Operational Plan, Effect Verification, and Key Technical Settings for a Stadium-Scale Artificial Rain Reduction Experiment

Yuquan ZHOU, Siyao LIU, Miao CAI, Junlin LONG, and Jia WANG

1 CMA Cloud-Precipitation Physics and Weather Modification Key Laboratory (CPML), Weather Modification Centre,China Meteorological Administration (CMA), Beijing 100081

2 School of Electronic Engineering, Chengdu University of Information Technology, Chengdu 610225

3 College of Meteorological Observation, Chengdu University of Information Technology/CMA Key Laboratory of Atmospheric Sounding, Chengdu 610225

4 Jiangsu Weather Modification Center, Nanjing 210000

ABSTRACT To explore the key technologies of artificial weather modification for specific targets (e.g., a stadium) and improve the efficiency of artificial rainfall modification for major events, this study conducts an artificial rainfall reduction experiment for the closing ceremony of Nanjing Youth Olympic Games on 28 August 2014.Satellite retrievals, radar observations, sounding data, and other sources of information as well as Cloud and Precipitation Accurate Analysis System (CPAS) are used in this study.The main conclusions are as follows.(1) On 28 August 2014, a large-scale cumulus cloud system with mixed-phase stratocumulus and stratus precipitation was observed.This system was influenced by the weak shear of a low-level trough and the precipitation was dominated by cold clouds with dry layers between clouds.Thereby, we adopted the crystal-priming over-catalytic hypothesis and conducted a rocket-catalytic rain abatement operation at a certain distance (100-25 km) from the stadium.Rocket shootings of different intensities were implemented for two echoes that affected the stadium successively (two rounds of 15 rocket shootings within 15 min for an isolated weak echo IA; multiple rounds of 156 rocket shootings within 80 min for a strong echo IB).Amazingly, after the shootings with the catalysis in the air, reflectivity of the two echoes was reduced at all altitudes with the most significant reduction at the 2-km altitude, and the time needed for the obvious reduction was 40 min.The most obvious reduction of the two echoes then maintained for 60 and 53 min, respectively, and the operation time needed for the echo zone to recover after the stop of rocket shooting was 108 min for echo IA and 90 min for echo IB.The two echoes moving across the stadium during the time period of the closing ceremony (2000-2130 local time) were at their minimal strengths, with almost no echo over the target stadium.This demonstrates that the rocket shooting strategy of over-crystallization catalysis is effective, and the shooting site, time, and dose are reasonable.The following technical parameters were used during this experiment.At about 80-25 km away from the target stadium in the west, the rocket shooting lasted for 15-80 min and the doses were not less than 1 shot min-1 (1 shot min-1 for echo IA, 2.25 shots min-1 for echo IB).The attenuation rate was 0.21 dBZ min-1 for the average 15 dBZ of echo IA.For the average 25 dBZ of echo IB, the attenuation rate was 0.27 dBZ min-1.The above technical settings helped achieve the goal of reducing rain over the stadium to almost zero for nearly 1-h period during the critical time of the event.

Key words: target area, artificial rainfall abatement, radar monitoring, weather modification plan, effect analysis

1.Introduction

Weather modification technologies sometimes are used with the hope to achieve rain reduction in specific areas (event venues, celebration squares, etc.) and at specific times (opening and closing ceremony periods, celebration periods, etc.) to reduce adverse effects of precipitation and ensure smooth operations of some major events.With the increase in various national events, the demand for weather modification to achieve specific goals has also increased.

Artificial rainfall reduction is a type of weather modification, and scientists at home and abroad have proposed three scientific hypotheses for artificial rainfall reduction: “early rainfall,” “excessive seeding,” and “dynamic descent.” In 1950, Langmuir first proposed the concept of “excessive seeding” for artificial rainfall reduction (Langmuir, 1950), which involves seeding an excessive amount of catalysts at appropriate time and location to produce high concentrations of ice crystals that compete with supercooled liquid water (vapor) and thereby reduce the size of water droplets and achieve the goal of inhibiting or delaying rainfall.Ultimately this method can result in the reduction of precipitation in the protected area during the protection period.The “White-Top” project in the United States confirmed the correctness of Langmuir’s “excessive seeding” theory for rain reduction (Lovasich et al., 1971).“Dynamic Sinking” is mainly a cloud suppression technology applied to convective clouds, which involves spreading coarse, dispersed particles at the top of the cloud to induce downdrafts that affect the cloud dynamic process and suppress the development of convective clouds, ultimately achieving the goal of reducing convective clouds(Mason, 1971).“Early precipitation” was initially derived from small-scale artificial seeding experiments and numerical studies, which showed the weakening of precipitation or movement of precipitation downwind under certain conditions.Some studies suggest that small-scale seeding may alter the distribution of mesoscale precipitation through the transport of catalysts and changes in atmospheric water content as well certain dynamic effects,which is also called the extraterritorial effect (Dennis and Brown, 2012).

The above three hypotheses of rain-reducing methods have been explored by scientists internationally.Among the three hypotheses, the theory of “excessive seeding”for rain reduction is relatively mature (Dennis and Koscielski, 1969; Miller et al., 1979; Ye et al., 1998; Zhang et al., 2008).However, the issue of optimal amount of seeding for increasing or reducing precipitation in different types of cloud remains unclear.The theoretical optimal concentration of precipitation particles that can be catalyzed is approximately 1 L-1and the maximum is about 10 L-1.If the concentration exceeds this range, the particles become too small to fall.Several studies that combined practical operations and numerical simulations were done.Wang et al.(2005) found that within an operational area of the volume ofV= 100 km3, 2.2 rockets are required for rain enhancement.Li et al.(2005)showed that 1-3 rockets need to be launched for enhancement of rain from stratiform clouds.Wang et al.(2007) suggested that 2-3 rockets should be launched for a single rain enhancement operation.Wang et al.(2008)found that for an operational area ofV= 240 km3volume, 5.3 rockets are needed for rain enhancement.The above studies generally suggest that the number of rockets needed for cloud seeding is usually below 5.However, for rainfall reduction, an excessive amount is required, and the threshold for excess is still under exploration.In practice, a catalytic operation with as much substance as possible that exceeds the amount used for cloud seeding is often employed.

Investigations on the time of physical response after dissemination of the catalyst were also performed.Wang et al.(2021) and Yue et al.(2021) studied a cold cloud seeding operation by airplane in Shanxi Province on 19 March 2017.They found that cloud gullies were visible in the Moderate Resolution Imaging Spectroradiometer(MODIS) images from theTerrasatellite 38 min after seeding and in the Visible and Infrared Radiometer(VIRR) images from theFengyun-3C(FY-3C) satellite 49 min after seeding.Impact echoes appeared on the radar about 18 min after seeding and lasted for about 268 min.Cai et al.(2013) analyzed the artificial rainfall operation process in Hebei Province on 18 April 2009 using the K-value method, and found that the rainfall in the operation area began to increase 1 h after the operation and reached its peak 2 h after the operation.They believed that the operation effect lasted for 3 h.Many numerical simulation studies have shown that there is an increase in rainfall after 20-30 min of catalysis, and the maximum net rainfall reduction occurs between 60-90 min.The rainfall then gradually decreases and the process generally lasts for 3 h.The response time of rainfall enhancement may vary depending on different cloud conditions(He et al., 2013; Liu W.G.et al., 2016, 2018; Liu X.E.et al., 2016).However, the above studies mostly focus on understanding the effects of rainfall enhancement.

Some experiments have been carried out both domestically and internationally on artificial rain reduction over specific target areas such as a square or a sports field.In 1986, severe leakage of the Chernobyl nuclear power plant occurred in the former Soviet Union.If there had been rainfall following the leakage, it would have caused water source pollution in the entire watershed and exacerbated the impact of the accident.The meteorological department used more than 10 aircrafts for artificial rain reduction, and there was no ground precipitation in the nuclear radiation-affected area for half a month.In those important conferences and events such as the “Eight-Nation Summit” in St.Petersburg, the Moscow Red Square military parade, the opening ceremony of the International Youth Games, etc., attempts have been made to carry out weather modification operations, and these celebratory events were not affected by precipitation (Zhang et al.,2006).However, most of these conclusions are based on news and/or reports, whereas no relevant scientific publication has been found.

