Trends in downward surface shortwave radiation from multi‐source data over China during 1984–2015
Trends in downward surface shortwave radiation from multi‐source data over China during 1984–2015
复制标题
DOI：
10.1002/joc.6408
复制
发表时间：
2019-12
期刊：
影响因子：
--
通讯作者：
Zhigao Zhou;A. Lin;Lunche Wang;Wenmin Qin;Yang Zhong;Lijie He
中科院分区：
文献类型：
--
作者：
Zhigao Zhou;A. Lin;Lunche Wang;Wenmin Qin;Yang Zhong;Lijie He
明确地表太阳辐射的年代际变率对认识水文、生物过程和气候预测具有重要意义。然而，现有的研究表明，由于灵敏度漂移，不准确的校准和仪器更换，在中国上空观测到的SSR可能会有很大的差异和年代际变率的不均匀性。因此，本研究提出一种新的测站选择程序，以消除误差，并得出“真实”的SSR值。随后，两颗卫星回收SSR，包括云和地球辐射能系统-能量平衡和填充产品（CERES‐EBAF）（第4版）和全球能量和水循环实验-地面辐射收支（GEWEX-SRB）（3.0版）和三个再分析产品，包括国家环境预测中心-国家大气研究中心（NCEP‐NCAR），国家环境预测中心‐/能源部（NCEP-DOE）和现代研究和应用回顾性分析，版本2（MERRA‐2）使用“true”进行评价中国气象局39个同质站的SSR值，发现尽管所有五个产品都高估了SSR，两颗卫星的反演结果显示出更好的精度，CERES-EBAF的总体R为0.95，均方根误差（RMSE）为20.4 W m−2，平均绝对偏差误差（MAE）为14.9 W m−2，总体R为0.92，GEWEX-SRB在中国的RMSE为27.7 W m−2，MAE为21.2 W m−2。同时，本研究中观测趋势与两颗卫星检索趋势的相互比较表明，两颗卫星检索得出的新趋势（+0.78 W m-2 decade-1）与1994年至2015年的观测趋势（+0.92 W m-2 decade-1）具有良好的一致性。然而，由于1984年至1993年期间的灵敏度漂移和仪器更换，现场测量的SSR趋势（+5.8 W m−2 decade− 1）与两次卫星反演的SSR趋势（-3.7 W m−2 decade−1）仍然不一致。对1984 - 2015年中国SSR年代际变化的可能原因进行了分析，发现气溶胶光学厚度（AOD）和气溶胶-云相互作用的变化可能是1984 - 2015年中国SSR年代际变化的主要影响因素。
The clear knowledge of decadal variability of surface solar radiation (SSR) is of vitally significant for understanding hydrological and biological processes and climate prediction. However, existing studies have shown observed SSR over China may have large discrepancies and inhomogeneity in decadal variability due to sensitivity drift, inaccurate calibrations and instrument replacement. Therefore, a new procedure of station selection was proposed to eliminate errors and to derive “true” SSR values in this study. Afterward, two satellite retrieves of SSR, including Clouds and the Earth's Radiant Energy System‐energy balanced and filled product (CERES‐EBAF) (edition 4) and Global Energy and Water Cycle Experiment‐Surface Radiation Budget (GEWEX‐SRB) (Version 3.0), and three reanalysis products, including National Centers for Environmental Prediction‐National Center for Atmospheric Research (NCEP‐NCAR), national centers for environmental prediction‐/department of energy (NCEP‐DOE) and Modern‐Era Retrospective analysis for Research and Applications, version 2 (MERRA‐2) were evaluated using “true” SSR values at 39 homogeneous stations from the China Meteorological Administration and it was found that although all five products overestimated SSR, two satellite retrieves showed better accuracy with an overall R of 0.95, an root mean squared error (RMSE) of 20.4 W m−2 and mean absolute bias error (MAE) of 14.9 W m−2 for CERES‐EBAF and an overall R of 0.92, an RMSE of 27.7 W m−2 and MAE of 21.2 W m−2 for GEWEX‐SRB across China. Meanwhile, inter‐comparisons between trends of observations and trends of two satellite retrieves in this study showed that the new trends derived from two satellite retrieves (+0.78 W m−2 decade−1) were good agreement with trends of observation (+0.92 W m−2 decade−1) from 1994 to 2015. However, trends of SSR (+5.8 W m−2 decade−1) in situ measurements were still in disagreement with the trends of SSR (−3.7 W m−2 decade−1) derived from two satellite retrieves from 1984 to 1993 because of the sensitivity drift and instrument replacement in this period. The possible reasons for decadal variability of SSR in China were detected and it was found that variations in aerosol optical depth (AOD) and aerosol‐cloud interaction, rather than cloud, were suggested to be likely the main influencing factor of decadal variability of SSR across China from 1984 to 2015.