• Ao, J., and J. Q. Sun, 2014: Difference of water vapour transportation in the interannual and decadal variations of winter precipitation over western China (in Chinese). Climatic Environ. Res., 19, 497506.

    • Search Google Scholar
    • Export Citation
  • Chen, B., and X. D. Xu, 2016: Spatiotemporal structure of the moisture sources feeding heavy precipitation events over the Sichuan Basin. Int. J. Climatol., 36, 34463457, https://doi.org/10.1002/joc.4567.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chen, B., X. D. Xu, and X. H. Shi, 2011: Estimating the water vapor transport pathways and associated sources of water vapor for the extreme rainfall event over east of China in July 2007 using the Lagrangian method (in Chinese). Acta Meteor. Sin., 69, 810818.

    • Search Google Scholar
    • Export Citation
  • Chow, V. T., D. R. Maidment, and L. W. Mays, 1988: Applied Hydrology. McGraw-Hill, 572 pp.

  • Ciric, D., R. Nieto, A. Ramos, A. Drumond, and L. Gimen, 2017: Wet spells and associated moisture sources anomalies across Danube River basin. Water, 9, 615, https://doi.org/10.3390/w9080615.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Cui, C. X., 2013: Study on water vapor characteristic of typical heavy snowstorm case in Northern Xinjiang (in Chinese). Plateau Meteor., 32, 11151125.

    • Search Google Scholar
    • Export Citation
  • Dai, X. G., W. J. Li, Z. G. Ma, and P. Wang, 2007: Water-vapor source shift of Xinjiang region during the recent twenty years. Prog. Nat. Sci., 17, 569575, https://doi.org/10.1080/10020070708541037.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Drumond, A., R. Nieto, and L. Gimeno, 2008: A Langangian identification of major source of moisture over Central Brazil and La Plata Basin. J. Geophys. Res., 113, D14128, https://doi.org/10.1029/2007JD009547.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Drumond, A., R. Nieto, and L. Gimeno, 2011: Sources of moisture for China and their variations during drier and wetter conditions in 2000–2004: A Lagrangian approach. Climate Res., 50, 215225, https://doi.org/10.3354/cr01043.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gimeno, L., A. Drumond, R. Nieto, R. M. Trigo, and A. Stohl, 2010: On the origin of continental precipitation. Geophys. Res. Lett., 37, L13804, https://doi.org/10.1029/2010GL043712.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gustafsson, M., D. Rayner, and D. L. Chen, 2010: Extreme rainfall event in southern Sweden: Where does the moisture come from? Tellus, 62A, 605616, https://doi.org/10.1111/j.1600-0870.2010.00456.x.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Huang, Y. J., and X. P. Cui, 2015: Moisture sources of torrential rainfall events in the Sichuan basin of China during summers of 2009–13. J. Hydrometeor., 16, 19061917, https://doi.org/10.1175/JHM-D-14-0220.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Huang, R. H., J. L. Chen, and Y. Liu, 2010: Interdecadal variation of the leading modes of summertime precipitation anomalies over eastern China and its association with water vapor transport over East Asia (in Chinese). Chin. J. Atmos. Sci., 35, 589606.

    • Search Google Scholar
    • Export Citation
  • James, P., A. Stohlet, N. Spichtinger, and S. Eckhardt, 2004: Climatological aspects of the extreme European rainfall of August 2002 and a trajectory method for estimating the associated evaporative source regions. Nat. Hazards Earth Syst. Sci., 4, 733746, https://doi.org/10.5194/nhess-4-733-2004.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Liu, J. R., X. F. Song, G. F. Yuan, X. M. Sun, X. Liu, F. Chen, Z. M. Wang, and S. Q. Wang, 2008: Characteristics of δ18O in precipitation over Northwest China and its water vapor sources (in Chinese). Acta Geogr. Sin., 63, 1222.

    • Search Google Scholar
    • Export Citation
  • Ma, S. H., and Y. W. Xi, 1997: Some regularities of storm rain fall in Xinjiang, China (in Chinese). Acta Meteor. Sin., 55, 239248.

  • Newell, R. E., N. E. Newell, Y. Zhu, and C. Scott, 1992: Tropospheric rivers? A pilot study. Geophys. Res. Lett., 19, 24012404, https://doi.org/10.1029/92GL02916.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Nieto, R., and L. Gimeno, 2019: A database of optimal integration time for Lagrangian studies of atmospheric moisture sources and sinks. Sci. Data, 6, 59, https://doi.org/10.1038/s41597-019-0068-8.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Pan, Y., Y. Shen, J. J. Yu, and P. Zhao, 2012: Analysis of the combined gauge-satellite hourly precipitation over China based on the OI technique (in Chinese). Acta Meteor. Sin., 70, 13811389.

    • Search Google Scholar
    • Export Citation
  • Shen, Y., Y. Pan, J. J. Yu, P. Zhao, and Z. J. Zhou, 2013: Quality assessment of hourly merged precipitation product over China (in Chinese). Trans. Atmos. Sci., 36, 3746.

    • Search Google Scholar
    • Export Citation
  • Shi, Y. G., and Z. B. Sun, 2008: Climate characteristic of water vapor transportation and its variation over Xinjiang (in Chinese). Plateau Meteor., 27, 310319.

    • Search Google Scholar
    • Export Citation
  • Sodemann, H., C. Schwierz, and H. Wernli, 2008: Interannual variability of Greenland winter precipitation sources: Lagrangian moisture diagnostic and North Atlantic Oscillation influence. J. Geophys. Res., 113, D03107, https://doi.org/10.1029/2007JD008503.

