设为首页
加入收藏
Experiments on Improving Temperature and Humidity Profile Retrieval for Ground-based Microwave Radiometer
-
摘要: 为提升地基微波辐射计在不同天气条件下, 特别是云天条件下温湿廓线的反演精度, 利用2011年1月—2016年12月中国气象局北京国家综合气象观测试验基地探空数据, 在微波辐射计反演温湿度廓线的过程中通过区分晴天和云天条件并引入全固态Ka波段测云仪云高及云厚信息, 对反演输入亮温进行质量控制和偏差订正, 建立BP神经网络模型, 采用2017年1月—2018年3月微波辐射计探测数据评估检验, 结果表明：在亮温订正前提下, 晴天温度模型、云天温度模型、晴天相对湿度模型和云天相对湿度模型反演结果与探空的相关系数分别为0.99, 0.99, 0.80和0.78, 均方根误差为2.3℃, 2.3℃, 9%和16%, 较微波辐射计自带产品（LV2产品）减小约0.4℃, 0.3℃, 11%和9%, 准确性提升约30%, 28%, 64%和45%；温度模型偏差在±2℃以内、湿度模型偏差在±20%以内的占比分别为68%, 70%和95%, 78%, 较LV2产品分别提高了7%, 5%和27%, 23%, 其中相对湿度改善明显。可见亮温订正、区分天气类型训练反演模型有利于改善地基微波辐射温湿廓线反演精度。Abstract: Ground-based microwave radiometer (MWR) has a crucial role in scientific researches, weather modification service and climate change studies. MWR adopts passive remote sensing technology which has smaller volume, lower power consumption. Ground-based microwave radiometer detects atmospheric temperature and humidity by receiving atmospheric microwave radiation, which can conduct 24-hour unattended, high-resolution observation. It can detect short-time variation of atmospheric elements. Many studies show that different seasons, different weather conditions, quality control algorithms, and changes in environments have certain effects on retrieval results of MWR. In the case of cloudy condition, the uncertainty of cloud absorption coefficient leads to the increase of retrieval error and incorrect data of MWR especially. In order to improve temperature and relative humidity detection capabilities of MWR, the experiment builds BP neural network algorithm with 6 years(from 2011 to 2016) sounding data. The experiment builds two types of retrieval methods because there are some differences of microwave radiation transfer between clear and cloudy samples. The test uses measured brightness temperature data (requiring correction) and cloud data of millimeter-wavelength cloud radar as model inputs and then uses sounding data to evaluate model outputs (temperature and relative humidity profiles) from 2017 to 2018.Results show that correlation coefficients between outputs of 4 models (clear sky sample temperature model, cloudy sample temperature model, clear sky sample relative humidity model and the cloudy sample relative humidity model) and sounding data are 0.99, 0.99, 0.80 and 0.78. Taking sounding profiles as reference, root mean square errors (RMSE) of retrieval results of 4 models are 2.3℃, 2.3℃, 9%, 16%. Comparing with the MWR original profiles, RMSEs of 4 models are reduced by 0.4℃, 0.3℃, 11% and 9%, accuracies are improved by about 30%, 28%, 64% and 45%. In particular, the deviation of temperature model and humidity model within ±2℃ and ±20% account for 68%, 70% and 95%, 78%, which are 7%, 5% and 27%, 23% higher than MWR original profiles. The bias correction of brightness temperature and the training retrieval model of distinguishing weather samples are helpful to improve the retrieval accuracy of MWR temperature and humidity profiles. The network model combined with cloud radar information has obviously better effects on the retrieval result under cloudy samples.Through these experiments, the quality control of brightness temperature, combination of active and passive retrieval algorithms are well improved. The combination of active and passive retrieval effectively improves the performance of MWR, which will lay a foundation for the development of the comprehensive observation system of atmospheric profiles.