In recent years, artificial weather modification in China not only has played an important role in agriculture production and disaster prevention and relief, but also has been actively implemented in the field of artificial rainfall suppression, which is increasingly demanded for major events (Zhang et al., 2008; Hong and Lei,2012; Guo et al., 2019).For example, various artificial weather modifications were conducted in the 2008 Beijing Olympic Games, the 2014 Nanjing Youth Olympics, the 2016 G20 Summit, the 2019 National Parade, the 2021 celebration of the centennial of the Communist Party of China, etc.Analysis of the effects of artificial rainfall reduction operations, however, is a difficult issue worldwide.Considerable different methods including numerical experiments and statistical and physical tests have been tried.Among these methods, physical test using satellite observations, radar data, and ground rainfall measurements is a commonly used method.Many studies such as Zhang et al.(2008), Zhou et al.(2009), Li et al.(2011), and He and Ma (2008) have analyzed the macroscopic and microscopic characteristics of clouds in the rainfall reduction operation during the 2008 Olympic Games using rainfall, radar, and satellite data.The results show that the large-dose rocket seeding operation overall has a certain inhibitory effect on the development of precipitation.However, due to the complexity of cloud microphysics and precipitation, it is still difficult to determine whether the evolution of precipitation in the entire cloud is affected by the local catalytic effect.

Weather radars can obtain observations of three-dimensional cloud and precipitation structures on high spatial-temporal resolutions, which makes them an important tool in the study of the effects of both cloud seeding and rain suppression.Zhou et al.(2001) found that the cloud seeding process can cause an enhancement in the maximum echo intensity of the seeded cloud, and the enhancement has a good correlation with the catalytic process.Jiang et al.(2006) analyzed radar echo changes in the seeding and control areas during two artificial rainfall experiments in the Jianghuai area of central China.They found that both seeding operations had the effect of rainfall enhancement with the relative increase of 120%-140%.Many researcher have attempted to automatically track radar echoes to explore the effects of cloud seeding operations.Foote et al.(2005) and Krauss and Sinkevich (2007) used the Thunderstorm Identification, Tracking, Analysis, and Nowcasting (TITAN) to track radar echoes and analyzed the effects of cloud seeding, and achieved good results for convective clouds.However, TITAN is not suitable for layer clouds typically targeted by cloud seeding.Chen et al.(2012) used the Tracking Radar Echo by Correlation (TREC) method to track radar echoes and determine the movement of the seeding area at different times.They analyzed a rocket cloud-seeding operation on a stratocumulus and stratus mixed cloud system in Beijing and obtained some understanding of the seeding effects.

Ni (2017) employed the cross-correlation TREC method to track individual echoes based on observations of a C-band dual-polarization full-coherent pulse Doppler weather radar system and sounding data as well as ground precipitation data.The effects of weather modification during the opening and closing ceremonies of the Nanjing Youth Olympic Games were analyzed by calculating changes in the ratio of echo intensity between upstream and downstream of the operation point.The relative reduction in reflectivity factor was used to reflect the effectiveness of the operation.During the target period of rainfall reduction, both the echo intensity and echo top height showed a decreasing trend.However, the TREC algorithm itself has some drawbacks.First, it assumes that radar echoes are a linear evolution process and only considers the horizontal movement of the echoes.However, in the actual movement of the cloud, the corresponding size, position, and intensity of the echoes are undergoing complex changes.Second, it only infers echoes based on radar images at different times within a certain time period, and the extrapolation time is short.Finally and most importantly, the splitting and merging of radar echoes is not well handled by the TREC algorithm.These uncertainties have a significant impact on the analysis of the subtle changes in radar echoes caused by artificial operations.Using weather radar data, Song et al.(2017) conducted a detailed analysis of the evolution process of a locally generated meso-γ-scale convective cloud during the rain mitigation operation for the 70th Anniversary Parade Commemorating the Victory of the Chinese People’s War of Resistance against Japanese Aggression.They found that the echo of the cloud weakened after the operation, but its lifespan was significantly longer than that of a natural cloud.This study well explained the effect of rain mitigation operation.However, similar to other studies, no analysis or summary is provided in the study of Song et al.(2017) regarding the design of the operation plan, the implementation of specific technical parameters, and their impact on the operation’s effectiveness.

The present paper is based on accurate spatial-temporal matching of radar and operation information and subjective judgment of the evolution process of echoes.Detailed analyses of different echo blocks at different times and heights are combined to avoid certain uncertainties in previous studies and overcome the shortcomings of automatic tracking methods.The method used in the present study is a trustworthy method for in-depth analysis of the subtle dynamic changes in echoes before and after the rain reduction operation.

In summary, currently there is still limited research on the design and effect analysis of artificial operations targeting specific areas, and there is few detailed analysis and evidence to prove that the cloud or rain changes in the target area are caused by the catalytic operations.Relevant operation plans and technical parameters are also lacking in analysis and research, and the explanation of the principle hypotheses for operation implementation is not clear.Complete pre-design of artificial operation and verification of effects of weather modification are far less than sufficient, and there is a lack of summary and understanding of the relevant operational parameters that have been proven to be effective in artificial weather modification operations.

At present, artificial rain reduction is still in the experimental stage and thus is within the scope of scientific experiments worldwide.The technology for artificial rain reduction needs to be validated in more physical experiments that are well designed with science-based hypothesis, and the results should be verified against observations.To meet the increasing demand for artificial rain reduction in China, it is imperative to address key difficulties and carry out scientific and meticulous case studies to better verify the effectiveness of artificial rain reduction and summarize the operational characteristics of artificial rain reduction.Research and refinement of relevant key technical parameters are exploratory work that is also practically valued and needed at the current stage.

On 28 August 2014, the second Youth Olympic Games (YOG) came to its last day at the Nanjing Olympic Sports Center Stadium.In order to ensure a successful closing ceremony, an artificial rain reduction experiment was organized and carried out from 2000 to 2130 local time (LT), following a pre-designed technical plan with the clear goal of rain reduction at the above specific location and period.As the technical expert group leader, the first author of the present paper took part in the entire process, from scheme design, proposal of technical settings, on-site command and coordination,and verification of the results against observations.

Based on analysis of the catalysis principle and the characteristics of the target clouds, the present study first proves that excessive catalysis of cold clouds is required for this operation, and then provides the corresponding design for this artificial rain reduction.Based on radar data and operational information, a detailed analysis is then given on the operational conditions (indicators) and evaluation methods as well as appropriate parameters for the effectiveness of the artificial rain reduction during the closing ceremony of the YOG.The present study attempts to explore and verify the rationality of the operation plan, and summarize and refine the key technical parameters used in the artificial rain reduction operation.

2.Data, platform, and operational zone

2.1 Data

Doppler radar can obtain three-dimensional structure of cloud and precipitation.The continuous observations at high temporal resolution of 6-min are very helpful for us to conduct detailed analysis of the evolution of cloud and precipitation structure.In this study, observations of the S-band 10-cm radar in Nanjing are used to conduct a detailed analysis of the horizontal and vertical structures and evolution process of cloud and precipitation during the focus period of 1600-2130 LT 28 August 2014.

The FY-2 series of satellites are the first generation of geostationary meteorological satellites independently developed in China.So far, six satellites (FY-2A/B/C/D/E/F) have been launched.FY-2C was officially put into operation on 1 June 2005.It has a rich observation channel and can provide abundant information for the study of cloud parameters.The use of the FY-2C geostationary satellite allows for continuous observations over a fixed hemisphere.With a high temporal frequency, it can provide a set of observation data every half hour, which is very helpful for the study of the evolution of various weather processes and cloud systems.Satellite remote sensing provides an important means for cloud physics research (Mao, 2008).In Section 2.1 of this paper, we take advantage of the continuity of FY-2C geostationary satellite observations to analyze the weather background and cloud evolution.