    • Search Google Scholar
    • Export Citation
  • Stohl, A., and P. James, 2004: A Lagrangian analysis of the atmospheric branch of the global water cycle. Part I: Method description, validation, and demonstration for the August 2002 flooding in central Europe. J. Hydrometeor., 5, 656678, https://doi.org/10.1175/1525-7541(2004)005<0656:ALAOTA>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Stohl, A., and P. James, 2005: A Lagrangian analysis of the atmospheric branch of the global water cycle. Part II: Moisture transports between Earth’s ocean basins and river catchments. J. Hydrometeor., 6, 961984, https://doi.org/10.1175/JHM470.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Stohl, A., C. Forster, A. Frank, P. Seibert, and G. Wotawa, 2005: Technical note: The Lagrangian particle dispersion model FLEXPART version 6.2. Atmos. Chem. Phys., 5, 24612474, https://doi.org/10.5194/acp-5-2461-2005.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Stohl, A., F. Forster and H. Sodemann, 2008: Remote source of water vapor forming precipitation on the Norwegian west coast at 60 N°—A tale of hurricanes and an atmospheric river. J. Geophys. Res., 113, D05102, https://doi.org/10.1029/2007JD009006.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sun, B., and H. J. Wang, 2014: Moisture sources of semiarid grassland in China using the Lagrangian particle model FLEXPART. J. Climate, 27, 24572474, https://doi.org/10.1175/JCLI-D-13-00517.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sun, B., and H. J. Wang, 2015: Analysis of the major atmospheric moisture sources affecting three sub-regions of East China. Int. J. Climatol., 35, 22432257, https://doi.org/10.1002/joc.4145.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Trenberth, K. E., 1998: Atmospheric moisture residence times and cycling: Implications for rainfall rates and climate change. Climatic Change, 39, 667694, https://doi.org/10.1023/A:1005319109110.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Trenberth, K. E., and D. P. Stepaniak, 2003: Covariability of components of poleward atmospheric energy transports on seasonal and interannual timescales. J. Climate, 16, 36913705, https://doi.org/10.1175/1520-0442(2003)016<3691:COCOPA>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • van der Ent, R. J., and O. A. Tuinenburg, 2017: The residence time of water in the atmosphere revisited. Hydrol. Earth Syst. Sci., 21, 779790, https://doi.org/10.5194/hess-21-779-2017.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, K. L., H. Jiang, and H. Y. Zhao, 2005: Atmospheric water vapor transport from westerly and monsoon over the Northwest China (in Chinese). Shuikexue Jinzhan, 16, 432438.

    • Search Google Scholar
    • Export Citation
  • Wang, X. R., X. D. Xu, and W. G. Wang, 2007: Characteristic of spatial transportation of water vapor for Northwest China’s rainfall in spring and summer (in Chinese). Plateau Meteor., 26, 749758.

    • Search Google Scholar
    • Export Citation
  • Wernli, H., 1997: A Lagrangian-based analysis of extratropical cyclones II: A detailed case-study. Quart. J. Roy. Meteor. Soc., 123, 16771706, https://doi.org/10.1002/qj.49712354211.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Yang, L. M., X. Li, and G. X. Zhang, 2011: Some advances and problems in the study of heavy rain in Xinjiang (in Chinese). Climatic Environ. Res., 16, 188198, https://doi.org/10.3878/j.issn.1006-9585.2011.02.08.

    • Search Google Scholar
    • Export Citation
  • Yang, L. M., Y. H. Zhang, and H. Tang, 2012: Analyses on water vapor characteristic in three heavy rainstorm processes of Xinjiang in July 2007 (in Chinese). Plateau Meteor., 31, 963973.

    • Search Google Scholar
    • Export Citation
  • Zhang, Y. H., B. X. Yu, Z. K. Wang, and L. H. Jia, 2018: Dynamic mechanism and water vapor transportation characteristics of two extreme rainstorm events in Ili River valley in summer of 2016 (in Chinese). Torrential Rain Disasters, 37, 435444.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    (a) Topographic distribution (m) and (b) distribution (mm) of annual average precipitation during 2008–15 and (c) distribution (mm) of total torrential rainfall (24-h accumulated precipitation > 24 mm) during April–September 2008–15 in Xinjiang.

  • View in gallery

    Probability density function (PDF; gray solid line) and cumulative distribution function (CDF; black solid line) of torrential rainfall in study areas a, b, c, and d.

  • View in gallery

    The large maps show the distribution of total torrential rainfall (24-h accumulated precipitation > 24 mm) during April–September 2008–15 in study areas a, b, c, and d in Xinjiang, and the small maps show the distribution of total rainfall (mm) of selected torrential events for substudy areas marked with rectangles in the large maps.

  • View in gallery

    Trajectories of the target particles 1–10 days before reaching the study area a, b, c, and d obtained by FLEXPART. The color of the trajectories refers to the AGL (m).

  • View in gallery

    Trajectories (500 randomly selected trajectories) of the target particles 1–10 days before reaching the study areas a, b, c, and d obtained by FLEXPART. The color of the trajectories refers to the specific humidity (g kg−1).

  • View in gallery

    Values of EP (mm) diagnosed from trajectories of the target particles 1–10 days before reaching the study areas a, b, c, and d obtained by FLEXPART. Black rectangles represent the moisture source regions to be examined in section 5: European region (A), north Asia region (B), Mediterranean/Black/Caspian Sea region (C), central Asia region (D), East Asia region (E), Bay of Bengal/Arabian Sea region (F), and the local region (T).

  • View in gallery

    Ratios (%) of the water vapor uptake from the examined water vapor source regions in the entire atmospheric layer to the total water vapor release within the study areas a, b, c, and d. These consist of three parts, namely, the part lost en route (orange), the part released over the study area (green), and the part that reached the study area but was not released (blue). The moisture source regions A–F are shown in Fig. 6; the T indicates the study area.

