Response of a westerly-trough rainfall episode to multi-scale topographic control in southwestern China

Wei Luo , Hnyue Yin , Shui Yng , , Yushu Zhou , Lingkun Rn , Bofeng Jio , Ziyng Li

a CHN Energy Dadu River Big Data Services Co., LTD, Chengdu, China

b Laboratory of Cloud-Precipitation Physics and Severe Storms (LACS), Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, China

Keywords:Heavy rainfall Trough Topography Numerical simulation

ABSTRACT This study aims to quantify the response of a westerly-trough rainfall episode that occurred in summer 2020 to multi-scale topographic control in southwestern China, based on observations and numerical simulations.The multi-scale topography is composed of the Tibetan Plateau, Hengduan Cordillera (HC), and Sichuan Basin(SB).The westerly trough was characterized by southeastward deepening together with an in-phase propagating rainfall episode.By utilizing the results of numerical experiments, how the multi-scale topography impacted this westerly trough rainfall episode is explored.It is found that HC was the pivotal topographic factor affecting the southeastward extension of the trough and related rainfall, while SB accerelated the eastward movement of the westerly trough and changed the tilting direction of the trough line, thus further changing the location and orientation of precipitation.For extreme rainfall with intensity exceeding 10 mm h ? 1 , a roughly threefold rise in the cover ratio (from 1.8% to 7.2%) and fourfold increase in the areal rainfall amount per hour occurred by removing the HC barrier, due to the strongest vorticity and long-distance transport capacity to potential vorticy mass accompanying the southeast-stretching trough.Our results quantitatively reveal a strong response of westerly trough rainfall to multi-scale topographic control in southwestern China, therefore serving as an important reference for future decision making and effective model improvement.

1.Introduction

Under complex topographic forcing, heavy rainfall frequently occurs in mountainous southwestern China every year ( Luo et al., 2016;Fu et al., 2019 ; Romanic, 2021 ), forming one of the rainiest regions in China and causing severe flood disasters.The particular multi-scale topography of this region, composed of the Tibetan Plateau (TP), Hengduan Cordillera (HC), and Sichuan Basin (SB), brings great difficulty to making accurate predictions of heavy rainfall, or even leads to unstable numerical simulation over steep terrain transition zones.

The thermodynamic effect and mechanical dynamic forcing of multiscale topography provide suitable environmental conditions to breed vortices or troughs over southwestern China, which are conducive to active convection development and frequent rainfall occurrence in summer ( Gao, 2000; Yu et al., 2007; Zhou et al., 2019 ; Lin et al., 2021; Wang et al., 2021 ).Plateau vortices and southwest vortices are important vortex systems inducing heavy rainfall in mountainous southwestern China( Lu et al.2020 ).Beyond that, westerly troughs are also a favorable synoptic system, as a significant disturbance superimposed on the midlatitude westerlies, to easily induce rainfall in this area ( Wang and Chen,2007; Yu et al., 2007 ; Si et al., 2018 ).

The westerly trough is a typical mobile system, which commonly brings rainfall in front of the trough by positive vorticity advection and ascending motion.The dynamical and physical processes leading to vorticity growth under westerly-trough forcing has been studied for many years ( Leroux et al., 2013; Liu et al., 2021 ; Ren et al., 2021 ).These analyses demonstrate that terrain might play a forcing role in the evolution of the westerly trough, thus further influencing associated rainfall activities.Once the trough moves southeastwards out of the TP, it also has the potential to greatly impact the densely populated downstream areas(e.g., the SB and the middle and lower reaches of the Yangtze River),easily causing geological disasters such as mudslides, mountain torrents,and urban waterlogging.Therefore, southeastward-moving type systems and related physical processes during heavy rainfall in southwestern China should be paid more attention.

In view of the significant correlation between trough evolution and precipitation development under complex topographic forcing, the impact of multi-scale topography on westerly troughs and related rainfall should be detected, which might be the key to improving the simulation and prediction of westerly trough precipitation from a terrain perspective in this region.Therefore, we aim to reveal the key topographic factor impacting westerly-trough rainfall based on terrain-sensitive numerical experiments in a case study.The horizontal pattern and temporal evolution of the rainfall case is shown in Figs.1 and 2 .The remainder of this paper is organized as follows: The model and experimental design are briefly described in Section 2 .The evolution of the westerly trough,its correlation with the evolution of precipitation, and the multi-scale topographic controls on the westerly-trough and extreme rainfall, are investigated in Section 3 .A summary is given in Section 4 .