The L-band sounding is a high-altitude meteorological detection system in China’s operational network.It can continuously and automatically measure air temperature,humidity, pressure, wind direction, and wind speed at high altitudes, and the measurement altitude can be up to 30 km.Zhou and Ou (2010) proposed a method for cloud vertical structure analysis using the relative humidity threshold of sounding data.In Section 2.2 of this paper,we use the relative humidity threshold method to analyze vertical structure of clouds on the day of the artificial operation.

The main method of this rainfall mitigation operation is rocket seeding with AgI catalyst.The rocket operation information includes the location of the operation point,the elevation angle, azimuth, amount of rockets used, operation time, and rocket model.The rocket model is WR-98 with a nucleation rate of 1.8 × 1015at -10°C.Precise spatial-temporal matching of the rocket operation information and radar data is crucial for the operation design and analysis of operation effects.

2.2 Platform and restricted operational zone

Cloud and Precipitation Accurate Analysis System(CPAS) is a technology platform independently designed and developed by Chinese scientists.The core of the platform is a real-time and fine-grained cloud physics processing and analysis system, and human decisionmaking, command and control, and performance analysis are the main functions of the platform.On the one hand,the CPAS system integrates fused data of multisource retrievals and observations from satellites, radars, soundings, and airplanes with multiple cloud model fusion analysis technologies to achieve real-time monitoring, recognizing, tracking analysis, operation designing, and warning of the macro-micro structure development of cloud and precipitation; at the same time, it integrates operation point information, operation elevation angle and azimuth, and the amount of catalyst used to accurately calculate the dispersion and transport range of each rocket catalyst and provide support for real-time analysis of the operation effects.

To meet the needs of the YOG artificial rainfall reduction, the CPAS has been transplanted to the operation command site, integrated with local observations and operation information, and formed the YOG version(CPAS-YOG), which is employed to support the design and real-time command of the YOG operation plan.This platform has a powerful capability for cloud forecast and analysis and it also functions well in processing and analyzing multisource information and real-time operation data.The real-time forecast and analysis of cloud macromicro structures and operation conditions as well as identification of supercooled liquid water based on satellite and radar observations are combined with calculations of the established operation condition identification model to enable real-time judgement of different cloud conditions for rainfall increase and decrease, eventually leading to reasonable, real-time instructions for artificial rainfall modification at different locations.At the same time,the CPAS-YOG has the capability to calculate catalyst diffusion and transport, and thus can identify the area influenced by the operation in real-time manner.As a result, precise spatial-temporal matching between radar echoes and the areas influenced by the operation can be achieved based on the spatial-temporal accuracy of the radar echo data (at 6-min intervals with horizontal resolution of 1 km).This provides favorable conditions for the recognition and analysis of the operation effect.In this paper, the CPAS-YOG system is used to discuss the operation effect and related operation technology through fine-grained analysis by integrating satellite and radar data with operation information after real-time operation design is completed and incident commands are issued.

Based on statistical characteristics of speed and direction of cloud and precipitation movement in Nanjing in August and considering the needs to protect the target area (the YOG venues), a pre-specified protection zone and the layout of operation points for weather modification activities are designed.This includes a venue protection zone with a radius of 25 km centered on the YOG Center, and three rocket operation protection zones with radii of 50 km (the first line of defense), 100 km (the second line of defense), and 150 km (the third line of defense), respectively.Within the second protection zone,ground-based rocket operation stations are mainly located in densely populated areas.

The principle and basic considerations of the design of the three defense zones are as follows.The area within the radius of 25 km is an urban area, in which operation is prohibited and the central venue is the target area where no rain is expected.The design of the three defense lines and zones is mainly based on the speed of the echo movement (water condensation) and the time period when the operation effect appears.According to the echo statistics in this area, the average speed of echo movement in this season is 50 km h-1.Relevant research results have shown that the effective period after the operation is 0.5-1 h, the most effective period is 1-2 h, and the period when the effect is not very obvious and eventually disappears is 2-4 h.Therefore, we have designed three operation zones corresponding to these time periods: 25-50-, 50-100-, and 100-200-km radius.Considering the characteristics of the air space and operation tools, the outer ring of the 100-150-km zone is designated as the aircraft operation zone, while the middle ring and the innermost ring of 25-50 km are designated as the rocket operation zone.Furthermore, the 25-50-km zone near the protection zone is designated as the reinforce operation zone, where the ground operation points are further densified to prepare for reinforce operation to handle newly generated echoes in adjacent areas.

Figure 1 shows the layout of the operation in Nanjing for the YOG displayed in the CPAS.This includes the venue protection area, the three designated operation circles (red circles), the rocket launch sites (denoted by black numbers), and their corresponding launch ranges(black circles).

According to the YOG organization committee, it is preferred that no precipitation occurs during the closing ceremony from 2000 to 2130 LT 28 August.The green box in Fig.1 represents the main operation area for the rockets on that day (detailed information about the target clouds on that day is provided in Table 1).Hereafter, the red circles in the radar and satellite images in this paper refer to the same operation zones and will not be clarified again.

3.Cloud and precipitation characteristics and the operation design

3.1 Synoptic weather and cloud evolution

At 0800 LT 28 August, the weather analysis showed that Jiangsu Province was located ahead of a high-altitude trough at 500 hPa, and a shear line at 850 hPa was located in the central area of Jiangsu Province.At 2000 LT, Jiangsu Province was located in the rear of the highaltitude trough at 500 hPa, and the 850 hPa shear line was located in the central to southern part of Jiangsu Province.Nanjing was located in an area of maximum westerly wind with the wind speed of 10 m s-1and was mainly affected by a weak shear associated with the low trough, which resulted in a large area of stratocumulus and stratus mixed cloud system and uneven surface precipitation.

Fig.1.Distribution of the protection zone for the Nanjing Youth Olympic Games venue as well as the three operation defense zones and ground operation points on CPAS-YOG.The red circles represent the defense lines (from inside out: the venue protection zone with a radius of 25 km;the first defense zone of 25-50 km; the second defense zone of 50-100 km; the third defense zone of 100-150 km).The black numbers in the figure indicate the designated operation points, and the corresponding black circles represent the range of coverage of rocket catalyst.

Fig.2.Infrared temperature images of FY-2 satellite from 1600 to 2100 LT 28 August.The red circles on the image indicate the layout of operation protection lines for the YOG.

The FY-2 satellite infrared brightness temperature for the period from 1600 to 2100 LT 28 August is shown in Fig.2.Corresponding to the high-altitude trough, there existed a northeast-southwest oriented cloud band, which was located to the west of the venue and continued to move from west to east.From the previous satellite image, it can be seen that convective clouds began to form in the south of the defense line outside the defense zone at 1300 LT, and rapidly intensified and expanded to the north after their formation.The satellite cloud image at 1600 LT shows that the convective clouds have reached the outermost defense line at this time.Due to the formation of the convective cloud clusters on the south side of the cloud band, the cloud band basically remained stable along the north-south direction.From 1600 to 2100 LT,it is found that these convective cloud clusters did not continue to move northward but slightly retreated to the south, causing no significant impact on the venue during the protection period.Therefore, this paper does not analyze them.The cloud that mainly affected the venue was the northeast-southwest oriented cloud band, which began to affect the venue area at 1700 LT and was the main precipitation cloud system affecting the venue.

