Advances in Meteorology

Volume 2016 (2016), Article ID 5375918, 13 pages

http://dx.doi.org/10.1155/2016/5375918

## Statistical Analysis of Relationship between Daytime Lidar-Derived Planetary Boundary Layer Height and Relevant Atmospheric Variables in the Semiarid Region in Northwest China

^{1}Key Laboratory of Arid Climatic Changing and Reducing Disaster of Gansu Province, College of Atmospheric Sciences, Lanzhou University, Lanzhou, Gansu 730000, China^{2}The Central Meteorological Observatory of Lanzhou, Lanzhou 730020, China

Received 9 November 2015; Revised 19 January 2016; Accepted 18 April 2016

Academic Editor: Jose D. Fuentes

Copyright © 2016 Ruijun Dang et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

#### Abstract

Accurate identification of key parameters for data assimilation is important in simulating the planetary boundary layer height (PBLH) and structure evolution in numerical weather prediction models. In this study, surface observational data and lidar-derived PBLH on 42 cloudless days from June 2007 to May 2008 are used to quantify the statistical relationships between surface parameters and the PBLH at a semiarid climate observational site in Northwest China. The results indicate that surface upward long wave radiation, surface temperature, and surface sensible heat fluxes show strong correlations with the PBLH with correlation coefficients at a range of 0.63–0.72. But these parameters show varying correlation response time to the different stages of PBL development. Furthermore, the air temperature shows the highest correlation with the PBLH near the surface and the correlation decreases with increasing height.

#### 1. Introduction

The atmospheric boundary layer, also known as the planetary boundary layer (PBL), is the turbulent layer near the Earth’s surface. It is directly affected by the underlying surface conditions and intimately associated with human activities [1]. The transfers of momentum, heat, and moisture between the surface and atmosphere are mainly based on turbulence. As the atmosphere is always in turbulent status in the layer, the PBL is crucial to surface-atmosphere exchanges of substances and energy. PBLH is of major relevance in boundary layer research as a key parameter characterizing the structure of the boundary layer [2, 3]. Observations of the PBLH are significant for theory and applications. Because it is closely related to turbulence, the PBLH is not observed by standard measurements. It is currently determined mainly from indirect measurements. For a convective boundary layer at noon, the PBLH is more or less identical with the mixed layer height. Due to the vertical turbulent mixing, wind velocity and potential temperature are well mixed within the layer. In most cases, wind and potential temperature are usually constants in the mixed layer. However, at the top of mixed layer, there is a sharp increase in wind speed and potential temperature caused by the abrupt decrease in turbulence intensity [4]. Therefore, the characteristics of wind speed and potential temperature can be used to calculate the PBLH when atmosphere is neutral or unstable. In addition, the PBL is moist relative to the upper free atmosphere, and a strong gradient in relative humidity exists at the top of PBL, which can also be utilized to determine the daytime PBLH [5]. At night when atmosphere is in a stable condition, inversion lid always exists at the top of boundary layer, and the nocturnal PBLH is usually represented by the thickness of surface temperature inversion layer. Above all, the PBLH can be determined from different instruments-derived profiles of thermodynamic variables like temperature, humidity, and horizontal wind speed. The difficulty in directly observing the thermodynamic structures of the atmosphere makes ground-based remote sensing technique an attractive choice. For instance, lidar provides vertical profiles of backscatter from aerosol particles with high temporal and spatial resolutions in the atmosphere. The aerosol concentration within the PBL is much higher than that in the free atmosphere. Therefore, a significant difference in aerosol concentration exists between the top of the PBL and the free atmosphere, which is reflected as a sudden attenuation of the lidar echo signals. On the basis of this characteristic of aerosols in the PBL, aerosol particles can be used as tracers to determine the PBLH. However, in the presence of optically thick clouds, the resulting PBLH using lidar data is unrealistic because of the high signal gradient generated by the clouds [6, 7]. Therefore, lidar data in clear sky conditions are chosen to calculate PBLH in this paper.

As the backscatter signal generally decreases most rapidly at the top of the boundary layer, the gradient of the aerosol concentration obtained from the lidar data can be utilized to retrieve PBLH. Many methods have been used to calculate the PBLH from lidar backscatter, including the gradient method [8, 9], the wavelet transform method [10–12], the standard deviation method [13], and the curve fitting method [14, 15]. Each method has its advantages and limitations. The gradient method is simple and easy to use; however, it is sensitive to local minima in the profile either atmosphere or noise induced nearly always occurring in a turbulent PBL [16]. The standard deviation method is not suitable for the situation of weak inversion layer [17]. Although the curve fitting method is relatively computationally expensive, it is barely affected by the local structure of the signal and generally generates the stable PBLH [14]. Therefore, the curve fitting method is used to retrieve daytime PBLH in the paper.

