设为首页
加入收藏
Retrieval of Atmospheric Boundary Layer Height from Ground-based Microwave Radiometer Measurements
-
摘要: 采用2013年中国科学院大气物理研究所香河大气综合观测试验站的地基微波辐射计和激光雷达观测数据，以激光雷达探测的大气边界层高度为参考，分别利用非线性神经网络和多元线性回归方法建立微波亮温直接反演大气边界层高度的算法，并对比两种方法的反演能力, 同时分析非线性神经网络算法在不同时段及不同天气状况下反演结果的差异。结果表明:非线性神经网络算法的反演能力优于多元线性回归算法，其反演结果与激光雷达探测的大气边界层高度有较好一致性，冬、春季的相关系数达到0.83，反演精度比线性回归算法约高26%;对于不同时段和不同天气条件，春季的反演结果最好，晴空的反演结果好于云天; 四季和不同天气状况的划分也有利于提高反演精度。Abstract: Atmospheric boundary layer is a key parameter for boundary layer studies, including meteorology, air quality and climate. The atmospheric boundary layer height estimates are inferred from local radiosonde measurements or remote sensing observations from instruments like laser radar, wind profiling radar or sodar. Methods used to estimate atmospheric boundary layer height from radiosonde profiles are also used with atmospheric temperature and humidity profiles retrieved by microwave radiometers. An alternative approach to estimate atmospheric boundary layer height from microwave radiometer data is proposed based on microwave brightness temperatures, instead of retrieved profiles. Using the ground-based microwave radiometer and laser radar atmospheric boundary layer height obtained in 2013 at Xianghe Station, algorithms for retrieving atmospheric boundary layer height from 14-channel microwave brightness temperatures are developed based on the nonlinear neural network and multiple linear regression methods. The atmospheric boundary layer height is derived from laser radar backscattering data using the algorithm that retrieves the most significant gradients in profiles using gradient method. Root mean square errors (RMSEs) and correlation coefficient with two kinds of method are obtained to analyze which method is better through comparison. Retrieval results with the neural network method are compared in different periods of time and weather conditions. It shows that neural network algorithm is better than the multiple linear regression algorithm because results are more consistent with the observation. The correlation coefficient between the lidar-detected and neural network algorithm retrieved boundary layer height is 0.83, which is about 26% higher than the multiple linear regression algorithm retrieved result. Also, RMSEs of the neural network algorithm retrieved values (268.8 m) are less than the multiple linear regression algorithm retrieved values (365.1 m). For different time periods and weather conditions, retrievals in spring are best of four seasons, retrievals in the clear sky are better than those in the cloudy sky. But RMSEs in the cloud sky are less than those in the clear sky. Overall, correlation coefficients in four seasons are close to 0.80. It suggests that in order to improve the retrieval precision, specific retrievals under different conditions (such as different seasons and different skies) should be carried out.
-
图 1 验证激光雷达探测边界层高度
(a)2013年6月13日07:15香河站气溶胶消光系数廓线, (b)2013年6月13日07:15北京市观象台探空位温廓线
Fig. 1 Verification of laser radar boundary layer height
(a) aerosol extinction coefficient profile at Xianghe Station at 0715 BT 13 Jun 2013, (b) potential temperature profile from sounding at Beijing Weather Observertory at 0715 BT 13 Jun 2013
图 2 2013年香河站激光雷达探测大气边界层高度与北京市观象台探空廓线计算大气边界层高度散点图
Fig. 2 Scatte-plots of laser radar boundary layer height and sounding boundary layer height in 2013
图 3 2013年8月24日香河站激光雷达气溶胶消光系数的时间-高度剖面图
(黑点为梯度法计算的大气边界层高度)
Fig. 3 Time-height section of laser radar aerosol extinction coefficient at Xianghe Station on 24 Aug 2013
(black dots denote boundary layer height derived with the gradient method)
图 4 2013年香河站激光雷达探测与微波辐射计测试样本反演边界层高度
(a) BP神经网络算法，(b) 多元线性回归算法
Fig. 4 Scatte-plots of laser radar boundary layer height and ground-based microwave radiometer boundary layer height for test data samples at Xianghe Station in 2013
(a) BP neural network, (b) multiple linear regression
图 5 2013年香河站地基微波辐射计和激光雷达大气边界层高度比较
(a) BP神经网络算法, (b) 多元线性回归算法
Fig. 5 Comparisons of boundary layer height derived from ground-based microwave radiometer and laser radar at Xianghe Station in 2013