3.2 Cloud vertical structure and evolution

Figure 3 shows the soundings from 2000 LT 27 to 0800 LT 29 August at Nanjing weather station.The cloud analysis method proposed by Zhou and Ou (2010)is employed to analyze the cloud system.The green shading represents the determined cloud region, and the solid green, blue, and red lines represent temperature,water vapor relative humidity (RH), and ice relative humidity profiles, respectively.The pink solid line represents the height of the 0°C layer.It can be seen that there was a thick cloud layer over Nanjing from 2000 LT 27 to 0800 LT 29 August.The sounding data from 0800 to 1400 LT on the closing ceremony day (28 August) indicate that the height of the 0°C layer was around 4800 m above the ground level (AGL), and a dry layer (with relatively low relative humidity) existed between the high and low clouds.From 0800 to 1400 LT, the thickness of the dry layer decreased slightly, ranging within 2500-5500 m AGL in height.The sounding cloud structure at 2000 LT indicates that the dry layer had disappeared,suggesting that the dry layer gradually weakened from 1400 to 2000 LT and eventually disappeared.It should be noted that according to meteorological records, continuous rain and fog persisted in Nanjing from 27 to 29 August, which was reflected on the sounding analysis as ground-based clouds.

Fig.3.Analysis of cloud structure and evolution at Nanjing sounding station from 27 to 28 August.

3.3 Evolution of cloud and precipitation echoes

Corresponding to the maintenance and movement of the northeast-southwest oriented cloud band shown on the satellite image that affected the venue as well as the low trough weather system, radar echoes constantly developed and moved, which could possibly cause precipitation at the venue.Based on the characteristics of echo evolution combined with operation information, the evolution process of the radar echoes is divided into three stages: the nascent stage (before 1804 LT), the development stage (1804-1823 LT), and the operation stage(1823-2111 LT).In order to present the echoes continuously, the radar echoes from 1804 to 2111 LT are marked consecutively from (1) to (30) in this paper.

Monitoring and analysis of the Hefei radar echoes at 6-min intervals on the command platform indicate that there were band-shaped echoes moving towards the venue direction before 1804 LT (see Fig.4).These band-shaped echoes corresponded to the northeast-southwest oriented cloud band, which was shown in satellite cloud image and analyzed in Section 2.1.Further calculations suggest that the cloud system corresponding to these echoes was approaching the venue at a nearly constant speed of 1 km min-1, and it might move to the venue during the period of 2000-2130 LT, which is the protection period, and could possibly affect the closing ceremony.

Figure 4 shows the reflectivity of Nanjing radar at 6-min intervals from 1804 to 1816 LT.During this period,the radar echoes moved steadily towards the venue with strong echo intensity, indicating that the venue was about to be affected.The horizontal structure of the echoes and the process of their eastward movement can be observed from Fig.4.To investigate the vertical structure of the precipitation echoes, vertical cross sections are produced along the moving direction across the venue (indicated by the red thin arrow in Fig.4) to demonstrate vertical structures of the radar echoes at different times and locations.The reflectivity before the operation is displayed in Fig.4 (1-3), which shows that a weak isolated echo (IA,inside the blue elliptical shape) was moving towards the venue from west to east at a speed of 1 km min-1.The red rectangular box in the cross section shows the corresponding structure of the echo IA, and the red pentagram indicates the position of the venue center.A stable echo band (IB, highlighted in this study and will be analyzed in more detail later) right behind the echo IA was also moving towards the venue from west to east with a speed of 1 km min-1.

In summary, the cloud system that could soon affect the venue was part of a stratocumulus and stratus mixed cloud, which was relatively stable and had a weakening dry layer below.The isolated weak echo IA on the west side of the venue and the stronger echo IB on the echo band were moving towards the venue from west to east at a speed of 1 km min-1.

3.4 Rainfall attenuation hypothesis and operation design

Fig.4.Evolution of the radar echo horizontal and vertical profiles every 6 min before the operation (1804-1816 LT).The red thin arrow indicates the location of the radar vertical profile passing through the venue along the direction of echo movement.The thick solid arrow in the profile denotes the location of the operation.The vertical structures of the radar echoes at different times and locations are shown on the right of each panel.The blue elliptical area represents the area of interest, and the red rectangular box in the plot corresponds to the profile structure.The red pentagon indicates the location of the venue center.

From the vertical structure of the radar, it can be seen that the 0°C layer was located at the height of about 4800 m AGL.Strong echoes generally occurred above the 0°C layer, and the strongest echoes appeared near or slightly below the 0°C layer, indicating that the particle concentration and mass of precipitation were larger in the upper layers, and the growth occurred above the 0°C layer and weakened below it.This suggests that the precipitation formation and growth mechanism corresponded to typical cold cloud process (initiated by the Bergeron process).If there were no dry layer in the lower layers, the precipitation would continue to grow.That is, the production and development of precipitation particles occurred through a cold cloud process rather than a warm rain process (in which the particle size increases as it moves to lower levels).Numerical simulation results also confirmed this analysis.The goal of this operation was to prevent as much as possible the rain from falling around the venues during the specific period (2000-2130 LT).Based on the forecast of the cloud model, the satellite and radar monitoring and analysis of cloud structure characteristics from soundings during the period 0800-1400 LT, a rainfall reduction plan using cold cloud seeding was proposed, and the operation design and expected effect are described as follows.

(1) Principle hypothesis for this operation: using a cold cloud catalyst to scatter a large amount of ice-forming nuclei in the supercooled zone of the mixed cloud, increasing the concentration of ice crystals.The large amount of artificial and natural ice crystals will compete for limited supercooled liquid water, which will lead to limited growth of large-size ice crystals and slower falling speed.This process is conducive to more evaporation in the air.When the ice crystals fall into the warm zone (lower level) and melt into small raindrops, they encounter a relatively dry layer in the low levels (dry layer with RH < 75), which is unfavorable for the growth of raindrops.Small raindrops fall slowly and stay in the air for a long time, resulting in large evaporation.In the dry layer, small-size raindrops further evaporate, and eventually evaporate completely before reaching the ground.As a result, the raindrop concentration decreases and ground rainfall decreases correspondingly.This is how the goal of reducing ground precipitation in the venue is achieved.

(2) Pre-design of the operation plan: based on existing data and observations, the ice crystals are estimated to grow to a size that could fall in approximately 30-60 min after catalysis, and the operation effect can last for about 3-4 h after 30 min of catalysis (Zhou and Zhu, 2014).The cloud echo velocity on the day of operation was 1 km min-1.Considering that no operations can be carried out within 25 km of the venue, the operation will be conducted at areas 25-80 km away from the venue.If a general WR-98 rocket (with a nucleation rate of 1.8 × 1015at-10°C) is used, the dosage should be at least 10 times the amount of rain enhancement for excessive catalytic operations.Of course, based on the air space, operation equipment, and cloud conditions, more rockets should be used.

(3) Pre-estimate of the operation effect: the mass of high-altitude ice crystals is supposed to increase, and the average diameter of the ice crystals will decrease.Since the radar echo is proportional to the sixth power of the ice crystal diameter, the radar echo intensity will decrease, although not significantly.In the low-altitude (especially dry) layer, however, the radar echo intensity will show a significant decrease and the decrease will be more significant after catalysis.The most significant decrease will occur at the bottom of the dry layer.For the commonly used composite reflectivity, the distribution of the strongest vertical echoes is often described.Note that the strongest vertical echoes in the case of the present study were concentrated near the 0°C layer (bright band),and they are supposed to be less affected by catalysis.Thereby it is hard to detect significant changes.The use of vertical profiles may reveal the weakening of the echo intensity in the low-altitude layer.A single vertical profile has spatial limitations, and the implementation of multiple layers of Constant Altitude Plan Position Indicator (CAPPI) is expected to show the catalytic effect on the intensity changes in three-dimensional space.

