**1. Introduction**

Precipitation types can be mainly classified as stratiform precipitation (SP) and convective precipitation (CP) [1]. There are significant differences between SP and CP regarding the growth of precipitation particles by aggregation, riming, and coalescence, which is due to the different thermal dynamics and microphysics processes of the two types of precipitation [2,3]. The vertical structure of precipitation can reflect the characteristics of dynamics and microphysical of hydrometeor growth attenuation in the precipitation clouds. These microphysical and thermodynamic processes can affect the precipitation efficiency, and the intensity of surface precipitation, likewise, plays a role in determining precipitation type to some extent [4–7]. The topography has a very important influence on the vertical structure of precipitation and clouds [8–10], manifested in that topographical thermodynamic and

**Citation:** Shen, C.; Li, G.; Dong, Y. Vertical Structures Associated with Orographic Precipitation during Warm Season in the Sichuan Basin and Its Surrounding Areas at Different Altitudes from 8-Year GPM DPR Observations. *Remote Sens.* **2022**, *14*, 4222. https://doi.org/10.3390/ rs14174222

Academic Editors: Massimo Menenti, Yaoming Ma, Li Jia and Lei Zhong

Received: 26 July 2022 Accepted: 24 August 2022 Published: 27 August 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

dynamic processes affects atmospheric circulation, thereby the initiation and development of rainfall systems was significantly affected.

Sichuan province is located in the interior of southwest China, with the Qinling Mountains to the north, the Yun-Gui Plateau to the south, and the eastern edge of the Qinghai-Tibet Plateau. The terrain of Sichuan is high west and low east that the topography is diverse, and the weather processes are also complex and changeful. Heavy rainfall is one of the meteorological disasters with the highest frequency and the severest personal casualty and property loss in the Sichuan Basin and its surrounding areas. This region not only has high annual average precipitation, but is also prone to short-time heavy precipitation because of the complex terrain. In the past, research on mountain precipitation was difficult due to scarce surface observational data. At the same time, the detection of weather radar in mountainous areas is limited by terrain [11]. The Global Precipitation Measurement (GPM) mission was initiated by the National Aeronautics and Space Administration (NASA) and the Japan Aerospace and Exploration Agency (JAXA), which is the successor to Tropical Rainfall Measuring Mission (TRMM). The GPM observatory carries the first spaceborne dual-frequency phased array precipitation radar (DPR), developed by JAXA and National Institute of Communication Technology (NICT), and a conical-scanning multichannel microwave imager, developed and built by the Ball Aerospace and Technology Corporation under contract with NASA's Goddard Space Flight Center (GSFC). The GPM sensor has higher sensitivity and wider measurement range compared with TRMM instruments, and can provide more accurate precipitation microstructure information [12–14]. The detection of precipitation radar (PR) from satellite is not restricted to the geographical environment. Thus, it is feasible for PR to monitor cloud clusters on the distant seas and oceans, the plateau, or mountains where ground-based PR is hard to deploy, which can effectively make up for the deficiency of ground-based PR. In addition, the large precipitation particles are usually located in the middle and lower layers of cloud cluster, in the upper layer of cloud cluster, radar beam form satellite attenuated more lightly than that from groundbased because of the work type of satellite PR detection is top-down, which is available for obtaining the structural information of the upper layer of cloud cluster [15].