2.Numerical model and experimental design

2.1. Model

Version 4.2.1 of the Weather Research and Forecasting model was used to perform numerical simulations ( Fig.1 ).The domain covered southwestern China with 601 × 601 horizontal grid points and 2.7 km grid spacing centered at (28.1°N, 102.2°E) and with 61 vertical levels from the surface to 10 hPa.We adopted the YSU boundary layer scheme,the unified Noah land surface model, RRTM longwave/Dudia shortwave radiation options, and the WSM6 microphysics scheme.

The initial and lateral boundary conditions were derived from the GFS 0.25° × 0.25° operational analyses available every 6 h.The model was integrated from 0000 UTC 17 June 2020 and lasted 48 h.Hourly CMORPH (CPC Morphing Technique) precipitation with a 0.1° resolution was used to examine the simulation of precipitation.

2.2. Experimental design

Besides the above control (CNTR) model run, we carried out three additional sensitivity experiments ( Fig.3 (a)), named TP, TP + HC, and ALL (i.e., TP + HC + SB), to examine the individual topographic effects of TP, HC, and SB on the westerly-trough rainfall ( Fig.3 (b–e)).In the three experiments, the topographic elements were gradually added and their differences compared.

In terms of the idealized terrain construction (as shown in Fig.3 (a1–a3)), the geometric shapes were composed of an ellipse, a circle, and a nearly rounded rectangle to approximate the TP, SB, and HC.For more details, please refer to Wang and Tan (2006 , 2014) and Li et al.(2021) .Note that such a design is convenient to adjoin and split certain regular terrain components freely, so as to facilitate resolving/distinguishing the individual roles of terrain components.

3.Results

3.1. General description of the rainfall event

In summer 2020, a heavy rainfall episode occurred in Sichuan Province, featuring southeastward propagation from the TP to HC and SB ( Fig.1 ).The peak rainfall intensity exceeded 100 mm h?1and 250 mm/24 h, resulting in catastrophic floods and serious geological disasters near the steep terrain adjoining the TP, HC, and SB.

The main rain band presented a “U-shaped ” pattern ( Fig.1 ).The lefthand branch of the rainfall (west of 105°E) propagated southeastwards from the TP along the steep terrain adjoining the TP, SB, and HC from 0600 UTC 17 to 1200 UTC 18 June 2020 ( Fig.1 (a–f)).From the synoptic chart, the U-shaped rain band along the steep terrain was closely associated with the passage of a westerly trough.The right-hand branch of the rainfall (east of 105°E) maintained stably ( Fig.1 (a1–f1)), with a strong rainfall center near 108°E ( Fig.1 (a1–c1)) due to a low vortex (c.f., 700-hPa synoptic chart, omitted herein).In view of the dual influences of complex terrain on the trough itself and evolution of precipitation, the terrain-related rainfall will be the focus in the following analysis.

3.2. Model validation

Fig.1 (a2–f2) shows the evolution of the simulated 6-h accumulative precipitation and the moving westerly trough.Compared with the observation ( Fig.1 (a1–f1)), the model reproduces the pattern and trend of the precipitation and trough, especially the U-shaped rain band along with the southeastward-stretching trough from 1200 UTC 17 to 0000 UTC 18 June 2020 ( Fig.1 (b1–d1)).The locations of simulated strong precipitation centers also approach the observed ones, e.g., the precipitation between 28° and 30°N, near 102°E at 1800 UTC 17 June 2020.

The precipitation occurs in front of the trough along the steep terrain( Fig.1 (a2–f2)), dominated by positive vorticity advection.The precipitation propogates southwards and moves eastwards, synchronous with the trough evolution.

3.3. Westerly-trough rainfall evolution features

Fig.2 shows longitude–time diagrams of hourly precipitation and zonally averaged (between 26° and 30°N) relative vorticity and vertical velocity at 500 hPa for the CNTR run.During the deepening stage of the trough before 0600 UTC 18 June, the strong vorticity concentrates west of 104°E, with the maximum intensity exceeding 2.0 × 10?4s?1, and propagates eastwards from 0000 UTC 17 to 0600 UTC 18 June( Fig.2 (c)).The positive vorticity zone expands between 98° and 104°E( Fig.2 (c)), accompanied by active convection ( Fig.2 (b)) and precipitation ( Fig.2 (a)).The vertical velocity is larger than 0.6 m s?1, and the precipitation intensity reaches up to 3.5 mm h?1.

After 1200 UTC 18 June, plateau precipitation between 98° and 100.5°E generates again, characterized by a significant diurnal cycle of nocturnal rainfall (compared with the rain band over the plateau during 1200 UTC 17 to 0000 UTC 18 June).Meanwhile, the vorticity and convection are obvious within the precipitation corridor.In the first half of the event, the drastic vorticity growth due to southward movement and deepening of the trough before 0600 UTC 18 June, together with the convection development in front of the trough, are the main causes of heavy rainfall.