In summary, based on cloud forecast and radar analysis, it is determined that the cold cloud and precipitation mechanism is dominant for this case.Therefore, cold cloud seeding is appropriate for this operation.The dry layer in the low warm area of the cloud is conducive to evaporation.Cold cloud seeding can reduce the diameter of ice crystals, result in smaller cloud droplets and stronger evaporation, and reduce ground precipitation.The effect of seeding can be tested using multilayer CAPPI.Based on the above hypothesis of reducing rainfall, a rocket operation plan is further proposed.According to the movement path of the precipitation system(from west to east), artificial operations can be carried out in a stepwise manner from far to near.When the system moves within 60-90 km away from the key protection area, the second line of defense rocket operation can be initiated to carry out high-dose seeding.When the system moves within 15-60 km away from the key protection area, the first line of defense rocket operation can be initiated to achieve even larger doses of excessive seeding.After a large amount of continuous operations in time and space, it is expected that the radar echoes over the stadium will weaken (especially in the low altitude),and the corresponding ground rainfall will decrease.

4.Artificial rainfall reduction and effect analysis

4.1 Introduction to the operation

Following the overall plan of the operation, real-time precise command was achieved based on tracking and monitoring of the cloud system as well as certain principles.To achieve excessive catalysis, as many as possible sites were selected during the actual command and as many as possible emissions were made at each site,while the intervals between emissions were made as short as possible to achieve multiple rounds of high-dose and time-space continuous operations on the moving radar echoes.During actual operations, specific parameters for each operation point (time, elevation angle, azimuth, and amount of ammunition used) were calculated in real-time by the decision and command platform(CPAS-YOG) deployed on-site based on characteristics of the radar echoes, rocket trajectory curves, and the position of the operation point according to the existing models.The operating instructions (operating plan parameters) for each site were formed, and the operating parameters were updated for each operation point every 6 min based on the latest radar echoes.At the same time,specific air space is applied for the operation.Whether the operation could be implemented according to the plan instructions at relevant points was also dependent on whether the air space could be approved for the operation.Therefore, the actual situation of the operation information and the best instructions were not always consistent.

The actual process of this operation was divided into two major stages.During both stages, high-density seeding of AgI was conducted at 5500-6500 m (temperature range from -8 to -10°C) for the target IA and IB cloud clusters, which might occur and affect the protected area.Details of the operation situations are given in Table 1.Considering dispersion and other parameters based on the launch angle, the operations recorded in this table are found to be effective operations that were applied to the target IA and IB clouds.Other irrelevant operations are not included.

As shown in Table 1, detailed operation situations for the two target clouds are as follows.

(1) From 1800 to 1920 LT, two suites of operation were carried out on the target cloud IA.As shown in Fig.4 (1-3) and Fig.5 (4-13), the echo intensity of IA was weak.However, an isolated strong echo cloud located to the west side of venue and the in front of the cloud band had moved across the second line of defense by 1758 LT,and its echo intensity kept growing.According to the operation instructions and the air space approved for operation, two suites of operation were carried out from 1820 to 1825 LT (8 rounds) and from 1830 to 1835 LT (7 rounds), respectively.

Table 1.The operation information during the closing ceremony

(2) From 1920 to 2130 LT, multiple rounds of highdose catalytic operation were carried out on the target cloud IB, which was located right on the echo band and had a strong and stable echo intensity.As displayed in Fig.5 (5-8), accompanied with the low-pressure trough,a deep precipitation cloud system was moving and developing, which was clearly shown in the radar echo image of IB at 1919 LT.This precipitation cloud system had moved across the second line of defense to the west of the venue with a very strong echo intensity.Following the principle mentioned above, according to the direction and intensity of the echo movement, multiple rounds of high-dose catalytic operations were carried out at multiple points from 1920 to 2050 LT (as shown in Table 1),aiming to achieve an overall excessive catalytic effect and weaken the precipitation cloud system.

(3) Due to the weak strength of the IA, it was treated with two suites of continuous operations, using a total of 15 rockets.Based on the strength of the IA radar echo and its movement, it was determined that the target area was no longer affected by IA, and thus the operation was stopped.However, the IB had a strong radar echo and was stable in the echo belt, which tended to move eastward and possibly affect the venue.The radar echo was strong and covered a long period and a large area,thereby a long-duration, multi-operation-point, and continuous catalysis was employed to achieve excessive catalysis in a certain time and space.

4.2 Vertical structure of radar echoes before and after the operation

Based on the analysis of the catalytic principle, there would have significant changes in the low-level radar echoes after the operation due to the presence of the dry layer (evaporation would occur during the falling process of the upper layer particles), while the changes in the high-level radar echoes near and above the catalytic height would be relatively small compared to the changes in lower levels.It was found that the radar composite reflectivity could not well reflect the operation effect.Therefore, the overall changes in radar echoes at different positions and times before and after the operations were analyzed using composite reflectivity and profile structure, and the vertical structure changes of the radar echoes before and after operations on IA and IB were used to intuitively analyze the operation effect.Furthermore, changes in relevant parameters before and after the operations were explored based on radar reflectivity at different heights.

4.2.1Vertical structure change after the operation on echo IA

Figure 5 shows the evolution process (1823-1919 LT)of the radar echo of IA in terms of its horizontal distribution (blue elliptical shape) and vertical (red rectangular shape) cross section.The detailed operation information on IA is marked in the figure, with the solid red circles indicating the locations of the operation points and the black numbers beside them indicating the corresponding operation amounts.As shown in Fig.4, IA crossed the second line of defense from the west at 1804 LT, and the echo intensity gradually increased from 1804 to 1823 LT.As shown in the vertical cross section, the echoes in the upper and lower levels grew significantly, and two rounds of rocket-assisted seeding were carried out during 1820-1825 LT (8 rockets) and during 1830-1835 LT(7 rockets).The echo intensity of IA began to weaken at 1829 LT and decreased significantly at around 2000-4000 m AGL, which was consistent with the prediction that the low-level radar echoes would undergo significant changes after the seeding due to the existence of the dry layer in the mid- to lower levels.At 1913 LT, the seeded echoes began to move towards the venue, and an area with echoes of almost minimal intensity, referred to as an “echo hole,” appeared in the venue center.The appearance of this echo structure was attributed to the generation of a large number of artificial ice crystals in the cloud after the seeding, which increased the concentration of ice particles and decreases their diameters due to the “consumption” of supercooled liquid water.As the radar reflectivity is proportional to the sixth power of the particle diameter, the radar reflectivity decreased; and considering that the particle falling speed is proportional to its diameter, the falling speed also decreased, leading to increased evaporation and ultimately weakening the rainfall.This echo hole proved the effectiveness of the seeding.By the time IA entered the venue at 1931 LT,the echoes basically had dissipated.

Fig.5.Evolution of radar echoes of IA in both horizontal distribution and vertical cross section (1823-1919 LT).The red thin arrows indicate the position of the vertical cross section of the radar echoes passing through the venue in the direction of movement.The thick solid arrow in the profile represents the location of the operation.The vertical cross sections of different radar echoes at different times and locations are shown on the right side of each panel.The black ellipse represents IA, and the black rectangular box in the cross section represents the corresponding profile structure of IA.The red ellipse represents IB, and the red rectangular box in each vertical cross section represents the corresponding profile structure of IB.The red pentagon indicates the position of the center of the venue.

It should be noted that in addition to the “echo hole”position indicated by the red ellipse, there were also two weak echo areas to the left of the “echo hole” position in Fig.5.From the continuous echo observations shown in Fig.5, it can be seen that these two weak echo areas had been present throughout the entire process, and they were not significantly related to the operation.The “echo hole”within the red ellipse was the remnant of the echo that was strong at the beginning and gradually weakened after the operation.Apparently, it was the result of the operation (the same phenomenon can also be found in the analysis of IB, which will not be explained here).