-
图 1 实测与模拟亮温拟合
(a)晴天条件, 22.24 GHz通道, (b)晴天条件, 58.00 GHz通道, (c)云天条件, 22.24 GHz通道, (d)云天条件, 58.00 GHz通道
Fig. 1 Linear fitting of measured and simulated brightness temperatures
(a)channel of 22.24 GHz in clear sky samples, (b)channel of 58.00 GHz in clear sky samples, (c)channel of 22.24 GHz in cloudy samples, (d)channel of 58.00 GHz in cloudy samples
图 2 晴天温度模型反演结果及LV2产品与探空对比
(a)2018年2月23日19:15温度廓线, (b)2017年3月27日07:15温度廓线, (c)2017年1月—2018年3月平均偏差廓线, (d)2017年1月—2018年3月均方根误差廓线
Fig. 2 Comparison of temperature profiles of retrieval model, LV2 and sounding in clear sky samples
(a)case at 1915 BT 23 Feb 2018, (b)case at 0715 BT 27 Mar 2017, (c)mean bias from Jan 2017 to Mar 2018, (d)root mean square error from Jan 2017 to Mar 2018
图 3 云天温度模型反演结果及LV2产品与探空对比
(a)2017年1月17日07:15温度廓线, (b)2017年10月31日07:15温度廓线, (c)2017年1月—2018年3月平均偏差廓线, (d)2017年1月—2018年3月均方根误差廓线
Fig. 3 Comparison of temperature profiles of retrieval model, LV2 and sounding in cloudy samples
(a)case at 0715 BT 17 Jan 2017, (b)case at 0715 BT 31 Oct 2017, (c)mean bias from Jan 2017 to Mar 2018, (d)root mean square error from Jan 2017 to Mar 2018
图 4 晴天相对湿度模型反演结果、LV2产品廓线与探空对比
(a)2017年3月27日07:15相对湿度廓线, (b)2017年4月26日07:15相对湿度廓线, (c)2017年1月—2018年3月平均偏差廓线, (d)2017年1月—2018年3月均方根误差廓线
Fig. 4 Comparison of relative humidity profiles of retrieval model, LV2 and sounding in clear sky samples
(a)case at 0715 BT 27 Mar 2017, (b)case at 0715 BT 26 Apr 2017, (c)mean bias from Jan 2017 to Mar 2018, (d)root mean square error from Jan 2017 to Mar 2018
图 5 云天相对湿度模型反演结果、LV2产品与探空对比
(a)2018年2月1日19:15低云, (b)2017年8月1日13:15低云, (c)2017年8月8日07:15中云, (d)2018年1月17日07:15中云, (e)2017年12月14日07:15高云, (f)2017年1月24日19:15高云
Fig. 5 Comparison of relative humidity profiles of retrieval model, LV2 and sounding in cloudy samples
(a)low cloud samples at 1915 BT 1 Feb 2018, (b)low cloud samples at 1315 BT 1 Aug 2017, (c)medium cloud samples at 0715 BT 8 Aug 2017, (d)medium cloud samples at 0715 BT 17 Jan 2018, (e)high cloud samples at 0715 BT 14 Dec 2017, (f)high cloud samples at 1915 BT 24 Jan 2017
图 6 云天相对湿度模型反演结果及LV2产品偏差廓线
(a)平均偏差廓线, (b)均方根误差廓线
Fig. 6 Relative humidity profiles of retrieval model and LV2 in cloudy samples
(a)mean bias, (b)root mean square error
表 1 亮温订正关系式
Table 1 Relation between the simulated and the measured brightness temperatures
