduced in the corresponding paper. Our OSSEs use the JPL-developed Carbon Monitoring System Flux (CMS-Flux) 4-D variational (4D-Var) inversion system (Liu et al 2017) to optimize weekly estimates of net ecosystem exchange (NEE) and ocean fluxes at 4º × 5º. Two OSSEs are conducted, differing in assimilated data: (1) OCO-2 and (2) OCO-2+AURORA. Both OSSEs are run for a period of two years (2015-2016) to examine recovery of the mean seasonal cycle and flux anomalies. Details regarding the model setup can be found in Section 2.3 of the publication associated with this dataset. These files contain global prior, posterior, and true flux for each OSSE from 2015 to 2016 at daily 4º × 5º resolution. Separate files are provided for each year. Prior and True fluxes are the same for each OSSE and denoted in the file as ‘Prior’ and ‘True’, respectively. Posterior fluxes for OCO-2 and OCO-2+AURORA are denoted ‘Posterior_OCO2’ and ‘Posterior_NIR_OCO2’, respectively. Flux variables have dimensions of 365 x 46 x 72, representing days per year, latitude, and longitude. Daily values are based on interpolation from weekly inversion estimates; we thus recommend to average results to weekly or monthly bins. References Liu, J., Bowman, K.W., Schimel, D.S., Parazoo, N.C., Jiang, Z., Lee, M., Bloom, A.A., Wunch, D., Frankenberg, C., Sun, Y. and O’Dell, C.W., 2017. Contrasting carbon cycle responses of the tropical continents to the 2015–2016 El Niño. Science, 358(6360), p.eaam5690.

本数据集为一系列模拟试验的成果，旨在评估大气二氧化碳（atmospheric CO2）星载采样（spaceborne sampling）对北极地区（Arctic region）陆气二氧化碳交换（land-atmospheric CO2 exchange）估算结果的影响。
具体而言，本研究采用通量反演观测系统敏感性试验（flux inversion observing system sensitivity experiments，OSSE），对比低地球轨道（low earth orbit，LEO）的低频采样方案与高椭圆轨道（highly elliptical orbit，HEO）的高频采样方案的相对影响。其中LEO采样方案基于2014年起在轨运行的轨道碳观测站2号（Orbiting Carbon Observatory 2，OCO-2）任务；HEO采样方案则基于本配套论文中提出的新型任务概念——北极傅里叶变换光谱仪探测计划（Arctic Fourier Transform Spectrometer Investigation，AURORA）。
本研究的OSSE采用喷气推进实验室（JPL）研发的碳监测系统通量（Carbon Monitoring System Flux，CMS-Flux）四维变分（4-D variational，4D-Var）反演系统（Liu等，2017），以4°×5°的空间分辨率对净生态系统交换（net ecosystem exchange，NEE）与海洋通量的周度估算结果进行优化。
本研究共设置两组OSSE试验，二者的差异在于所同化的观测数据类型：(1) 仅同化OCO-2观测数据；(2) 同时同化OCO-2与AURORA观测数据。两组试验均运行2年（2015-2016），以评估平均季节循环与通量异常的恢复情况。关于模型设置的详细细节，可参阅本数据集配套论文的第2.3节。
本数据集包含2015至2016年期间，两组OSSE的全球先验通量、后验通量与真实通量数据，空间分辨率为每日4°×5°。数据按年份分为两个独立文件。两组试验的先验通量与真实通量完全一致，在数据文件中分别以"Prior"与"True"作为标识。OCO-2组与OCO-2+AURORA组的后验通量则分别以"Posterior_OCO2"与"Posterior_NIR_OCO2"作为标识。
通量变量的维度为365×46×72，依次对应年度天数、纬度与经度。由于日度数据是基于周度反演估算结果插值得到的，因此建议将结果聚合至周度或月度统计区间后再开展分析。
参考文献：Liu, J.、Bowman, K.W.、Schimel, D.S.、Parazoo, N.C.、Jiang, Z.、Lee, M.、Bloom, A.A.、Wunch, D.、Frankenberg, C.、Sun, Y.及O’Dell, C.W.，2017。热带大陆对2015-2016年厄尔尼诺事件的碳循环响应差异。《科学》，358(6360)，eaam5690页。