Fig.6.(a) The radar profiles of operation echo IA and (b) the difference between reflectivity (per 1000-m altitude) after operation and before operation.The black dashed lines in the figure indicate the times when operations were conducted, and the corresponding black numbers indicate the number of operations; the red dashed box is the field protection area and (a) the blue dotted line in the figure is the height of 0°C layer determined from sounding (about 4732 m); (b) the solid lines of different colors in the figure represent the difference between the reflectivity after IB operation and before operation at different heights, respectively (blue: 2000 m, orange: 3000 m, green: 4000 m, red: 5000 m, purple: 6000 m).

Figure 6 shows the vertical evolution of the radar reflectivity of IA before and after the operation (Fig.6a)and the difference between reflectivity before and after the operation at every 1000-m height (Fig.6b).The calculation method for Fig.6a is the average of the Plan Position Indicator (PPI) echo area (54 km2) at different heights of IA.For Fig.6b, the reflectivity intensity at the start time of the operation (1820 LT) was considered to be 0, and the reflectivity at other times was obtained based on the reflectivity intensity at the corresponding times.The black dotted line in the figure represents the time of the operation, and the corresponding numbers indicate the amount of operations at that time.The red dashed line box indicates the time period when IA moved into the venue protection area.From the vertical profile of IA, it can be seen that the echo intensity gradually weakened from 1823 to 1925 LT after the operation, and the weakening trend spread from the upper to the lower layers.After reaching the minimum value at 1925 LT, it started to increase again.In the altitude range of 6000-8000 m, the echo intensity first decreased from 1823 to 1835 LT, then started to gradually increase after 1835 LT, and decreased again after 1900 LT, reaching a minimum value near the venue before gradually increasing again.The difference in reflectivity at every 1000-m height interval also shows that the reflectivity at 6000 m first decreased and then increased, and it remained a relatively high value.The reflectivity at 2000- and 3000-m heights decreased significantly after the operation, and reached the maximum decrease at 1925 LT before it started to decrease again.

4.2.2Vertical structure change after the operation on echo IB

From the horizontal and vertical structure diagrams of radar echoes (from 1925 to 2111 LT) shown in Fig.7, it can be seen that IB was the main echo band with strong echo intensity that affected the venue.Multiple rounds of multisite continuation operation were conducted on it from 1938 to 2046 LT.It can be seen that the echo intensity of IB was very strong from 1925 to 1956 LT.After the operation, the echo started to become weak since 1956 LT.The weakening was particularly evident at altitudes of 4000-2000 m.During the movement of the system, a weak echo zone gradually formed in front of the main echo band, and when the echo being operated at 2046 LT moved into the venue area, a clear “echo hole”similar to that of IA appeared above the venue (indicated by the red elliptical box in the figure), which lasted for nearly 50 min.The mechanism for its formation is the same as that for IA, so it will not be described again here.This “echo hole” ensured that the vertical precipitation echo near the venue did not pass through, and no significant precipitation occurred in the venue.

Figure 8 shows the vertical evolution of radar echoes of IB before and after the operation in terms of average reflectivity (Fig.8a) and the difference in reflectivity at every 1000-m altitude before and after the operation (Fig.8b).The calculation method of Fig.8a is the average area of PPI echoes at different heights (167 km2) of IB.The calculation method of Fig.8b is as follows.The reflection intensity at the start of the operation (at 1938 LT)was regarded as 0, and the reflection intensity at other times was obtained based on the reflection intensity at the corresponding times, which could clearly reflect relative increases and decreases of the reflectivity before and after the operation.The black dotted lines in Fig.8 represent the times when operations were conducted, and the corresponding numbers on the dotted lines represent the amount of operation at the corresponding times.The red dotted box indicates the time when IB moved into the stadium safeguard zone.

From the vertical profile shown in Fig.8a, it can be seen that after the first operation at 1938 LT, there was a phenomenon of echo weakening and then strengthening.From 1950 to 2034 LT, multiple rounds of rocket were conducted on IB.After the relatively high dose operation at 1950 LT (16 rockets) and 1956 LT (28 rockets),the echo gradually weakened.At 2059 LT, the echo reached its minimum and an “echo hole” appeared,which demonstrated the significant decrease of rainfall echo near the stadium and ensured there was no obvious rainfall in the stadium.The significant weakening of the echo after 2059 LT was the main concern of this study.The weakening was most significant in levels with temperature above 0°C, which was attributed to the presence of a dry layer in the lower level that caused the falling particles to evaporate.The echoes began to gradually increase after 2059 LT and moved out of the stadium.The difference in reflectivity at different altitudes shown in Fig.8b also indicated that the radar reflectivity at different altitudes decreased first before it increased, and highaltitude echoes weakened earlier than those in low-altitude.The echoes at 5000-6000-m altitude weakened the fastest but to a smaller extent, and the echoes at 2000-m altitude weakened the slowest but to the greatest extent,reaching a maximum weakening difference of about -15 dBZ at 2059 LT, and then the degree of weakening began to decrease.

Based on the analysis of the vertical structure evolution of the two radar echoes, regardless of IA or the stronger IB, a significant weakening of the echoes could be observed 30-45 min after each excessive catalytic operation, with significant decreases occurring in the 4000-2000-m altitude, and the most obvious decrease occurred at 2000 m.A distinct “echo hole” appeared in the vertical structure, which was caused by the increase in particle concentration after the operation.More ice crystals “consumed” supercooled liquid water, resulting in a decrease in particle diameter.As the radar reflectivity is proportional to the sixth power of the particle diameter, the radar reflectivity decreases.Since the falling speed of a particle is proportional to its diameter, decreases in falling speed resulted in enhanced evaporation,ultimately causing decreases in both lower-level echo and rainfall intensity.During this process, the decrease in particle diameter could inevitably lead to a decrease in radar reflectivity, but the rainfall intensity may not be significantly affected.This “echo hole” proved the effectiveness of this operation.

Fig.7.Evolution of the horizontal and vertical profile structures of radar echoes IB (1925-2111 LT).The red thin arrow in the figure shows the location of the radar vertical profile according to the moving direction via the venue, the thick solid arrow in the profile is the operation location,and the subplot on the right side of each panel shows the vertical structure of the radar echoes corresponding to different locations at different times.The black oval denotes IB, the black box in the figure is the profile structure corresponding to IB, and the red pentagram position is the position corresponding to the center of the venue.

The difference is that the reflectivity of IA decreased rapidly from upper to lower levels after the operation,while the reflectivity of IB first decreased and then increased.Only after multiple high-dose operations, did the phenomenon of the echo weakening and the appearance of an “echo hole” similar to that for IA to occur.There are two main reasons for this difference.First, IA was an isolated small echo, while IB was located on the echo band and required continuous operations to achieve the desired effect.Second, IB was located behind IA and was more moist, and the dry layer could promote more evaporation, which is a favorable factor for reducing rainfall.

Fig.8.(a) The radar profile of operation echo IB and (b) the difference between reflectivity (per 1000-m altitude) after operation and before operation.The black dashed lines in the figure indicate the times when operation were conducted, and the corresponding black numbers indicate the numbers of operation; the red dashed box is the field protection area and (a) the blue dotted line in the figure is the height of 0°C layer determined from sounding (about 4732 m); (b) the solid lines of different colors in the figure represent the differences in reflectivity after IB operation and before operation at different heights, respectively (blue: 2000 m, orange: 3000 m, green: 4000 m, red: 5000 m, purple: 6000 m).

4.2.3Comparison of echo IB and other echoes

As can be seen from the analysis above, the impacts of catalytic changes on echoes at different heights were not completely identical.The echo attenuation was significant at heights of 2000-4000 m, while the greatest attenuation occurred at the height of 2000 m and the degree of changes was relatively weak at 5000-6000 m.In order to further analyze the effectiveness of the operation and highlight differences between operational and non-operational echoes, the echoes at the height of 2000 m are selected for comparative analysis between operational IB and non-operational echoes (IIC) as well as between operational IB and operational echoes (IIID, IIIE, and IIIF).