The reliability of GPM DPR data has been evaluated and verified by many scholars. Liao et al. [16] conducted a physical evaluation of the rain profiling retrieval algorithms for DPR on board the GPM Core Observatory satellite and proved that the DPR dualwavelength algorithm can generally provide accurate rainfall rate. Jin et al. [17] evaluated the applicability of the GPM data in Mount Taishan area, and the results showed that the GPM had the highest accuracy in the mountainous area that could estimate the precipitation system with more accurate accuracy and lower error. Zhang et al. [18] revealed that the GPM DPR inversion product was more capable of revealing structures of both strong and light precipitation through individual cases and statistical analysis. Lasser et al. [19] compared the precipitation observation data measured by GPM DPR with that measured by ground weather stations, and the result showed that the precipitation observation data measured by GPM DPR was basically consistent with the data measured by ground meteorological station. Sun et al. [20] compared the GPM DPR data with the measured results of onedimensional (1-D) laser-optical particle size velocity (PARSIVEL) disdrometers over the Yangtze-Huai River Valley in central China, and found that the measured results were similar, the mean deviation was relatively small, and the skewness was close to zero. 16 laser-optical PARSIVEL disdrometers deployed in Sichuan were used to validate the GPM-DPR results by Li et al. [21] verified that the DPR data in this area are basically credible. However, it is also mentioned that due to the complex topography of Sichuan Province and the difference of measurement principles between DPR and disdrometers, it is difficult to have identical and ideal conditions to evaluate DPR data by PARSIVEL disdrometers. All of the above comparative evaluation works show that the GPM DPR data are very reliable.

Drop size distribution (DSD) and its spatial variability is not only essential in understanding the microphysical processes that occur at the different stages of a precipitating system, but also useful in microwave communication, soil erosion and landslide triggering

studies [22–24]. Seela et al. [25] studied the DSD variability in summer season rainfall between two islands in the western Pacific region and the results indicated that both orography and aerosol loading are responsible for the spatial variability of DSD. Harikumar et al. [26] made a comparative study on the data of four tropical stations in the peninsular India and found that orographic rain has larger drops when compared with non-orographic rain when the rain rate is high. Zwiebel et al. [27] studied the impact of complex terrain located in France on the structure of rainfall and mentioned that the evolution of the DSD over the transition and plain areas is dominated by coalescence and evaporation.

In this study, DPR observations are used to analyze the radar reflectivity factors and the vertical structural characteristic of DSD of SP and CP over different terrain regions (plains, mountains, high mountains) in the Sichuan Basin and its surrounding areas, expecting to obtain the characteristics of precipitation structure, which will help in further understanding the influence of mountain topography on precipitation structure and internal microphysical processes, and it also plays a very important role in deepening scientific recognition of mountain heavy rainfall mechanism. The remainder of this study is arranged as follows: The dataset used and the methodology are given in Section 2. Section 3 provides the results, including the vertical structure characteristics, horizontal distribution characteristics, and the relationship between each physical quantities of stratiform and CP over different types of terrain. The discussion is provided in Section 4 and followed by a summary in Section 5.

### **2. Data and Methods**

GPM Core Observatory takes about 93 min to fly around the earth. The global coverage is from 68◦S to 68◦N, and the flight heights is at the altitude of 407 km. The GPM radar is able to provide observations of the 3D structure of precipitation from the surface to 22 km altitude. The DPR instrument is made of a Ka-band precipitation radar (KaPR) operating at 35.5 GHz and a Ku-band precipitation radar (KuPR) at 13.6 GHz, and the KuPR and KaPR will provide coaligned 5-km-resolution footprints on Earth's surface. 2A.GPM.DPR is the dataset of DPR Ku and Ka-band radar reflectivity profile and radar-based precipitation retrievals. The dataset carries three radar profiles, including the Ku-band normal scan (NS), the Ka-band matched scan (MS), and the Ka-band high-sensitivity scan (HS). The NS has a nominal vertical range resolution of 250 m with cross-track swath widths of 245 km, the MS has a nominal vertical range resolution of 250 m with cross-track swath widths of 120 km, and the HS has a nominal vertical range resolution of 500 m with crosstrack swath widths of 120 km [12].The 2A.GPM.DPR database V06 (covering the period from May to September of 2014–2021) has been used in the study, which can provide detailed precipitation information, including reflectivity factor with attenuation correction (Ze), DSD, storm top altitude (STA), freezing height (FzH), rain rate (RR), precipitation type, and so on. In single frequency classification (CSF) modules, i.e., in Ku-only and Ka-only modules, precipitation type classification is made by a Vertical profiling method (V-method) and by an Horizontal pattern method (H-method) [28,29]. In the dual frequency module, in place of the V-method, the measured dual frequency ratio (DFR m) method is used for precipitation type classification and for BB detection in the inner swath [30,31], classifying rain into stratiform, transition, and convective. More details can be found online at https://gpm.nasa.gov/sites/default/files/2019-05/ATBD\_DPR\_201811\_with\_ Appendix3b.pdf (accessed on 23 August 2022).