  • View in gallery

    As in Fig. 7, but for ratios (%) of the moisture uptake from the examined moisture source regions in the atmospheric boundary layer to the total moisture release within the study areas.

  • View in gallery

    Contributions (%) of the moisture source regions (A–F in Fig. 6) to the total water vapor released within the study areas a, b, c, and d calculated from 10-day trajectories of the air particles during the selected torrential rainfall events. The T indicates the study area, and Total represents total moisture contributions from all of the examined moisture source regions.

  • View in gallery

    Contributions (%) of the moisture source regions (A–F in Fig. 5) to the total water vapor released within the study areas a, b, c, and d calculated from 10-day trajectories of the air particles during April–September 2008–15. The T indicates the study area, and Total represents total moisture contributions from all of the examined moisture source regions.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 28 28 19
PDF Downloads 20 20 11

An Analysis of Moisture Sources of Torrential Rainfall Events over Xinjiang, China

View More View Less
  • 1 Laboratory of Cloud-Precipitation Physics and Severe Storms, Institute of Atmospheric Physics, Chinese Academy of Sciences, and College of Earth Science, University of Chinese Academy of Sciences, Beijing, China
  • 2 Key Laboratory of Land Surface Process and Climate Change in Cold and Arid Regions, Northwest Institute of Eco-Environment and Resources, Chinese Academy of Sciences, Lanzhou, and University of Chinese Academy of Sciences, Beijing, China
  • 3 Laboratory of Cloud-Precipitation Physics and Severe Storms, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, and Nanjing University of Information Technology, Nanjing, China
© Get Permissions
Open access

Abstract

Water vapor is a primary rainfall source for the development of torrential rainfall events. By using a Lagrangian flexible particle dispersion model (FLEXPART), the water vapor transports associated with torrential rainfall over Xinjiang, China, during April–September of 2008–15 are examined in this study. The results show that water vapor related to torrential rainfall events is mostly transported by westerly winds. The moisture sources for the development of torrential rainfall over four areas (Altay, Ili Valley, Hami, and Aksu-Kashgar) are mainly from Xinjiang and central Asia. The north Asia area and the Mediterranean/Black/Caspian Sea region are also important contributors to moisture source over the Altay area. Over Ili Valley, both the central Asia area and Xinjiang contribute 40% of water vapor to rainfall sources. Over the Hami area, 70% of the moisture source is from the Xinjiang. Over the Aksu-Kashgar area, the central Asia region is the most important moisture source area.

Denotes content that is immediately available upon publication as open access.

This article is licensed under a Creative Commons Attribution 4.0 license (http://creativecommons.org/licenses/by/4.0/).

© 2019 American Meteorological Society.

Corresponding author: Yushu Zhou, zys@mail.iap.ac.cn

Abstract

Water vapor is a primary rainfall source for the development of torrential rainfall events. By using a Lagrangian flexible particle dispersion model (FLEXPART), the water vapor transports associated with torrential rainfall over Xinjiang, China, during April–September of 2008–15 are examined in this study. The results show that water vapor related to torrential rainfall events is mostly transported by westerly winds. The moisture sources for the development of torrential rainfall over four areas (Altay, Ili Valley, Hami, and Aksu-Kashgar) are mainly from Xinjiang and central Asia. The north Asia area and the Mediterranean/Black/Caspian Sea region are also important contributors to moisture source over the Altay area. Over Ili Valley, both the central Asia area and Xinjiang contribute 40% of water vapor to rainfall sources. Over the Hami area, 70% of the moisture source is from the Xinjiang. Over the Aksu-Kashgar area, the central Asia region is the most important moisture source area.

Denotes content that is immediately available upon publication as open access.

This article is licensed under a Creative Commons Attribution 4.0 license (http://creativecommons.org/licenses/by/4.0/).

© 2019 American Meteorological Society.

Corresponding author: Yushu Zhou, zys@mail.iap.ac.cn

1. Introduction

Although Xinjiang’s overall precipitation is small, torrential rainfalls occur with an average of 2.3 regional torrential rainfall events occurring every year (Yang et al. 2011). For example, a 50-yr-record-breaking torrential rainfall event occurred over Xinjiang in July 1996, resulting in catastrophic floods along dozens of rivers. The daily precipitation of the 20 meteorological stations in Xinjiang exceeded the grade of heavy rainfall during 18–20 July 2004. A 60-yr-record-breaking heavy rainfall event occurred in the Korla area on 4 June 2012. Floods, landslides, and debris flows caused by torrential rainfall often lead to severe losses of life and property damage and thus have a great impact on the national economy and people’s livelihoods in Xinjiang.

Persistent water vapor transport is a necessary condition for the development of torrential rainfall systems (Gustafsson et al. 2010), making the study of water vapor transport and sources for heavy rains important (Newell et al. 1992). Located in the inland areas of Eurasia and not affected by the monsoon, the climate over northwestern China is completely different from that of the eastern monsoon region of China. Due to the influence of the westerlies throughout the year, water vapor in northwestern China mainly consists of the evapotranspiration transported from the western continent and oceans (Wang et al. 2005; Dai et al. 2007; Wang et al. 2007; Liu et al. 2008; Huang et al. 2010; Drumond et al. 2011). The water vapor transport from the Indian Ocean also has an impact on precipitation in the northwest region of China, where water vapor is abundant before crossing the Tibetan Plateau. Due to the blocking effect of the Tibetan Plateau, only a small part of the water vapor can reach the northwest inland region of China (Wang et al. 2007; Ao and Sun 2014). However, with the subtropical high extending northward and westward, water vapor in eastern China can be transported into Xinjiang through the Hexi Corridor (Yang et al. 2012; Zhang et al. 2018).