Following the considerations and selection principles mentioned above, the non-operational echoes (IIC, green box, shown in Fig.7) at the altitude of 2000 m are selected.Note that we employ a relatively strict approach to select IIC.First, the echoes must be in the same cloud system as IB; second, the evolution characteristics of the echoes must be basically consistent with that of IB before the operation; third, the same echo area of 167 km2is selected.At the same time, in order to ensure the echo attenuation effect over the venue during the actual operation, quite a few operations (more than 4 rockets) were also carried out on the echoes near IB (blue box), even though these echoes were not the main target operation echoes.Hereafter these echoes are referred to as IIID(purple box), IIIE (orange box), and IIIF (pink box).It should be noted that although these three echoes were in the same cloud system as IB from the initial analysis,they were all located in the strong echo center of the mixed cloud.Therefore, their properties were not very similar to that of IB.Due to the influence of established facts (the operation had been completed and the operation area was determined), echoes that were very similar to IB are not selected.This study also analyzes such nontarget operation echoes and a brief comparative introduction is provided.

The average values of the echoes for IB, IIC, IIID,IIIE, and IIIF shown in Fig.9 are statistically analyzed,and the results are presented in Fig.10.It can be seen from Fig.10 that all echoes exhibited a weakening trend from 1906 to 2034 LT.All the echoes of IB and IIC showed similar intensity and trend of changes before the operation, making them good objects for comparison.However, the echoes of IB, IIID, IIIE, and IIIF were stronger and showed different trends of changes compared to that of IB and IIC.Looking at the red dashed box in Fig.10, the effect of the operation on the echoes is apparent.A significant period of weakening by about 10 dBZ occurred from 2046 to 2111 LT after the operation,followed by a significant increase in the echoes.The IIC echoes (non-operational echoes) steadily weakened and eventually disappeared.Looking at the change processes of other echoes of IIID, IIIE, and IIIF with a small amount of operation, it is found that although there was a small amount of operation, the echoes all exhibited a relatively steady weakening trend after the operation.

5.Variation of radar echoes of different intensities after operation

Fig.9.Temporal evolutions of IB reflectivity and non-operation (IIC)echoes as well as non-target operation echoes (IIID, IIIE, and IIIF) at 2000-m altitude.

In order to further analyze the relationship between operation techniques and operation effects, the most obvious changes that occurred at the altitude of 2000 m after the operation are selected based on the above analysis.The changes in echo intensity before and after the operation are statistically analyzed for the two typical areas of target echoes IA (54 km2) and IB (167 km2).At the same time, the operation effect is further analyzed by combining the operation time, distance, and dosage.The relevant operational technical parameters are summarized and the key techniques for artificial rainfall reduction in specific target areas are refined.

5.1 Technical requirements for operation on the weak isolated echo IA

Fig.10.Comparison of area (167 km2) mean values of the IB reflectivity and the non-operation echoes (IIC) as well as non-target operation echoes (IIID, IIIE, and IIIF) at 2000-m altitude over time.

Figure 11 shows temporal evolution of the average reflectivity and the number of points with a large CAPPI echo greater than 10 dBZ at 2000-m height for IA from 1649 to 2136 LT.The red dots indicate the times of the operation, and the corresponding numbers represent the total amount of operations at the corresponding times.Overall, IA began to weaken after the operation at 1829 LT, and the echo intensity recovered to the level similar to that before the operation after 2015 LT.The echo intensity experienced a process of weakening and then strengthening during this period (shown in the red box in Fig.11a).The black box indicates the time period when IA was in the venue, which was also the period when the echo weakened most and the echo intensity was lower than that before the operation.The detailed process is as follows.From 1810 to 1823 LT (i.e., before the operation), the echo was in the development and enhancement stage.After two rounds of operation (a total of 15 rockets) during 1823-1829 LT, the radar echo gradually weakened starting from 1835 LT.IA entered the protection zone at 1900 LT, and the echo reached its lowest value at 1925 LT.After 1925 LT, the echo intensity began to increase.At 1931 LT, the echo moved to the venue (the green shaded area in the figure) and generally remained weak.It moved out of the protection zone at 1950 LT.After 2015 LT, the reflectivity of the echo recovered to the level similar to that before the operation.

To summarize the technical parameters for operation on IA, the start time of the statistical parameters is taken as the midpoint of the operational period for the convenience of statistics, that is, the average time of 1823-1829 LT is 1826 LT.The statistical results are as follows.

(1) The period of weakening after operation is 1826-1925 LT, namely 59 min;

(2) The time period for the echo intensity to recover to the level similar to that before the operation is 1826-2015 LT, approximately 108 min;

(3) The period during which the operation effect on the echo weakening is significant is 1906-1938 LT,namely 32 min;

(4) The time period during which a significant change occurred after operation is 1826-1906 LT, namely 40 min;

(5) The echo weakening rate of IA is approximately 0.20 dBZ min-1, calculated by (12.34 dBZ - 0.61 dBZ)/59 min, where 12.34 is the average reflectivity of IA at 1834 and 1829 LT, and 0.61 is the reflectivity value at 1925 LT when the reflectivity of IA is minimized;

(6) The optimal starting distance and time for operation: starting from 1826 LT and operating until the echo reaches its weakest point at 1925 LT, which is about 59 min.The movement speed of the echo is 1 km min-1.Therefore, it can be calculated that the optimal operation distance is 60 km away from the venue protection zone,and the optimal operation time is 60 min before the protection period.

5.2 Technical requirements for operation on the strong echo IB

Fig.11.Area (54 km2) average of radar reflectivity of the operation echo IA at 2000-m altitude and the number of grid cells with strong echoes(> 10 dBZ) in this area during the period 1649-2136 LT.The red dots show the operation moments, the corresponding numbers are the operation numbers at the corresponding moments, and the area shaded in green shows the time when the operation echo moved to the stadium.

Figure 12 shows the time series of the average CAPPI radar reflectivity and strong echoes (> 15 dBZ) at the height of 2000 m for IB from 1823 to 2341 LT.The red dots indicate the times of operation, and the corresponding numbers represent the total amount of operations at corresponding times.Overall, after multiple rounds of high-dose operation from 1938 to 2034 LT, IB began to weaken, and the echo intensity returned to the level before the operation at 2216 LT.During this period, there was a process of echo weakening and then strengthening(shown in the red box in Fig.12a).The black box indicates the time period when IB was located in the protection area, which is also the period when the most significant weakening of echoes occurred and the echo intensity was much lower than that before the operation.

The detailed process is as follows.From 1841 to 1931 LT before the operation, IB showed a high reflectivity and a slow decreasing trend.From 1931 to 1938 LT, the reflectivity of IB slightly increased.Starting from 1938 LT, multiple rounds of high-dose operations were carried out on IB.After the first two operations (8 rockets and 16 rockets), the echo reflectivity fluctuated, but there was no obvious weakening.After the third operation with 28 rocket shootings at 1956 LT, the echo began to show a significant weakening trend, but it slightly increased after about 12 min.After the fourth (8 rockets),fifth (8 rockets), and sixth (24 rockets) operations were continuously conducted, the echo began to weaken significantly, especially after the sixth operation of 24 rockets at 2021 LT.After the final operation of 8 rockets at 2034 LT, the echo began to weaken rapidly.At 2136 LT, the echo reflectivity returned to the level similar to that after the last operation.When the echo moved to the venue(green shaded area in the figure) at 2046 LT, the echo was basically in a relatively weak state and the minimum value of the radar reflectivity was reached at 2059 LT.At 2118 LT, it moved out of the venue, and the reflectivity returned to the level before the operation at 2136 LT.

To summarize the operation technical parameters of IB, the start time of the statistical parameters is taken as the midpoint of all operation periods for the convenience of statistics, that is, the average time of 1938-2034 LT is 2006 LT.The statistical results are as follows.