通道频点/GHz 订正关系式(晴天) 订正关系式(云天) 22.24 Y=1.0897X+0.1843 Y=1.0081X+1.8127 23.04 Y=1.0899X+0.3553 Y=1.0068X+1.8865 23.84 Y=1.0942X+0.0848 Y=1.0062X+1.5239 25.44 Y=1.1114X-0.4975 Y=1.0001X+1.0236 26.24 Y=1.1227X-0.4410 Y=0.9944X+1.2145 27.84 Y=1.1281X-0.5029 Y=0.9734X+1.3946 31.40 Y=1.1299X-0.5855 Y=0.9250X+2.1566 51.26 Y=1.2011X-16.9122 Y=0.8558X+18.4323 52.28 Y=1.2191X-28.6074 Y=0.9048X+16.7461 53.86 Y=1.0996X-21.2750 Y=1.0418X-7.2022 54.94 Y=1.0239X-6.5648 Y=1.0161X-4.3417 56.66 Y=0.9989X-0.3136 Y=0.9971X+0.2349 57.30 Y=0.9970X+0.2410 Y=0.9945X+0.9747 58.00 Y=0.9969X+0.3437 Y=0.9939X+1.2098 
下载: 导出CSV
表 2 反演模型及对应试验方案
Table 2 Retrieval models and test schemes
方案 模型 是否进行亮温订正 是否考虑云的影响 反演结果 S1 晴天温度 否 否 SP1 S2 晴天温度 是 否 SP2 S3 云天温度 否 是 SP3 S4 云天温度 是 是 SP4 S5 晴天相对湿度 否 否 SP5 S6 晴天相对湿度 是 否 SP6 S7 云天相对湿度 否 是 SP7 S8 云天相对湿度 是 是 SP8
下载: 导出CSV
表 3 LV2产品、SP1和SP2与探空的相关性及偏差统计
Table 3 Correlation and bias of LV2, SP1, SP2 to sounding in clear sky samples
高度/km 对比项 相关系数 均方根误差 不同偏差范围占比/% 准确性提升率/% A B C LV2产品 0.9956 1.5 56 30 14 [0, 2.5] SP1 0.9903 2.0 47 28 25 SP2 0.9937 1.5 56 28 16 6.78 LV2产品 0.9852 3.2 17 20 63 (2.5, 4.5] SP1 0.9839 4.3 8 13 79 SP2 0.9879 2.2 34 30 36 42.08 LV2产品 0.9821 3.1 26 25 49 (4.5, 10] SP1 0.9810 4.3 14 15 71 SP2 0.9837 2.8 32 25 43 19.17 LV2产品 0.9954 2.7 35 26 39 [0, 10] SP1 0.9913 3.6 25 19 56 SP2 0.9946 2.3 41 27 32 29.98 注：误差占比范围A: [-1, 1], B:[-2, -1)∪(1, 2], C: (-∞, -2)∪(2, +∞), 单位:℃。表中仅统计SP2相对于LV2产品的提升率。
下载: 导出CSV
表 4 LV2产品、SP3和SP4与探空的相关性及偏差统计
Table 4 Correlation and bias of LV2, SP3, SP4 to sounding in cloudy samples
高度/km 对比项 相关系数 均方根误差 不同偏差范围占比/% 准确性提升率/% A B C LV2产品 0.9953 1.4 61 27 12 [0, 2.5] SP3 0.9812 2.4 37 36 27 SP4 0.9944 1.3 67 22 11 30.93 LV2产品 0.9873 2.5 31 25 44 (2.5, 4.5] SP3 0.9795 9.0 0 0 100 SP4 0.9829 2.0 40 31 29 35.29 LV2产品 0.9838 3.2 38 31 31 (4.5, 10] SP3 0.9807 11.6 0 1 35 SP4 0.9822 2.9 42 35 23 23.15 LV2产品 0.9953 2.6 41 24 35 [0, 10] SP3 0.9675 8.9 14 14 72 SP4 0.9946 2.3 45 25 30 27.76 注：误差占比范围A: [-1, 1], B:[-2, -1)∪(1, 2], C: (-∞, -2)∪(2, +∞), 单位:℃。表中仅统计SP4相对于LV2产品的提升率。
下载: 导出CSV
表 5 LV2产品、SP5和SP6与探空的相关性及偏差统计
Table 5 Correlation and bias of LV2, SP5, SP6 to sounding in clear sky samples
高度/km 对比项 相关系数 均方根误差 不同偏差范围占比/% 准确性提升率/% A B C LV2产品 0.8246 12 60 30 10 [0, 2.5] SP5 0.8349 11 71 21 8 SP6 0.8529 10 77 17 6 27.13 LV2产品 0.6880 17 42 35 23 (2.5, 4.5] SP5 0.6575 13 68 19 13 SP6 0.7059 10 73 19 8 50.42 LV2产品 0.3073 25 19 29 52 (4.5, 10] SP5 0.2426 9 87 9 4 SP6 0.3576 8 88 9 3 74.02 LV2产品 0.5800 20 38 30 32 [0, 10] SP5 0.7606 10 78 15 7 SP6 0.8006 9 81 14 5 64.28 注：误差占比范围A: [-10, 10], B:[-20, -10)∪(10, 20], C: (-∞, -20)∪(20, +∞), 单位:%。表中仅统计SP6相对于LV2产品的提升率。
