Autoregressive Prediction Method of E10.7 Index
-
摘要: 针对高层大气密度预报和轨道预报业务中对新型太阳紫外辐射指数E10.7的需求，基于TIMED-SEE观测仪器提供的0.1～105 nm太阳辐射强度数据，开展了E10.7指数反演和中期预报研究。 E10.7指数是太阳光谱中波长为0.1～105 nm的辐射流量，单位与F10.7指数相同（sfu，1 sfu=10–22 W·m–2·Hz–1）。 TIMED-SEE观测仪器提供的0.1～105 nm太阳辐射强度实测值具有高时间分辨率、延迟时间短和易获得的优势，利用最小二乘法拟合可反演出准实时的E10.7指数，均方根误差为5.445 sfu。利用高阶自回归模型对E10.7的中期预报效果尚佳，未来27天的预报值平均相对误差为7.83%。利用同样方法还开展了E10.7指数81天中心滑动平均值未来27天预报试验，未来27天的预报值平均相对误差仅为3.63%。Abstract: In response to the demand for the new solar ultraviolet radiation index E10.7 in the operation of upper atmospheric density prediction and orbit prediction, this paper carries out the research on E10.7 index inversion and medium-term prediction based on the integrated solar irradiance from 0.1 to 105 nm provided by TIMED-SEE instrument. The E10.7 index is the radiation flux in the solar spectrum at wavelengths from 0.1 to 105 nm in the same unit as the F10.7 index (sfu, 1 sfu = 10–22 W·m–2·Hz–1). The integrated solar irradiance from 0.1 to 105 nm provided by the TIMED-SEE instrument has the advantages of high time resolution, short delay time, and easy availability. The quasi-real-time E10.7 can be retrieved by least square fitting, and the root mean square error is 5.445 sfu. The average relative error of the next 27 days prediction is 7.83% using the autoregressive model. Using the same method, the 27 days prediction test of the 81-day central moving average of E10.7 is also carried out. The average relative error of the next 27-day predictions of the 81-day central moving average of E10.7 is 3.63%.
-
Key words:
- E10.7 index /
- Autoregressive model /
- Mid-term forecast
-
图 1 2002年2月7日至2006年6月30日的ETS和E10.7（红色点为异常数据）
Figure 1. ETS and E10.7 from 7 February 2002 to 30 June 2006 (The red points are abnormal)
图 2 2002年2月7日至2006年6月30日NOAA发布的E10.7与ETS日均值关系。R为二者的线性相关系数，RMSE为线性拟合均方根误差，红色实线为线性拟合结果
Figure 2. Relationship between E10.7 and ETS daily average values from 7 February 2002 to 30 June 2006. R is the linear correlation coefficient, RMSE is the root mean square error of linear fitting, and the red solid line is the result of linear fitting
图 3 2002年2月7日至2006年6月30日NOAA发布的E10.7日均值、利用式（3）计算的E10.7数据及滑动拟合得到的E10.7数据
Figure 3. E10.7 of NOAA calculaed by Eq.(3), and obtained by sliding window linear fitness from 7 February 2002 to 30 June 2006
图 4 2002年2月7日至2006年6月30日的T10.7的自相关（a）和偏自相关（b）结果（红色点为相关系数为0的点）
Figure 4. Autocorrelation function (a) and the partial autocorrelation function (b) of T10.7 from 7 February 2002 to 30 June 2006 (Red dot is the point with the correlation coefficient of 0)
图 5 2005年5月20日预报与观测的E10.7和E81C。 t为距2005年5月20日的天数，负值为距2005年5月20日的历史天数