Previous studies on this subject have used Eulerian methods (Shi and Sun 2008; Yang et al. 2012, Cui 2013). In this study, the Lagrangian flexible particle dispersion model (FLEXPART) is used to study moisture sources of multiple torrential rainfall events in Xinjiang. FLEXPART, developed by the Norwegian Institute (NILU), was originally applied to the point source diffusion of pollutants and hazardous substances (such as nuclear pollution) and was later applied to gas exchanges in the troposphere and stratosphere (Stohl et al. 2005). In recent years, FLEXPART has been widely used in studying the global and regional atmospheric water cycle (Stohl and James 2004, 2005; Stohl et al. 2008; Sodemann et al. 2008; Drumond et al. 2008, 2011; Gimeno et al. 2010; Chen et al. 2011; Sun and Wang 2014; Huang and Cui 2015; Chen and Xu 2016). Stohl and James (2004) conducted a comparison study of global water transport fluxes using both the Eulerian and Lagrangian methods and found similar results, indicating that the FLEXPART model accurately simulates global atmospheric water cycle.

The data, model, and methodology are discussed in the next section. The characteristics of water vapor are discussed in section 3, and the sources of water vapor are identified in section 4. The results of quantitative analysis are present in section 5. The conclusions are given in section 6.

2. Data, model, and methodology

The FLEXPART model in this study is driven by the National Centers for Environmental Prediction’s Final Analysis (NCEP FNL) Operational Global Analysis data, which are available every 6 h with a 1° × 1° resolution on 26 vertical levels (http://rda.ucar.edu/datasets/ds083.2/). The precipitation data used in this study are from a 0.1° × 0.1° resolution dataset from 2008 to 2015, which was generated through hourly precipitation observations by automatic weather stations in China and merged with the CPC morphing technique (CMORPH) for satellite data (Pan et al. 2012; Shen et al. 2013).

a. Study areas and selection of torrential rainfall events

Figure 1 shows the topographical height of Xinjiang, the annual average precipitation from 2008 to 2015, and the accumulated precipitation distribution of torrential rainfall during April–September 2008–15 (24-h cumulative precipitation > 24 mm). The distribution of torrential rainfall events is similar to that of the annual average precipitation in Xinjiang, being mainly distributed along the line of the Tianshan Mountains and the Altay area. In general, similar to the mainland, torrential rains in Xinjiang are mainly distributed along the local topography (Ma and Xi 1992). Figure 1c reveals four areas with significant rainfall amounts: the Altay area (area a), Ili Valley (area b), the Hami area (area c), and the Aksu-Kashgar area (area d), which are selected for further analysis. Figure 2 shows the probability density function (PDF) and cumulative distribution function (CDF) curves of the torrential rainfall in which the inflection points (the point where the gray dotted lines intersect in Fig. 2) corresponding to the cumulative percentages and the precipitation are 86.64% and 127.24 mm for the Altay area, 86.64% and 200 mm for Ili Valley, 88% and 100.76 mm for the Hami area, and 88.73% and 147.07 mm for the Aksu-Kashgar area. According to the threshold value of the precipitation corresponding to the inflection point of each area, the approximate value was used to get the precipitation distribution (the heavy rain magnitude indicated by the shading in Fig. 3). From there, more study areas of interest (referred as the substudy areas) from the study areas a–d were selected: study areas a1–a3, study areas b1–b4, study areas c1–c2, and study areas d1–d4. Next, two steps are used to select the torrential rainfall events that occurred in the substudy areas as research objects. The first step is to use daily precipitation to identify the torrential rainfall grid points (24-h accumulated precipitation > 24 mm on each grid). Second, for a torrential rainfall event examined in this study, we require that at least 10% of the grid points in the study area are identified as torrential rainfall. There are 32, 68, 22, and 43 torrential rainfall events selected for study areas a–d, respectively (Table 1). We noted from the total precipitation of selected torrential rainfall events in the substudy areas (Fig. 3, small maps on the right and left) that the total precipitation is almost located in the center of each selected substudy area, indicating that the substudy areas and the torrential rainfall events we selected are of a certain representativeness. Although the time for integration based on the duration of individual events would be better to analyze the circulation and complete moisture transport process (Ciric et al. 2017), this paper analyzes the moisture source using the above methodology to select cases; this methodology focuses on the torrential precipitation center of each study area and calculates the associated moisture source (Huang and Cui 2015).

Fig. 1.
Fig. 1.

(a) Topographic distribution (m) and (b) distribution (mm) of annual average precipitation during 2008–15 and (c) distribution (mm) of total torrential rainfall (24-h accumulated precipitation > 24 mm) during April–September 2008–15 in Xinjiang.

Citation: Journal of Hydrometeorology 20, 10; 10.1175/JHM-D-19-0010.1

Fig. 2.
Fig. 2.

Probability density function (PDF; gray solid line) and cumulative distribution function (CDF; black solid line) of torrential rainfall in study areas a, b, c, and d.

Citation: Journal of Hydrometeorology 20, 10; 10.1175/JHM-D-19-0010.1

Fig. 3.
Fig. 3.

The large maps show the distribution of total torrential rainfall (24-h accumulated precipitation > 24 mm) during April–September 2008–15 in study areas a, b, c, and d in Xinjiang, and the small maps show the distribution of total rainfall (mm) of selected torrential events for substudy areas marked with rectangles in the large maps.

Citation: Journal of Hydrometeorology 20, 10; 10.1175/JHM-D-19-0010.1

Table 1.

The latitude and longitude range of the substudy areas in study areas in Xinjiang and the number of selected cases and selected particles.