下载: 导出CSV
表 6 LV2产品、SP7和SP8与探空的相关性及偏差统计
Table 6 Correlation of LV2, SP7, SP8 to sounding in cloudy samples with the bias
高度/km 对比项 相关系数 均方根误差 不同偏差范围占比/% 准确性提升率/% A B C LV2产品 0.7599 19 37 32 31 [0, 2.5] SP7 0.8143 15 59 23 18 SP8 0.8439 14 63 23 14 35.50 LV2产品 0.6077 26 26 29 45 (2.5, 4.5] SP7 0.7417 28 33 20 47 SP8 0.7770 18 46 29 25 46.27 LV2产品 0.3145 28 20 23 57 (4.5, 10] SP7 0.6302 24 39 25 36 SP8 0.7477 18 45 29 26 48.91 LV2产品 0.5478 25 23 27 45 [0, 10] SP7 0.6764 22 45 24 31 SP8 0.7855 16 51 27 22 44.95 注：误差占比范围A: [-10, 10], B:[-20, -10)∪(10, 20], C: (-∞, -20)∪(20, +∞), 单位：%。表中仅统计SP8相对于LV2产品的提升率。
下载: 导出CSV
-
[1] 孙学金.大气探测学.北京:气象出版社, 2009. [2] 周秀骥.中尺度气象学研究与中国气象科学研究院.应用气象学报, 2006, 17(6):665-671. http://qikan.camscma.cn/jamsweb/article/id/200606115 [3] 陈明轩, 俞小鼎, 谭晓光, 等.对流天气临近预报技术的发展与研究进展.应用气象学报, 2004, 15(6):754-766. http://qikan.camscma.cn/jamsweb/article/id/20040693 [4] 郭丽君, 郭学良.利用地基多通道微波辐射计遥感反演华北持续性大雾天气温、湿度廓线的检验研究.气象学报, 2015, 73(2):368-381. http://d.old.wanfangdata.com.cn/Periodical/qxxb201502012 [5] Raju C S, Renju R, Antony T, et al.Microwave radiometric observation of a waterspout over coastal Arabian Sea.IEEE Geoscience & Remote Sensing Letters, 2013, 10(5):1075-1079. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=343856de9be826fc5f33977ee84a84f9 [6] Ware R, Carpenter R, Güldner J, et al.A multichannel radiometric profiler of temperature, humidity, and cloud liquid.Radio Science, 2016, 38(4):44-1-44-13. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=91eb3b202eb0ec2c092c73ca8f032e81 [7] 鲍艳松, 钱程, 闵锦忠, 等.利用地基微波辐射计资料反演0~10 km大气温湿廓线试验研究.热带气象学报, 2016, 32(2):163-171. http://d.old.wanfangdata.com.cn/Periodical/rdqxxb201602003 [8] 刘亚亚, 毛节泰, 刘钧, 等.地基微波辐射计遥感大气廓线的BP神经网络反演方法研究.高原气象, 2010, 29(6):1514-1523. http://d.old.wanfangdata.com.cn/Conference/7570338 [9] 王云, 王振会, 李青, 等.基于一维变分算法的地基微波辐射计遥感大气温湿廓线研究.气象学报, 2014, 72(3):570-582. http://d.old.wanfangdata.com.cn/Periodical/qxxb201403011 [10] 刘思波, 何文英, 刘红燕, 等.地基微波辐射计探测大气边界层高度方法.应用气象学报, 2015, 26(5):626-635. doi: 10.11898/1001-7313.20150512 [11] 段英, 吴志会.利用地基遥感方法监测大气中汽态、液态水含量分布特征的分析.应用气象学报, 1999, 10(1):34-40. http://qikan.camscma.cn/jamsweb/article/id/19990133 [12] 郝巨飞, 袁雷武, 李芷霞, 等.激光雷达和微波辐射计对邢台市一次沙尘天气的探测分析.高原气象, 2018, 37(4):1110-1119. http://d.old.wanfangdata.com.cn/Periodical/gyqx201804023 [13] 刘建忠.不同时次地基微波辐射计反演产品评估.气象科技, 2012, 40(3):332-339. http://d.old.wanfangdata.com.cn/Periodical/qxkj201203002 [14] 刘红燕.