Figure 5. E10.7 and E81C that are predicted and observed on 20 May 2005. Positive values in abscissa are the numbers of days after 20 May 2005, the negative values in abscissa are the numbers of days before 20 May 2005
图 6 2005年5月1日预报与观测的E10.7和E81C。 t为距离2005年5月1日的天数，负值为距2005年5月1日的历史天数
Figure 6. E10.7 and E81C that are predicted and observed on 1 May 2005. Positive values in abscissa are the numbers of days after 1 May 2005, the negative values in abscissa are the numbers of days before 20 May 2005
图 7 模型预测未来27天E10.7和E81C的平均逐日相对误差
Figure 7. Average daily relative error of E10.7 and E81C that are predicted in the next 27 days
表 1 2002年2月7日至2006年6月30日X级耀斑
Table 1. X class solar flares from 7 February 2002 to 30 June 2006
年–月–日 峰值时刻（UT） 耀斑等级 年–月–日 峰值时刻
（UT）耀斑
等级2002-04-21 01:51 X1.5 2003-11-04 19:50 X28.0 2002-05-20 15:27 X2.1 2004-02-26 02:03 X1.1 2002-07-03 02:13 X1.5 2004-07-15 01:41 X1.8 2002-07-15 20:08 X3.0 2004-07-15 18:24 X1.6 2002-07-18 07:44 X1.8 2004-07-16 02:06 X1.3 2002-07-20 21:30 X3.3 2004-07-16 10:41 X1.1 2002-07-23 00:35 X4.8 2004-07-16 13:55 X3.6 2002-08-03 19:07 X1.0 2004-07-17 07:57 X1.0 2002-08-21 05:34 X1.0 2004-08-13 18:12 X1.0 2002-08-24 01:12 X3.1 2004-08-18 17:40 X1.8 2002-08-30 13:29 X1.5 2004-10-30 11:46 X1.2 2002-10-31 16:52 X1.2 2004-11-07 16:06 X2.0 2003-03-17 19:05 X1.5 2004-11-10 02:13 X2.5 2003-03-18 12:08 X1.5 2005-01-01 00:31 X1.7 2003-05-27 23:07 X1.3 2005-01-15 00:43 X1.2 2003-05-28 00:27 X3.6 2005-01-15 23:02 X2.6 2003-05-29 01:05 X1.2 2005-01-17 09:52 X3.8 2003-06-09 21:39 X1.7 2005-01-19 08:22 X1.3 2003-06-11 00:02 X1.3 2005-01-20 07:01 X7.1 2003-06-11 20:14 X1.6 2005-07-14 1055 X1.2 2003-06-15 23:56 X1.3 2005-07-30 0635 X1.3 2003-10-19 16:50 X1.1 2005-09-07 17:40 X17.0 2003-10-23 08:35 X5.4 2005-09-08 21:06 X5.4 2003-10-23 20:04 X1.1 2005-09-09 03:00 X1.1 2003-10-26 06:54 X1.2 2005-09-09 09:59 X3.6 2003-10-26 18:19 X1.2 2005-09-09 20:04 X6.2 2003-10-28 11:10 X17.2 2005-09-10 16:43 X1.1 2003-10-29 20:49 X10.0 2005-09-10 22:11 X2.1 2003-11-02 17:25 X8.3 2005-09-13 19:27 X1.5 2003-11-03 01:30 X2.7 2005-09-13 23:22 X1.7 2003-11-03 09:55 X3.9 2005-09-15 08:38 X1.1 注 数据由美国国家海洋和大气管理局国家地球物理数据中心提供。 
下载: 导出CSV
表 2 测试集的E10.7和E81C预报误差
Table 2. Predicted error of E10.7 and E81C of all test sets
预报相对误差 平均值/(%) 最大值/(%) 最小值/(%) E10.7 7.83 15.08 2.40 E81 C 3.63 10.10 0.29
下载: 导出CSV
-