3.4. Multi-scale topographic influences

3.4.1.Spatialpattern

Fig.3 shows the spatial pattern of the westerly-trough rainfall under various configuration scenarios among the multi-terrain elements( Fig.3 (a)).The ALL run is used as a baseline to perform comparative analyses ( Fig.3 (b–e)).

For the ALL experiment ( Fig.3 (b3, c3, d3, e3)), it can be seen that the trough line extends southeast to 25°N at 500 hPa ( Fig.3 (b3)), producing strong cyclonic curvature circulation accompanied by strong vorticity(shaded) at 500 hPa and even 700 hPa ( Fig.3 (c3)).Convection and precipitation develop in front of the trough.

Relatively, if the SB is removed ( Fig.3 (a2, b2, c2, d2, e2)), the eastward transfer of the trough is slowed down in the TP + HC experiment(cf., Fig.3 (b2, b3)).In other words, the basin terrain attracts and accelerates the eastward transfer of the trough, induces a tilting trough line towards the basin side and a narrower trough width, and therefore changes the pattern and orientation of precipitation (cf., Fig.3 (d2, e2)and Fig.3 (d3, e3)).

Fig.1.(a–f) Observed (left-hand panels, a1–f1) and simulated (right-hand panels, a2–f2) 6-h accumulative precipitation (colored; units: mm/6 h) at 6-h intervals from 0600 UTC 17 to 1200 UTC 18 June 2020.The gray shading represents terrain height (units: m), and blue contours denote geopotential height (units: dgpm).

Fig.1.(Continued).

Further, the HC run ( Fig.3 (a1)) results in greater changes in the trough and related precipitation without the obstruction of the HC barrier.The southward extension of the trough is greatly accelerated and the trough line rapidly reaches 22°N to the south of the TP at the 500-hPa level ( Fig.3 (b1)).Correspondingly, the southward-moving rain band is accelerated, quickly sweeping most of the southern areas within the domain ( Fig.3 (d1, e1)).

Fig.2.Time–longitude diagrams of (a) hourly precipitation (shaded; units: mm), (b) zonally averaged (26°–30°N) vertical velocity (units: m s ? 1 ), and (c) relative vorticity (shaded; units: 10 ? 4 s ? 1 ) at 500 hPa for the CNTR run.

Overall, the HC is the pivotal topographic factor affecting the southward extension of the trough and related rainy region, while SB accerelates the eastward movement of the trough and changes the tilting direction of the trough line, thus further changing the location and orientation of precipitation.Relatively speaking, the degree of impact of the HC on the evolutions of the trough and precipitation is particularly significant.If both HC and SB are removed, the coverage of precipitation to the south of the TP increases dramatically.

3.4.2.Temporalevolution

From the spatial patterns of rainfall under various terrain scenarios( Fig.3 (d, e)), the TP experiment exhibits a special mode, but the other two simulations take on similar patterns.Therefore, we compare the cases of the TP and TP + HC runs ( Fig.4 ) to illuminate the role of TP in the evolution of precipitation.Both have clear southward-propagating trends from 34°N to 22°N, and obvious nocturnal rainfall between 33°and 30°N during 1200 UTC 17 June to 0000 UTC 18 June.However,the precipitation shifts faster for the TP run in the case without the HC barrier (cf., Fig.4 (a) and 4(d)), determined by quick-growing vorticity( Fig.4 (c, f)) and convection ( Fig.4 (b, e)) due to the rapidly deepening trough.For example, precipitation spreads southwards from 30°N to 26°N within 12 h (0000–1200 UTC 17 June) for the TP run, while it takes nearly 24 h (0000 UTC 17 to 0000 UTC 18 June) for rainfall to finish this route in the TP + HC experiment.Furthermore, compared to the TP + HC simulation, the TP run presents a multi-rainband mode and a longer duration, producing a stronger areal rainfall amount(cf., Fig.4 (a, d)).

3.4.3.Extremerainfall

To further quantify the respective effects of multiple terrain elements on precipitation, rainfall intensity, areal rainfall amount, and precipita-

Fig.3.Idealized terrain in the (a1) TP run, (a2) TP + HC experiment, and (a3) ALL run.(b, c) Vorticity (colored; units: 10 ? 5 s ? 1 ), terrain (gray contours; units: m),and geopotential height (blue contours; units: dgpm) at 500 hPa and 700 hPa at 1800 UTC June 2020.(d, e) 6-h precipitation (colored; units: mm), terrain (gray shading; units: m), and geopotential height (blue contours; units: dgpm) at 1200 UTC and 1800 UTC 17 June 2020.Left, middle, and right panels are cases for the TP, TP + HC, and ALL runs, respectively.

tion cover ratios are calculated as indices to estimate the differences among the three experiments ( Fig.5 ).