Table 1.

b. Model and methodology

In this study, by using the domain filling mode, FLEXPART model simulations are driven for the region 80°–130°E, 0°–80°N, with a total of 2.4 million particles released, and a height of 25 km above the ground. The model integration time is from 0000 UTC 1 January 2008 to 1800 UTC 31 December 2015, with outputs recorded every 6 h, including the identification numbers of all particles, the three-dimensional position (horizontal latitude, longitude, and vertical height abbreviated as AGL), specific humidity, atmospheric boundary layer height, air density, and temperature. To select enough particles of interest (named as target particles), particles having a water vapor release within the study areas are selected as the target particles in this study.

In the Lagrangian method, the idea of studying moisture changes is to track particles and to diagnose the change of moisture based on the change of specific humidity of particles with time (Wernli 1997; James et al. 2004; Stohl and James 2004, 2005) as follows:
ep=m(dq/dt),
where m and q are the mass and specific humidity for a particle, respectively, and t refers to time. Parameters e and p are rates of moisture increase and decrease along the trajectory, respectively. A diagnosis of the trajectories of all particles (target particles) that contribute to rainfall within the study areas can be used to determine where the moisture increases. Assuming an area A is large enough for the target particles but small enough for the simulated area, the value of that ep in all target particles residing in the atmospheric column over the area A divided by the area A is the surface net freshwater flux of the area A. That is,
EP[i=1N(ep)]/A,
where E and P are the evaporation and precipitation rates per unit area, respectively (Stohl and James 2004). In the FLEXPART model, only EP can be diagnosed. However, E and P cannot be diagnosed separately, although one of them is always dominant during a short time interval (a few hours) (Trenberth and Stepaniak 2003). If EP > 0, the evaporation process is dominant and the instantaneous evaporation rate is EEP. If EP < 0, the precipitation process is dominant, and the instantaneous precipitation rate is PEP. Therefore, when we diagnose torrential rainfall events, a region with EP ≫ 0 may be a strong moisture source where particles contributing to rainfall in the study areas uptake water vapor as they pass through and eventually carry water vapor into the study area. According to the distribution of EP, the locations where particles take up water vapor can be investigated, thereby determining the moisture sources.

When quantitatively estimating the contribution of the moisture sources, EP of the source regions cannot be simply accumulated, because the particles may undergo multiple cycles of moisture uptake and release along their way to the target region. Therefore, considering the moisture changes of target particles along their trajectories to the target region is extremely important. To address this issue, Sodemann et al. (2008) introduced a method called “moisture source contribution,” which calculates the contribution of each particle’s location along its trajectory to the precipitation falling in the target region. Based on the “moisture source attribution” method, Sun and Wang (2014) made some alterations and proposed an “areal source–receptor attribution,” which is applied in this study. Further details can be found in Sun and Wang (2014). Since water vapor evaporation mainly occurs on Earth’s surface, only the increase of water vapor in the atmosphere boundary layer is caused by evaporation. Meanwhile, increases in water vapor above the atmospheric boundary layer are mainly caused by convection, turbulence, precipitation, and then evaporation. Therefore, the water vapor contribution in the atmospheric boundary layer and in the entire atmosphere layer was calculated separately in this study.

3. Characteristics of moisture transport

In this study, we trace the target particles backward from each substudy area in the study areas a–d for all the selected torrential rainfall events within a 10-day transport time; the trajectories of these particles are shown in Fig. 4, and their colored trajectories are represented for the above ground level in Fig. 4. As seen from the figures, the target particles in the study areas a–d (Figs. 4a–d) are mostly from the regions west of Xinjiang, including central Asia, the western part of North Asia, Europe, north Africa, the Mediterranean Sea, the Black Sea, the Atlantic Ocean, and the Arctic Ocean. The target particles within all study areas are mainly transported into Xinjiang along the western water vapor channel. Furthermore, Fig. 4 shows that the trajectories in the lower layers are shorter than those in the upper layers, indicating that the transport of the particles in the lower layers is much slower than that in the upper layers. For the study areas b–d, particles in the upper layers from the north Indian Ocean account for larger amounts. Study area a is located in the most northern part of Xinjiang, and its particles in the lower layers mainly come from the North Asia. Study areas b and d are located in western Xinjiang, and the particles in the lower layers come from the Bay of Bengal and along the southwestern Tibetan Plateau into Xinjiang. Study area d is located in southwestern Xinjiang, and the particles in the lower layers come from the eastern China and through the Hexi Corridor into Xinjiang. Figure 5 shows the specific humidity distribution of 10-day back trace trajectories of study areas a–d. For study area a, the particles with specific humidity of greater than 4 g kg−1 are mainly located in Caspian Sea and central and southern Siberia. For study area b, the particles with specific humidity of greater than 4 g kg−1 are mainly located in Xinjiang and the Indian Peninsula. For study area c, the particles with specific humidity of greater than 4 g kg−1 are mainly located in Xinjiang, with some higher specific humidity values also distributed in central and southern Siberia. For study area d, the particles with specific humidity of greater than 4 g kg−1 are mainly located in the Indian Peninsula and Xinjiang.

Fig. 4.
Fig. 4.

Trajectories of the target particles 1–10 days before reaching the study area a, b, c, and d obtained by FLEXPART. The color of the trajectories refers to the AGL (m).

Citation: Journal of Hydrometeorology 20, 10; 10.1175/JHM-D-19-0010.1

Fig. 5.
Fig. 5.

Trajectories (500 randomly selected trajectories) of the target particles 1–10 days before reaching the study areas a, b, c, and d obtained by FLEXPART. The color of the trajectories refers to the specific humidity (g kg−1).