(1) The period of weakening after operation is 2006-2059 LT, namely 53 min;

(2) The time period for the echo intensity to recover to the level similar to that before the operation is 2006-2136 LT, approximately 90 min;

(3) The period during which the operation effect on the echo weakening is significant is 2046-2111 LT,namely 25 min;

(4) The time period during which a significant change occurred after operation is 2006-2046 LT, namely 40 min;

(5) The attenuation rate of the IB echo is (19.94 dBZ -5.54 dBZ)/53 min ≈ 0.27 dBZ min-1.Here, 19.94 is the average reflectivity of IB from 1938 to 2034 LT, and 5.54 is the reflectivity value when IB is attenuated to the lowest point at 2059 LT;

(6) The optimal distance and time for starting catalysis: starting from 2006 LT for operation until the echo reached its weakest point at 2059 LT.With the speed of 1 km min-1for the echo movement, the optimal operation distance is 53 km away from the venue (the best operation time is 53 min ahead of the operation).

Fig.12.Area (167 km2) average of radar reflectivity of the operation IB at 2000-m altitude and the number of grid cells with strong echoes (>15 dBZ) during the period of 1823-2341 LT.

By comparing and analyzing the operational parameters of IA and IB, it is found that IA and IB had a similar weakening period after the operation, which lasted for more than 50 min.The time required for IB to return to its echo intensity level prior to the operation was slightly smaller than that of IA.The durations of obvious operation effect period were 25 min for IA and 32 min for IB.It is worth noting that the time required for significant changes to occur after operation in both IA and IB is 40 min, which is consistent with the findings of Zeng et al.(1997) in their experiments on the macroscopic structural changes of clouds caused by operation in Gutian,Fujian.They found that significant changes in echo parameters occurred about 30-40 min after operation.Results of the present study are consistent with Zeng et al.(1997) because the echo intensity of IB is strong and the cloud is thick, and thus spatially and temporarily continuous and excessive suppression is required.More reliable statistical results regarding the specific rules and parameters will require further analysis of individual cases.

As can be seen, in the actual tracking and commanding of operations over specific period of time and area, it is necessary to comprehensively consider the characteristic conditions of the operation cloud and the characteristics of the direction and speed of the echo movement.Understanding these features are a prerequisite to determine the best advance time and distance for the operation and provide optimal real-time operation plans.From the perspective of operation dosage, IA was related to a weak isolated echo located ahead of the echo band, and 15 rockets had a significant operation effect on it.In contrast, IB was related to a stronger systemic echo on the echo band, and continuous high dosage operation (156 rockets) was required to weaken the echo of IB.In terms of the actual effect of the operation, the time, location,and dosage of the operation around the specific target area of venue protection in this operation were relatively appropriate.It can be seen that the pre-proposed operation plan and real-time command of the artificial rainfall reduction process for this specific target are reasonable and effective, and the goal of reducing rainfall and ensuring little or no rain in the specific area (venue) during a specific period of time (2000-2130 LT) has been achieved.

6.Conclusions

The artificial rainfall reduction experiment investigated in the present study is a scientifically designed physical experiment over a fixed area and period (Nanjing Olympic Sports Center, closing ceremony period 2000-2130 LT) with a clear goal to modify weather in the field.It is a complete artificial rainfall reduction operation, including the proposal of the technical principle, the design of the technical plan, the estimation of the effect,and the verification of the effect against observations.Results of the analysis in the present study confirm the correctness of the technical theory and the expected effect.Based on the operation effect, relevant technical parameters of artificial rainfall reduction for specific targets are summarized and refined.Some insights and results obtained from the present study are as follows.

(1) On 28 August 2014, Nanjing was affected by a weak shear line caused by a low-pressure trough, which resulted in a large area of mixed stratocumulus and stratus clouds and uneven precipitation on the ground.The mixed clouds in the venue area were relatively stable with weak dry layers (layers with low relative humidity)present at the lower part of the clouds.There were weak isolated echoes (IA) to the west of the venue and strong echoes (IB) on a stable echo band that was moving from west to east at the speed of 1 km min-1, which was expected to affect the venue.Based on analysis of cloud structure from soundings and precipitation structure from radar observations, a technical solution of “excessive catalysis” was proposed for the purpose to artificially reduce the rainfall.

(2) To reduce the impact of different intensity echoes IA and IB on the venue during various time periods, excessive catalysis was implemented.IA was an isolated echo located in front of the main echo band, and it was subjected to two rounds of rocket operations at the height of 5500-6000 m AGL.IB was the main echo that could affect the protection zone of the venue.It was located on the echo band of the low-pressure trough cloud system,and had a long duration and high intensity.Multiple stations were used for continuous operations to reduce its impact on the venue.The operations were conducted at the height of 5500-6000 m AGL.

(3) After the catalysis of IA and IB, the reflectivity decreased at each individual level of height, and the most obvious attenuation of echo occurred at the height of 2000 m.Both IA and IB showed an “echo hole” when they moved to the venue.The reason for the appearance of this echo structure is that after the catalysis operation,more ice crystals competed for supercooled liquid water,which reduced the particle diameter.Since the radar reflectivity is proportional to the sixth power of the particle diameter, the radar reflectivity decreased.The particle falling speed also decreased because the falling speed is proportional to the particle diameter.Additionally, the dry layer in the lower level led to enhanced evaporation,ultimately resulting in a decrease in rainfall intensity.During this process, the decrease in rainfall particle diameter could certainly lead to a decrease in radar reflectivity, but the changes in rainfall intensity might not be very obvious.The appearance of this “echo hole” above the venue confirmed the effectiveness of this operation.

(4) IIC, IIID, IIIE, and IIIF were in the same cloud system as IB.Changes in the average reflectivity over the regions (167 km2) of IIC, IIID, IIIE, and IIIF, where the pre-operation echo evolution characteristics were similar to that of IB, were compared with the changes in IB.It is found that IB experienced a significant weakening in the period after the operation with a decrease of about 10 dBZ.After the echo recovered and increased, IB then experienced a significant increase.Meanwhile, the unoperated echo had been steadily weakening and eventually disappeared.The echoes that were subjected to smallamount operations (but the operation amount was still higher than that for rainfall enhancement operation)showed a weakening trend after the operation.The most significant difference between IB and the echoes subjected to smaller-amount operations was that the reflectivity of IB greatly increased after a rapid and pronounced decrease.

(5) The durations of the weakening phase after the IA and IB operations are similar, both lasting for over 50 min.The period of obvious catalytic effect for IA is 32 min, and for IB it is 25 min.The time it took for the IB echo intensity to recover to the pre-operation level is slightly earlier than that for IA due to the larger echo intensity and deeper cloud related to IB.The attenuation rates of the two echoes were 0.21 and 0.27 dBZ min-1,respectively, which are slightly different with different operation amounts.It is worth noting that the time required for obvious changes to occur after the catalysis of IA and IB both is 40 min.In terms of the operation dose,IA received two consecutive operations with a total of 15 rockets, and the catalytic effect was significant.With a strong systemic echo on the echo band, IB required continuous high-dose operations (156 rockets) to weaken the echo.

(6) Key technologies for commanding artificial operations with clear goals (such as rainfall reduction) over fixed region and specific time periods (such as the closing ceremony of the Olympic Center) require comprehensive consideration of cloud characteristics as well as analysis of the movement direction and speed of the echo.Based on this, the optimal time, distance, and dose for implementation of the operation can be determined and the best operation plan can be provided in real-time.Looking at the actual effect of the operation, it is obvious that the time, location, and dosage of the operation in the specific target area were all suitable.

(7) This is a rare practice of artificial rainfall reduction.The current research in this paper is mainly based on analysis of satellite data and soundings as well as radar observations.In the future, further in-depth studies will be conducted from different perspectives by combining numerical simulations and other sources of observations.

Acknowledgments.We would like to thank Jiangsu Provincial Meteorological Bureau (JPMB) for providing the rocket operation data.Thanks also go to JPMB and Nanjing Meteorological Bureau for organizing the targetspecific artificial operations.