大气科学  2018, Vol. 42 Issue (2): 292-300 PDF

1 中国科学院大气物理研究所中层大气和全球环境探测重点实验室, 北京 100029
2 中国科学院大学, 北京 100049
3 南京信息工程大学气象灾害预报预警与评估协同创新中心, 南京 210044
4 中国电力科学研究院新能源与储能运行控制国家重点实验室, 北京 100192

Estimation of Radiative Forcing and Chore Length of Shallow Convective Clouds (SCC) Based on Broadband Pyranometer Measurements
SHI Hongrong1,2, CHEN Hongbin1,2,3, XIA Xiang'ao1,2,3, FAN Xuehua1, ZHANG Jinqiang1,2,3, CAO Yang1,2, ZHU Yanliang1, WANG Shu4
1 Key Laboratory of Middle Atmosphere and Global Environment Observation(LAGEO), Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing 100029
2 University of Chinese Academy of Sciences, Beijing 100049
3 Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters, Nanjing University of Information Science & Technology, Nanjing, 210044
4 State Key Laboratory of Operation and Control of Renewable Energy & Storage Systems, China Electric Power Research Institute, Beijing 100192
Abstract: We presented a method to identify and calculate cloud radiative forcing (CRF) and horizontal chore length (L) of shallow convective clouds (SCC) using broadband pyranometer measurements. The data analyzed were collected from the SCC campaign during two summers (2015-2016) at Baiqi site over the Inner Mongolia grassland. The SCC detection method was verified against observations from reports and cameras records. It was found that the results of the detection method and human observations were in agreement by about 75%. The deviation of L can be attributed to following factors:1) the accuracy of wind speed at the height of SCC and the ratio of horizontal to vertical length play a key role in determining the value of L; 2) the effect of variance in solar zenith angle can be negligible. The downwelling shortwave CRF of SCC was −134.1 W m−2. The average value of L of SCC was 1129 m. Besides, the distribution of normalized cloud chore length agreed well with a power-law distribution.
Key words: Shallow convective clouds      Horizontal chore length      Radiative forcing      Pyranometer
1 引言

2 数据及算法介绍

2.1 算法介绍

 图 1 不同类型（a）晴空、（b）阴天、（c）积雨云及（d）淡积云天空状况下地面向下短波辐射通量时间序列 Figure 1 Time series of downwelling shortwave irradiance at the surface under different conditions: (a) Clear sky; (b) overcast; (c) cumulonimbus; (d) shallow convective clouds (SCC)
 ${F_{{\rm{cloud}}}} = {F_{{\rm{obs}}}} - {F_{{\rm{clear}}}}$ (1)
 $\tau = {F_{{\rm{obs}}}}/{F_{{\rm{clear}}}}$ (2)
 $D = |\Delta {F_{{\rm{obs}}}}|/\Delta t$ (3)

 图 2 淡积云天（2015年8月6、8及9日）辐射强迫、透过率以及辐射通量变化率统计值。黑色实线为淡积云特征阈值（Fcloud＜−45，τ＜0.9，D＞150） Figure 2 Scatter plots of cloud radiative forcing (Fcloud), transmittance (τ) and gradient of irradiance (D) of SCC days (6, 8, and 9 August 2015). Black solid lines show characteristic thresholds of SCC (Fcloud < −45, τ < 0.9, and D > 150)

 图 3 淡积云检测识别算法流程图。（a）、（b）、（c）分别是阴天、积雨云和淡积云天示例 Figure 3 The flow chart of algorithm of shallow convective clouds (SCC) detection. (a), (b), (c) are the examples of overcast, cumulonimbus, and SCC days, respectively

 图 4 计算淡积云水平尺度方法示意图 Figure 4 Schematic diagram for calculating horizontal length of SCC
 $L = U \cdot \Delta t/(1 + \tan (\theta)/{A_r}),$ (4)

2.2 误差分析

3 淡积云辐射强迫及水平尺度分析

 图 5 2015年8月6日淡积云个例:（a）地表短波向下辐射通量观测值（obs）与晴空模拟值（fit），圆圈标记了淡积云遮挡辐射表的起始时间（黑色圆圈：开始；黑色叉：结束）；（b）淡积云水平尺度（蓝色）和淡积云云间距（黄色）；（c）云辐射强迫瞬时值（蓝色虚线）和小时平均值（黄色实线）；（d）云透过率瞬时值（蓝色虚线）和小时平均值（黄色实线） Figure 5 A sample of SCC day on 6 August 2015: (a) Observed and fitted values of downwelling shortwave irradiance (Black circle and cross are starting and ending time of clouds crossing the solar beam); (b) horizontal length (blue lines) and spacing (yellow lines) of SCC; (c) instantaneous (blue dashed lines) and 1 h average (yellow solid lines) values of Fcloud; (d) instantaneous (blue dash lines) and 1 h average (yellow solid lines) values of τ

 图 6 淡积云（a）水平尺度、（b）云间距、（c）云辐射强迫以及（d）透过率频率图 Figure 6 Histograms of probability density of (a) horizontal length, (b) cloud spacing length, (c) Fcloud, and (d) τ of SCCs.

 图 7 淡积云水平尺度概率密度分布，满足幂律分布（a=11.8，b=−0.91）。图b是图a的对数坐标形式，图中圈代表观测值，实线为拟合曲线，SSE为残差平方和，Rsquare为决定系数 Figure 7 Plot of normalized length density vs the length of SCC. Circles show observations of irradiance, lines are fitted values of irradiance. The fitted line agrees well with a power-law distribution (a=11.8; b=−0.91). Circles are observational values, solid line is the fitted line, SSE is sum of squares due to error, and Rsquare is coefficient of determination
4 总结