Citation: Journal of Hydrometeorology 20, 10; 10.1175/JHM-D-19-0010.1

4. Identification of moisture sources

Figure 6 shows values of EP diagnosed from trajectories of the target particles 1–10 days before reaching the substudy areas (shown as green rectangles in Fig. 6) in study areas a–d. Section 2b noted that if EP > 0, there are moisture net uptakes, indicating moisture sources (shown in red shading in the Fig. 6). When EP < 0, there are moisture net releases, indicating moisture sinks (shown in blue shading in Fig. 6) (Sun and Wang 2014; Huang and Cui 2015; Chen and Xu 2016). The moisture in target particles may experience several times of uptake and release before reaching the target areas. The net moisture uptake in study area a (Fig. 6a) is mainly distributed around the East European Plain, West Siberia, the eastern part of the Mediterranean/Black/Caspian Sea region, the northern part of central Asia, and the study area, while the uptake in study area b is distributed around West Siberia, the eastern part of the Mediterranean/Black/Caspian Sea region, central Asia, and Xinjiang. Figure 6c shows that the net moisture uptake in study area c is mainly distributed around West Siberia, the eastern part of the Mediterranean/Black/Caspian Sea region, central Asia, and the study area. From Fig. 6d, the net moisture uptake in study area d is mainly distributed around the eastern part of the Mediterranean/Black/Caspian Sea region, central Asia, Xinjiang, and the Hexi Corridor. Otherwise, for study area a, there are net moisture release distributions in the coast of Europe region and study areas. For study area b, there are moisture net release distributions in the coast of Europe region, Indian Peninsula region, and adjacent region of the study areas. For study area c, there are net moisture release distributions in the study area and other regions in Xinjiang, and for study area d, there are net moisture releases in the coast of Europe region, Indian Peninsula region, and the adjacent region and local areas. The distribution of water vapor releases indicates the impact of the different locations of the study area. For example, in the Altay area, which is located in northern Xinjiang, the net moisture uptake is mostly located in the East European Plain and Siberia. While in the Aksu-Kashgar area, the net moisture uptake is mostly located in central Asia and north Indian Peninsula, and there is a large amount of moisture released in the Indian Peninsula and southwestern Tibetan plateau.

Fig. 6.
Fig. 6.

Values of EP (mm) diagnosed from trajectories of the target particles 1–10 days before reaching the study areas a, b, c, and d obtained by FLEXPART. Black rectangles represent the moisture source regions to be examined in section 5: European region (A), north Asia region (B), Mediterranean/Black/Caspian Sea region (C), central Asia region (D), East Asia region (E), Bay of Bengal/Arabian Sea region (F), and the local region (T).

Citation: Journal of Hydrometeorology 20, 10; 10.1175/JHM-D-19-0010.1

5. Quantitative estimation of water vapor contributions in moisture source

To better understand the contribution of different moisture sources to the torrential rainfall in different study areas, based on the geographical distribution, the particle trajectories (Fig. 4), and distribution of EP (Fig. 6), moisture source regions are divided into seven regions labeled A, B, C, D, E, F, and T (Fig. 6). They are European region (A), north Asia region (B), Mediterranean/Black/Caspian Sea region (C), central Asia region (D), East Asia region (E), Bay of Bengal/Arabian Sea region (F), and the local region (T, small green rectangle). For the East Asia region (E), most of EP is located in the Xinjiang region, therefore, this region can represent the Xinjiang region. To quantitatively study the moisture uptake and releases in particles from different sources to the study area, the moisture uptake in particles within different water vapor sources is divided into three parts, namely, the part that is lost en route (Loss), the part that is released over the study area (Released), and the part that has reached the study area but does not release (Unreleased). These methods are based on those of Sun and Wang (2015) and Huang and Cui (2015). Uptake, Loss, and Released in each moisture source are calculated by using the method called areal source–receptor attribution, and the calculation method for Unreleased is as follows: Unreleased = Uptake − Loss − Released.

Figures 7 and 8 show that the ratios of the water vapor in the entire atmosphere layer (Uptake AL) and in the atmospheric boundary layer (Uptake BL) that are taken up by particles from different moisture sources to the total water vapor released in the study area (Released total) are Uptake AL/Released total × 100% and Uptake BL/Released total × 100%, respectively, both of which consist of the part lost en route: Loss/Released total × 100% (orange rectangle); the released part over the study area: Released/Released total × 100% (green rectangle); and the unreleased part over the study area: Unreleased/Released total × 100% (blue rectangle). The local moisture uptake in the study area was also estimated, indicated by “T” in Figs. 7 and 8. For study areas b and d, the central Asia region (D) is the region with the most moisture uptake among all the moisture source regions examined. For study area a, the Mediterranean/Black/Caspian Sea region (C) has the highest uptake, but for study area c the East Asia region has the highest. The moisture source regions where the water vapor uptake ranked second and third are the north Asia region (B) and the central Asia region (D) for study area a; the Mediterranean/Black/Caspian Sea region (C) and East Asia region (E) for study areas b and d; then the Mediterranean/Black/Caspian Sea region (C) and the central Asia region (D) for study area c. By comparing AL/Released total × 100% and Uptake BL/Released total × 100%, respectively (Figs. 7 and 8), we observe that the moisture uptake in the atmospheric boundary layer is close to half of that in the entire atmosphere layer; this is related to the evaporation in the atmospheric boundary layer, but the ratio in the different source regions examined is the same as that of the entire atmosphere. Moreover, the part lost en route (orange rectangle) accounts for more than 50% of the total amount of the moisture in source regions examined for all study areas, indicating that the moisture is released before it is transported into the study area because of precipitation and other processes. That is, the water vapor transported from the source regions loses more than half its moisture en route before it reaches Xinjiang, and more than half of the water vapor in source regions has no effect on the precipitation in the target areas.

Fig. 7.
Fig. 7.