(a) BP neural network, (b) multiple linear regression回归
图 6 河站晴空和云天红外辐射亮温变化
(a)2013年1月1—2日 (晴空), (b) 2013年4月19—20日 (云天)
Fig. 6 Variation of infrared brightness temperature under different weather conditions at Xianghe Station
(a) clear sky during 1-2 Jan in 2013, (b) cloudy sky during 19-20 Apr in 2013
图 7 香河站地基微波辐射计反演大气边界层高度与北京市观象台探空廓线计算大气边界层高度散点图
(a)2013年检验, (b)2014年检验
Fig. 7 catte-plots of ground-based microwave radiometer boundary layer height at Xianghe Station vs sounding boundary layer height at Beijing Weather Observatory
(a) the test in 2013, (b) the test in 2014
表 1 不同季节反演结果的统计比较
Table 1 Statistic comparison of retrieval results for different seasons
季节 训练样本量 测试样本量 偏差/m 均方根误差/m 标准差/m 相关系数 冬季 320 84 2.5 228.4 387.1 0.82 春季 900 212 8.6 266.6 400.0 0.82 夏季 12000 2579 1.0 257.4 352.6 0.79 秋季 6200 1557 -0.2 251.6 362.5 0.80 
下载: 导出CSV
-
[1] Stull R B. 边界层气象学导论. 杨长新, 译. 北京: 气象出版社, 1991. [2] Liu S, Liang X Z.Observed diurnal cycle climatology of planetary boundary layer height.J Climate, 2010, 22:5790-5809. https://www.researchgate.net/publication/258402747_Observed_Diurnal_Cycle_Climatology_of_Planetary_Boundary_Layer_Height [3] Scarino A J, Obland M D, Fast J D, et al.Comparison of mixed layer heights from airborne high spectral resolution lidar, ground-based measurements, and the WRF-Chem model during CalNex and CARES.Atmos Chem Phys, 2014, 14:5547-5560. doi: 10.5194/acp-14-5547-2014 [4] Seibert P, Beyrich F, Gryning S E, et al.Review and intercomparison of operational methods for the determination of the mixing height.Atmos Environ, 2000, 34:1001-1027. doi: 10.1016/S1352-2310(99)00349-0 [5] 王珍珠, 李炬, 钟志庆, 等.激光雷达探测北京城区夏季大气边界层.应用光学, 2008, 29(1):96-100. http://www.cnki.com.cn/Article/CJFDTOTAL-YYGX200801021.htm [6] 贺千山, 毛节泰.微脉冲激光雷达及其应用研究进展.气象科技, 2004, 32(4):219-224. http://www.cnki.com.cn/Article/CJFDTOTAL-QXKJ200404004.htm [7] 贺千山, 毛节泰.北京城市大气混合层与气溶胶垂直分布观测研究.气象学报, 2005, 63(3):374-384. doi: 10.11898/1001-7313.20050312 [8] Fabio D F, Giovanni S.A combined natural orthogonal functions Ⅰ neural network technique for the radiometric estimation of atmospheric profiles.Radio Science, 1998(2):405-410. https://www.researchgate.net/publication/224266201_A_neural_network_algorithm_for_the_retrieval_of_atmospheric_profiles_from_radiometric_data [9] 雷恒池, 魏重, 沈志来.微波辐射计探测降雨前水汽和云液水.应用气象学报, 2001, 12(增刊Ⅰ):73-79. http://www.cnki.com.cn/Article/CJFDTOTAL-YYQX2001S1009.htm [10] 姚展予, 王广河, 游来光, 等.寿县地区云中液态水含量的微波遥感.应用气象学报, 2001, 12(增刊Ⅰ):88-95. http://www.cnki.com.cn/Article/CJFDTOTAL-YYQX2001S1011.htm [11] 魏重, 雷恒池, 沈志来.地基微波辐射计的雨天探测.应用气象学报, 2001, 12(增刊Ⅰ):65-72. http://www.cnki.com.cn/Article/CJFDTOTAL-YYQX2001S1008.htm [12] 陈洪滨, 林龙福.从118.75 GHz附近六通道亮温反演大气温度廓线的数值模拟研究.大气科学, 2003, 27(5):894-899. http://www.cnki.com.cn/Article/CJFDTOTAL-DQXK200305009.htm [13] 姚志刚, 陈洪滨.利用神经网络从118.75 GHz附近通道亮温反演大气温度.气象科学, 2006, 26(3):252-259. http://www.cnki.com.cn/Article/CJFDTOTAL-QXKX200603002.htm [14] Cimini D, De Angelis J C, Dupont S P, et al.Mixing layer height retrievals by multichannel microwave radiometer observations.Atmos Meas Tech Discuss, 2013, 6:4971-4998. doi: 10.5194/amtd-6-4971-2013 [15] 黄治勇, 徐桂荣, 王晓芳, 等.地基微波辐射资料在短时暴雨潜势预报中的应用.应用气象学报, 2013, 24(5):576-584. doi: 10.11898/1001-7313.20130507 [16] 王振会, 李青, 楚艳丽, 等. 地基微波辐射计工作环境对k波段亮温观测影响. 应用气象学报, 2014, 25(6): 711-721. http://qikan.camscma.cn/jams/ch/reader/view_abstract.aspx?file_no=20140607&flag=1 [17] 车云飞, 马舒庆, 杨玲, 等.云对地基微波辐射计反演温度廓线的影响.应用气象学报, 2015, 26(2):193-202. doi: 10.11898/1001-7313.20150207 [18] 夏俊荣.华北地区大气气溶胶垂直分布特性的观测与分析.北京:中国科学院大气物理研究所, 2010. [19] 蒋维楣, 徐玉貌, 于洪彬.边界层气象学基础.南京:南京大学出版社, 1994. [20] Simon Hayki. 神经网络原理. 叶世伟, 史忠植, 译. 北京: 机械工业出版社, 2004: 33-175. [21] Hagan M T, Demuth H B, Beale M H.神经网络设计.北京:机械工业出版社, 2002. [22] 刘旸, 官莉.人工神经网络法反演晴空大气湿度廓线的研究.气象, 2011, 37(3):318-324. doi: 10.7519/j.issn.1000-0526.2011.03.009 [23] 谷良雷, 胡泽勇, 吕世华, 等.敦煌和酒泉夏季晴天和阴天边界层气象要素特征分析.干旱区地理, 2007, 30(6):871-878. http://www.cnki.com.cn/Article/CJFDTOTAL-GHDL200706016.htm -
计量
- 摘要浏览量: 4791
- HTML全文浏览量: 1436
- PDF下载量: 1437
- 被引次数: 0