三年地基微波辐射计探测温度廓线的精度分析.气象学报, 2011, 69(4):719-728. [15] 侯叶叶.地基微波辐射计反演水汽密度廓线精度分析.气象科技, 2016, 44(5):702-709. http://d.old.wanfangdata.com.cn/Periodical/qxkj201605002 [16] 张文刚, 徐桂荣, 颜国跑, 等.微波辐射计与探空仪测值对比分析.气象科技, 2014, 42(5):737-741. http://d.old.wanfangdata.com.cn/Periodical/qxkj201405002 [17] 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://d.old.wanfangdata.com.cn/Periodical/yyqxxb2001z1009 [18] Liljegren J C, Clothiaux E E, Mace G G, et al.A new retrieval for cloud liquid water path using a ground-based microwave radiometer and measurements of cloud temperature.J Geophys Res Atmos, 2001, 106(D13):14485-14500. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=191a8b41836bfa2fdf7f1fe28ca194b0 [19] Frate F D, Schiavon G.A combined natural orthogonal functions/neural network technique for the radiometric estimation of atmospheric profiles.Radio Science, 1998, 33(2):405-410. [20] 车云飞, 马舒庆, 杨玲, 等.云对地基微波辐射计反演湿度廓线的影响.应用气象学报, 2015, 26(2):193-202. doi: 10.11898/1001-7313.20150207 [21] 茆佳佳, 张雪芬, 王志诚, 等.多型号地基微波辐射计亮温准确性比对.应用气象学报, 2018, 29(6):86-98. doi: 10.11898/1001-7313.20180608 [22] 王志诚, 张雪芬, 茆佳佳, 等.不同天气条件下地基微波辐射计探测性能比对.应用气象学报, 2018, 29(3):282-295. doi: 10.11898/1001-7313.20180303 [23] 唐英杰, 马舒庆, 杨玲, 等.云底高度的地基毫米波云雷达观测及其对比.应用气象学报, 2015, 26(6):680-687. doi: 10.11898/1001-7313.20150604 [24] 袁野, 朱士超, 李爱华.黄山雨滴下落过程滴谱变化特征.应用气象学报, 2016, 27(6):734-740. doi: 10.11898/1001-7313.20160610 [25] 陶法, 胡树贞, 张雪芬.地基可见光/红外全天空成像仪数据融合.气象, 2018, 44(4):52-59. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=qx201804005 [26] 周毓荃, 欧建军.利用探空数据分析云垂直结构的方法及其应用研究.气象, 2010, 36(11):50-58. http://d.old.wanfangdata.com.cn/Periodical/qx201011008 [27] Clough S A, Shephard M W, Mlawer E J.Atmospheric radiative transfer modeling:A summary of the AER codes.Journal of Quantitative Spectroscopy & Radiative Transfer, 2005, 91:233-244. doi: 10.1016-j.jqsrt.2004.05.058/ [28] Pierdicca N, Pulvirenti L, Marzano F S.A model to predict cloud density from midlatitude atmospheric soundings for microwave radiative transfer applications.Radio Science, 2016, 41(6):1-12. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=cb8a7699ce88e212ef2279d8279fe9e4 [29] 顾成明, 王云峰, 张晓辉, 等.云参数对微波亮温模拟计算的影响试验.应用气象学报, 2016, 27(3):380-384. doi: 10.11898/1001-7313.20160313 [30] 傅新姝, 谈建国.地基微波辐射计探测资料质量控制方法.应用气象学报, 2017, 28(2):209-217. doi: 10.11898/1001-7313.20170208 [31] 气象观测专用技术装备测试方法地面气象观测设备(试行).北京: 中国气象局, 2015. -
计量
- 摘要浏览量: 3006
- HTML全文浏览量: 1669
- PDF下载量: 216
- 被引次数: 0