Ratios (%) of the water vapor uptake from the examined water vapor source regions in the entire atmospheric layer to the total water vapor release within the study areas a, b, c, and d. These consist of three parts, namely, the part lost en route (orange), the part released over the study area (green), and the part that reached the study area but was not released (blue). The moisture source regions A–F are shown in Fig. 6; the T indicates the study area.

Citation: Journal of Hydrometeorology 20, 10; 10.1175/JHM-D-19-0010.1

Fig. 8.
Fig. 8.

As in Fig. 7, but for ratios (%) of the moisture uptake from the examined moisture source regions in the atmospheric boundary layer to the total moisture release within the study areas.

Citation: Journal of Hydrometeorology 20, 10; 10.1175/JHM-D-19-0010.1

To explain more clearly the contribution of moisture sources to torrential rainfall in the study areas, the areal source–receptor attribution method is used to calculate the moisture contribution. Figure 9 shows the calculated ratios of the water vapor that is taken up within the entire atmosphere layer, and in the atmospheric boundary layer within the moisture source regions examined for water vapor release in the four study areas during the selected torrential rainfall events. Thus, the total moisture contribution of all the source regions can be calculated, which is represented by “Total” in the figure. The moisture source region where the contribution level ranks first is the East Asia region (E) except for study area d, with the contribution value varying slightly for each study area. Additionally, eastward locations of the study areas showed a greater moisture contribution from the East Asia region. According to the distribution of EP (Fig. 6), the circulation affecting Xinjiang is dominated by westerlies, so the more eastward the region, the greater the evaporation. For study area e (Fig. 6d), due to its location to the west, the water vapor is mainly supplied by the central Asia region (D); however, particles can blow into the study area a from the west gap of Xinjiang and the Hexi Corridor, which also brings much water vapor. Similarly, except for study area d, the moisture contribution of the central Asia region (D) ranks second for the other study areas. If a location is far away from the central Asia region, then the region’s moisture contribution will decrease. This is because the more eastward the study area, the longer the distance the water vapor needs to be transported, resulting in water vapor losing more moisture en route. Otherwise, the North Asia region (B) and the Mediterranean/Black/Caspian Sea region (C) are also important moisture source regions for study area a. Both regions account for more than 30% of the study area’s water vapor contribution. However, the contributions of other moisture source regions of all study areas account for less than 10%, proving that the Mediterranean/Black/Caspian Sea region (C) is the important moisture source region for torrential rainfall in Xinjiang. Moisture source regions vary according to the geographical differences of torrential rainfall centers: the more northerly the center, the more northerly the moisture source region may be distributed, and vice versa. However, the main moisture sources for Xinjiang’s torrential rains are Xinjiang and the central Asia region.

Fig. 9.
Fig. 9.

Contributions (%) of the moisture source regions (A–F in Fig. 6) to the total water vapor released within the study areas a, b, c, and d calculated from 10-day trajectories of the air particles during the selected torrential rainfall events. The T indicates the study area, and Total represents total moisture contributions from all of the examined moisture source regions.

Citation: Journal of Hydrometeorology 20, 10; 10.1175/JHM-D-19-0010.1

To get the difference of characteristics of moisture between the climatology and the torrential rainfall events, the water vapor contribution rate of moisture sources is selected during the period April–September 2008–15 and the settings of the FLEXPART are the same as the above experiments. Figure 10 shows the calculated ratios of the water vapor that is taken up within the entire atmosphere layer, and the atmospheric boundary layer within the moisture source regions is examined for water vapor release in the four study areas during April–September 2008–15. Like the torrential rainfall events, the central Asia region (D) and Xinjiang region (E) are main moisture source regions making the large amount of water vapor released within the four study areas, which are the common key moisture source for study areas. From the perspective of climatology, the contribution of moisture from the Xinjiang region (E) for the Ili Valley area, the Hami area, and the Aksu-Kashgar area apparently is reduced compared with the contribution during the torrential rainfall events, especially for the Aksu-Kashgar area. Meanwhile, the contribution of northwestern moisture source regions increases to some extent, yet the contribution of moisture from the central Asia region (D) and Xinjiang region (E) for the Altay area has apparently increased while the contribution of the north Asia region (B) has decreased to some extent.

Fig. 10.
Fig. 10.

Contributions (%) of the moisture source regions (A–F in Fig. 5) to the total water vapor released within the study areas a, b, c, and d calculated from 10-day trajectories of the air particles during April–September 2008–15. The T indicates the study area, and Total represents total moisture contributions from all of the examined moisture source regions.

Citation: Journal of Hydrometeorology 20, 10; 10.1175/JHM-D-19-0010.1

6. Conclusions and discussion

In this study, the Lagrangian flexible particle dispersion model (FLEXPART) is applied to investigate the main water vapor sources for torrential rainfall over many years within Xinjiang’s complex topography. Based on the torrential rainfall distribution in Xinjiang, four study areas are selected: the Altay area, Ili Valley, the Hami area, and the Aksu-Kashgar area. Main water vapor transport paths, source regions, and contributions to the torrential rainfall study areas are determined. Our major conclusions are as follows:

  1. Trajectories show that target particles entering Xinjiang almost all come from continents and oceans west of Xinjiang. Moreover, in Hami and the Aksu-Kashgar area, some particles with a lower height come from eastern China and through the Hexi Corridor into Xinjiang. In Ili Valley and the Aksu-Kashgar area, which are located in the western area of Xinjiang, some particles with a lower height come from the Indian Peninsula region and adjacent region. Furthermore, the specific humidity value of the trajectory depends mainly on the height and area position. When the lower layer trajectories located in the Indian Peninsula area and adjacent area, south of central Siberia, the Caspian Sea area, and Xinjiang, it could have a greater specific humidity value. However, if the moisture (specific humidity) of the trajectories releases (decreases) before it enters the study area, its influence on precipitation in Xinjiang will decrease. For example, the particles located in the Indian Peninsula area have larger specific humidity, but they also have a lot of moisture released before transport into study area. Therefore, the moisture contribution rate of the area in which these particles are located has little effect on the study area.
  2. A quantitative analysis of contribution for EP in the examined areas for torrential rainfall shows that there are two major moisture source regions: Xinjiang and the central Asia region. The north Asia and Mediterranean/Black/Caspian Sea regions are important moisture source regions for torrential rainfall events that occur in the Altay area. The torrential rainfall occurs in eastern Xinjiang Province, and the moisture contribution of the Xinjiang region accounts for a large proportion. The total moisture contribution of the entire atmosphere layer in the examined moisture source regions exceeds 90%, which means that the moisture source regions selected in this study have represented most of the moisture sources for each torrential rainfall study area.