The atmospheric boundary layer is largely governed by land surface processes, including the absorption of solar radiation by the land surface, transmission of heat energy between the atmosphere and soil, and mechanical processes. The surface temperature is an important external forcing factor to the thermal convection. The variation in surface temperature reflects the heating result of net radiation on the surface [18]. For net radiation, the contribution of the long and short wave components varies with atmospheric conditions. On sunny days, the upward long wave contributes most to the net radiation, and the contribution of upward short wave is minimum [19]. Besides, the development and maintenance of the thermal boundary layer mainly rely on the heat transmission through the sensible heat flux [20]. Therefore, the radiation variables, surface temperature, and sensible heat flux make major contributions to the formation and development of the PBL [21–24]. The assimilation of PBLH may be implemented by updating the first guess field of a numerical model with these variables.

For PBLH assimilation in the numerical model with Ensemble Kalman Filter (EnKF), it needs to confirm which variables are well correlated with PBLH. In addition, the influence radius for spatial and temporal domain should also be set. So the purpose of this study is to determine the statistical correlations between PBLH and conventional atmospheric variables, as well as influence radius of variables using the routine observations at the Semi-Arid Climate and Environment Observatory of Lanzhou University (SACOL) and to provide basis and support for PBLH assimilation. Due to the limitation of a single observational point, the radius of influence in horizontal direction cannot be found out. In the vertical direction, the vertical air temperature profiles are provided by a Radiometrics Profiling Radiometer (TP/WVP-3000). The observations of variables and PBLH in the following hours are used to analyze the temporal influence radius.

In this study, 42 cloudless sunny days (nonprecipitation, being without thunderstorm, no cloud or total-cloud covers less than 20 percent all day, and being with a clear structure of backscatter signals of lidar) are selected from June 2007 to May 2008, and the PBLH is calculated by retrieving lidar data using the curve fitting method over the Lanzhou suburb in the Yuzhong area at SACOL. The correlations between related variables and PBLH as well as lagged correlations between them are calculated to determine the major variables which affect the formation and development of boundary layer. The correlation coefficients between PBLH and air temperature at different heights are also calculated. Finally, through temporal variations of PBLH and atmospheric variables on the four typical examples 15 July 2007, 20 November 2007, 5 January 2008, and 9 April 2008, the lagged correlations between different variables and PBLH and the physical mechanisms behind the statistical correlations are specifically discussed.

#### 2. Data and Methods

The PBLH and statistical correlations in this paper are calculated with data collected at SACOL (35.57°N, 104.08°E; 1965.8 m above sea level), which is the suburb of Lanzhou on the southern bank of the Yellow River, a typical semiarid region. The instruments include air temperature and relative humidity sensors (HMP45C-L, Vaisala), a Precision Infrared Thermocouple Sensor (IRTS-P, Apogee), upward and downward pyranometers (CM21, Kipp & Zonen), upward and downward pyrgeometers (CG4, Kipp & Zonen), an atmospheric pressure sensor (RPT410F-3143, Druck), a Radiometrics Profiling Radiometer (TP/WVP-3000, Radiometrics), and a Micro-Pulse Lidar System (MPL-4, Sigma Space). The vertical resolutions of temperature profiles measured by the radiometer for the layers 1 km below and above are 100 m and 250 m, respectively. MPL-4 has one measurement channel at 527 nm, which records backscatter signals up to a height of 20+ km with a vertical resolution of 75 m. All the conventional atmospheric observations are subjected to basic quality control (QC). Only observations with a relatively high accuracy are selected. The SACOL MPL-4 is part of the MPLNET (Micro-Pulse Lidar Network) [25], and the observation follows the relevant uniform rules. Meanwhile, a series of corrections such as background correction, overlap correction, and range correction have been done for lidar data [26].

The curve fitting method first proposed by Steyn et al. [14] is used to retrieve PBLH from the lidar data. The technique uses the gradient of the lidar backscatter signal and fits an idealized backscatter profile to the observed backscatter profile by minimizing the measure of agreement between the two profiles. The form of the idealized backscatter profile iswhere the error function ( is defined as where and are the mean backscatters in the mixed layer and in air immediately above the mixed layer, respectively; is the depth of the mixed layer; is related to the thickness of the entrainment layer. The four parameters are determined by minimizing the root-mean-square deviation between and. When the root-mean-square deviation gets the minimum, represents the PBLH.

#### 3. Statistical Correlations between PBLH and Variables

##### 3.1. Statistical Correlations between Averages

The dates chosen for PBLH retrieval and correlation analysis are listed in Table 1. On these 42 cloudless sunny days, conventional observations are complete. The structure of the lidar backscatter signals is also very clear. To ensure representativeness, the selected days are from all four seasons. Because some data are unavailable for 8–30 September 2007, the cases in autumn are relatively less. But the representativeness of the statistical correlations is not affected.