Fig.4.Time–zonal diagrams of the meridional-mean (100°–105°E) (a) hourly precipitation (shaded; units: mm), (b) vertical velocity (shaded; units: m s ? 1 ), and (c)relative vorticity (shaded; units: 10 ? 4 m s ? 1 ) for the TP experiment.(d–f) As in (a–c) but for the TP + HC run.

The TP case brings about the maximal rainfall intensity(5.84 mm h?1) and cover ratio (42.1% of the target region) due to drastically-deepening trough southwards, and therefore produces the strongest areal rainfall amount (182281.55 mm h?1) relative to the other two model runs in which the three indices remain equivalent in magnitude.For extreme rainfall (rainfall intensity>10 mm h?1herein), the TP run still accounts for the largest proportion of areal rainfall amount (122481.25 mm h?1), because of the coaction of the combined wider coverage (7.2% cover ratio) and stronger rainfall intensity ( ～21.44 mm h?1).Comparing the TP and TP + HC runs, a roughly 3–4 times increase in the cover ratio (from 1.8% to 7.2%) and areal rainfall amount (from 22863.42 to 122481.25 mm h?1) occurs by removing the HC barrier.

Vorticity and potential vorticity (PV) can be utilized as metrics to indicate the strength and influence range of the trough.A southwardstretching trough signifies robust long-distance transport of PV mass from the high PV reservior at high latitudes towards the trough bottom (see the strong signal of PV in Fig.6 (a)).Therefore, we estimate the top-10% vorticity and standardized distance of PV meridional transport (scaled by using the meridional range of the domain as a reference length) for the three model runs, to quantify and explain how precipitation change responds to strength and stretching variations of the trough due to the various topographic effects ( Fig.6 (b)).The results show the strongest vorticity (4.12 × 10?4) and the farthest transport path (0.85,a dimensionless variable) in the TP run, relative to the other two experiments.Due simply to the super strength and range of influence of the trough, large rainfall intensity and coverage ratios are brought, therefore producing the large areal rainfall amount.

4.Summary

A westerly trough heavy rainfall episode near the steep terrain adjoining the TP, HC, and SB was investigated to explore the influence of multi-scale topography.The main conclusions can be summarized as follows.The HC is the pivotal topographic factor affecting the southward extension of westerly-trough rainfall.The SB accerelates the eastward movement of the system, changes the tilting direction of the trough line, and therefore the location and orientation of precipitation.Relatively speaking, the former is particularly significant for the evolution of rainfall.For extreme rainfall, the rainfall coverage to the south of TP increases dramatically (a roughly threefold rise from 1.8% to 7.2%in the cover ratio) without HC blocking, resulting in an increased areal rainfall amount (from 22863.42 to 122481.25 mm h?1) by removing the HC barrier.Two metrics to characterize trough strength and the range of influence of the trough –vorticity and the standardized distance of PV meridional transport –exhibited strong signals in the TP run, which was responsible for the change of rainfall due to the HC effect.

In a previous study, Li et al.(2021) analyzed the multi-scale topographic influence on the vortex development for an eastwardpropagating rainfall event that occurred in southwestern China.They found that the HC played a key role in the formation of southwest vortex,while SB influenced its location and intensity.Wang and Tan (2014) also emphasized the significance of the HC topography, besides the TP, on southwest vortex formation through providing the source of the vortex stream.

Our results focus on the westerly trough, a low-value system in nature, similar to a vortex to some degree, consistent with these previous studies.We reveal a strong response of the westerly trough to HC change.Furthermore, we demonstrate the influence of HC on westerlytrough and extreme rainfall, and show some quantitative analysis results.In future work, the impacts of slope change near the steep terrain at the boundary among TP, HC, and SB on the evolution of troughs and related rainfall are expected to be explored.Beyond that, more thermodynamic and dynamic aspects based on vorticity and PV equations are needed to reveal the mechanism responsible for the genesis and dispersion of the westerly trough itself, and its correlation with rainfall variation.

Fig.5.The (a) rainfall intensity (units: mm h ? 1 ), (b) areal rainfall amount(units: mm h ? 1 ), and (c) precipitation coverage ratio (units: %) for various topography scenarios.

Fig.6.The (a) potential vorticity (units: PVU) and (b) peak distance and vorticity (units: 10 ? 5 s ? 1 ) in the TP run, TP + HC experiment, and ALL run.

Funding

The authors were supported by the National Key Research and Development Program on the Monitoring, Early Warning and Prevention of Major Natural Disasters [grant number 2018YFC1507104],the National Natural Science Foundation of China [grant numbers 41875079, 41875056 , and 41975137 ], and a Key Technology Research project on multi-source meteorological data fusion in medium and small basins [grant number DSJ-KY-2021-004 ].