The research regions are the Sichuan Basin and its surrounding areas of China (99◦E– 109◦E, 27◦N–33◦N). ETOPO1 is a 1 arc-minute global relief model that was developed by the National Geophysical Data Center (NGDC), an office of the National Oceanic and Atmospheric Administration (NOAA), which was used to divide the research regions in this study. Fan et al. [32] found the best statistical window of topographic relief in Sichuan to be 9.92 km2, while the statistical window of topographic relief is defined as 13.69 km<sup>2</sup> in this paper because of the limit of resolution of ETOPO1. Topographic relief is the difference between maximum and minimum of altitude. Referring to related research results [33–35],

dividing the research regions into three types of topography by ETOPO1 (Figure 1): 1. Plains with altitude from 0 to 1500 m, topographic relief < 100 m; 2. Mountains with altitude from 500 to 1500 m, topographic relief ≥ 200 m; 3. High mountains with altitude from 1500 to 4000 m, topographic relief ≥ 200 m. In Figure 1b, the white areas represent the place where topographic relief does not meet the requirements of three categories mentioned above, e.g., the white areas of 1500–4000 m above mean sea level represent high-altitude areas with flat terrain as well as with few rainy pixels, therefore not going to study such areas. For statistical analysis, the number of rainy pixel samples occurred in study regions during the period May to September of 2014–2021 was counted based on GPM DPR data (Table 1), and Figure 2 shows the spatial distribution of the number of stratiform and convective rainy pixels.

**Figure 1.** Administrative regional division of the People's Republic of China (**a**), and geographical division of Sichuan Basin and its surrounding areas; (**b**) The green, yellow and brown areas represent the plains, mountains and high mountains respectively.

**Table 1.** Number of convective and stratiform rainy pixels of different types of terrain in Sichuan Basin and its surrounding detected by GPM DPR from May to September of 2014–2021.


Numbers in brackets denote SP, and numbers outside brackets denote CP.

**Figure 2.** The spatial distribution of heavy stratiform (**a**) and convective (**b**) rainy pixels. The heavy precipitation occurs over plains (green dots), mountains (yellow dots) and high mountains (brown dots).

The contoured frequency by altitude diagram (CFAD) is an effective method to represent the vertical structure of precipitation that has been applied in many studies [36,37]. The entire frequency distribution normalized by dividing by the maximum absolute frequency of the samples within the region of analysis is NCFAD, which permits the comparison of CFADs between regions despite the different absolute frequencies [38]. In this paper, NCFAD is mainly used to statistically analyze the vertical structure characteristics of SP

and CP that occurred over plains, mountains and high mountains regions of the Sichuan Basin and its surrounding area. RR ≥ 20 mm/h is defined as heavy precipitation [39], and 2A.GPM.DPR retrievals with RR < 0.5 mm/h are discarded from this study because of the limitations of dual-frequency retrieval. For the detailed discussion, the ground precipitation grades were divided into 5 levels, including 0.5 < RR ≤ 2 mm/h, 2 < RR ≤ 4 mm/h, 4 < RR ≤ 8 mm/h, 8 < RR ≤ 20 mm/h and RR ≥ 20 mm/h. Since shallow rain is archived as CP in DPR retrievals, excluded it from the CP, and only precipitation events in which the phase behavior of near-surface particles is liquid are considered. The STA, FzH, and bright band height in this study are absolute heights altitude relative to sea level, not relative to the surface. In addition, the horizontal distribution of STA and other physical quantity of DPR orbit dataset were gridded at 0.25◦ × 0.25◦. For example, the average STA denotes the ratio of the sum of STA within the 0.25◦ × 0.25◦ to the total number of samples within that area.