[1] WOODS T N. Recent advances in observations and modeling of the solar ultraviolet and X-ray spectral irradiance[J]. Advances in Space Research, 2008, 42(5): 895-902 doi: 10.1016/j.asr.2007.09.026 [2] EVANS J V. The Sun’s influence on the Earth’s atmosphere and interplanetary space[J]. Science, 1982, 216(4545): 467-474 [3] DONNELLY R. Gaps between Solar UV and EUV radiometry and atmospheric sciences[C]//Solar Radiative Output Variation, Cambridge Research and Instrumentation. Boulder, 1988: 139-145 [4] KOCKARTS G. Effects of solar variations on the upper atmosphere[M]//DOMINGO V. Physics of Solar Variations. Dordrecht: Springer, 1981: 295-320 [5] ROBLE R G, RIDLEY E C, DICKINSON R E. On the global mean structure of the thermosphere[J]. Journal of Geophysical Research, 1987, 92(A8): 8745-8758 doi: 10.1029/JA092iA08p08745 [6] TOBISKA W K. Validating the Solar EUV Proxy, E10.7[J]. Journal of Geophysical Research: Space Physics, 2001, 106(A12): 29969-29978 doi: 10.1029/2000JA000210 [7] STORZ M F, BOWMAN B R, BRANSON M J I, et al. High accuracy satellite drag model (HASDM)[J]. Advances in Space Research, 2005, 36(12): 2497-2505 doi: 10.1016/j.asr.2004.02.020 [8] TOBISKA W K, BARTH C A. A solar EUV flux model[J]. Journal of Geophysical Research: Space Physics, 1990, 95(A6): 8243-8251 doi: 10.1029/JA095iA06p08243 [9] TOBISKA W K, WOODS T, EPARVIER F, et al. The SOLAR2000 empirical solar irradiance model and forecast tool[J]. Journal of Atmospheric and Solar-Terrestrial Physics, 2000, 62(14): 1233-1250 doi: 10.1016/S1364-6826(00)00070-5 [10] 汪宏波, 赵长印. 不同太阳辐射指数对大气模型精度的影响分析[J]. 中国科学 G辑: 物理学 力学 天文学, 2009, 52(7): 1120-1128WANG HongBo, ZHAO ChangYin. Effects of various solar indices on accuracy of Earth’s thermospheric neutral density models[J]. Science in China Series G-Physics Mechanics & Astronomy, 2009, 52(7): 1120-1128 [11] 范磊, 韦飞, 符养, 等. 双通道辐射量计算太阳极紫外辐射E10.7指数方法初探[J]. 空间科学学报, 2013, 33(4): 403-409 doi: 10.11728/cjss2013.04.403FAN Lei, WEI Fei, FU Yang, et al. Use double channel solar irradiance flux to compute the E10.7 index[J]. Chinese Journal of Space Science, 2013, 33(4): 403-409 doi: 10.11728/cjss2013.04.403 [12] WOODS T N, BAILEY S, EPARVIER F G, et al. TIMED solar EUV experiment[J]. Physics and Chemistry of the Earth, Part C: Solar, Terrestrial & Planetary Science, 2000, 25(5/6): 393-396 [13] WOODS T N, EPARVIER F G, BAILEY S M, et al. Solar EUV Experiment (SEE): Mission overview and first results[J]. Journal of Geophysical Research: Space Physics, 2005, 110(A1): A01312 [14] 刘四清, 钟秋珍, 温靖, 等. 太阳10.7 cm射电流量中期预报模型研究(I)[J]. 空间科学学报, 2010, 30(3): 1-8 doi: 10.11728/cjss2010.01.001LIU Siqing, ZHONG Qiuzhen, WEN Jing, et al. Model research of 10.7 cm solar radio flux 27-day forecast (I)[J]. Chinese Journal of Space Science, 2010, 30(3): 1-8 doi: 10.11728/cjss2010.01.001 [15] 丁裕国, 江志红. 气象数据时间序列信号处理[M]. 北京: 气象出版社, 1998DING Yuguo, JIANG Zhihong. Time Series Signal Processing of Meteorological Data[M]. Beijing: China Meteorological Press, 1998 -
下载:
点击查看大图
计量
- 文章访问数: 88
- HTML全文浏览量: 31
- PDF下载量: 36
- 被引次数: 0