In this study, we created a composite analysis of the moisture transport paths and the moisture sources of the torrential rains in different regions of Xinjiang. Common characteristics of moisture sources for torrential rainfall events in Xinjiang were obtained: the moisture in Xinjiang’s rainfall is mainly transported through the westerlies; easterlies can also transport some moisture. Xinjiang and the central Asia region are the main moisture source regions for Xinjiang’s torrential rains, and the moisture source regions of different torrential rainfall centers vary according to their geographical locations. For the Altay area, the moisture source still mainly comes from the Xinjiang local region and the central Asia region, but the north Asia area and the Mediterranean/Black/Caspian Sea region also are important moisture sources; for Ili Valley, the central Asia area and Xinjiang area have the same contribution rate of water vapor, which is about 40%; for the Hami area, the primary moisture source is the Xinjiang area and its contribution rate almost accounts for 70%; for the Aksu-Kashgar area, the central Asia region is the most important moisture source area, while the contribution rate of water vapor in the Xinjiang area shows its subsequent importance.

In the FLEXPART model, the evaporation or precipitation is indivisible (Stohl and James 2004, 2005), and the calculation of the contribution of moisture sources is based on the net EP distribution (Sodemann et al. 2008; Sun and Wang 2014). Thus the uptake of moisture within the atmospheric boundary layer is commonly caused by evaporation. However, above the atmospheric boundary layer, the uptake of moisture may be caused by many processes (Chen et al. 2011). As in Fig. 9, half of the uptake processes occur above the atmospheric boundary layer; these parts of moisture uptake are hard to figure out by this method. Besides, the trajectory display, EP calculation, and calculation of the contribution of moisture source regions are based on the 10-day backtrace results. Nieto and Gimeno (2019) discussed the effect of integration time of Lagrangian moisture transport to analyze and identify the moisture sources and sinks. Furthermore, Chen and Xu (2016) discussed the temporal variation of moisture contribution with different integration times; their conclusion was that 10-day backtrace moisture sources account for almost all the moisture contribution of the study area. Otherwise, although 10 days is not a long time for a water vapor cycle, it is close to the global mean residence time of atmospheric water, which is about 10 days (Chow et al. 1988; Trenberth 1998; van der Ent and Tuinenburg 2017). The length of backtrace time can affect the transport of water vapor. However, it can represent most of the moisture source if the total contribution of the moisture source accounts for over 80%. In this work, we mainly used the Lagrangian method to figure out the characteristics of moisture sources in Xinjiang Province. Though this method can explain the characteristics of moisture sources, there is still a lot of space for improvement in this method. Comprehensive use of various methods, given full play to the advantages of each method, can be more comprehensive and bring more accurate understanding of the source of moisture in Xinjiang Province during torrential rain events.

Acknowledgments

This work was supported by the National Natural Science Foundation of China (41661144024), National Key Research and Development Program (2018YFC1507104), Special Fund Project of Institute of Desert Meteorology of China Meteorological Administration(IDM2019007), and Pakistan Science Foundation 408 [PSF/NSFC-Earth/C-COMSATS-lsb (07)].

REFERENCES

  • Ao, J., and J. Q. Sun, 2014: Difference of water vapour transportation in the interannual and decadal variations of winter precipitation over western China (in Chinese). Climatic Environ. Res., 19, 497506.

    • Search Google Scholar
    • Export Citation
  • Chen, B., and X. D. Xu, 2016: Spatiotemporal structure of the moisture sources feeding heavy precipitation events over the Sichuan Basin. Int. J. Climatol., 36, 34463457, https://doi.org/10.1002/joc.4567.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chen, B., X. D. Xu, and X. H. Shi, 2011: Estimating the water vapor transport pathways and associated sources of water vapor for the extreme rainfall event over east of China in July 2007 using the Lagrangian method (in Chinese). Acta Meteor. Sin., 69, 810818.

    • Search Google Scholar
    • Export Citation
  • Chow, V. T., D. R. Maidment, and L. W. Mays, 1988: Applied Hydrology. McGraw-Hill, 572 pp.

  • Ciric, D., R. Nieto, A. Ramos, A. Drumond, and L. Gimen, 2017: Wet spells and associated moisture sources anomalies across Danube River basin. Water, 9, 615, https://doi.org/10.3390/w9080615.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Cui, C. X., 2013: Study on water vapor characteristic of typical heavy snowstorm case in Northern Xinjiang (in Chinese). Plateau Meteor., 32, 11151125.

    • Search Google Scholar
    • Export Citation
  • Dai, X. G., W. J. Li, Z. G. Ma, and P. Wang, 2007: Water-vapor source shift of Xinjiang region during the recent twenty years. Prog. Nat. Sci., 17, 569575, https://doi.org/10.1080/10020070708541037.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Drumond, A., R. Nieto, and L. Gimeno, 2008: A Langangian identification of major source of moisture over Central Brazil and La Plata Basin. J. Geophys. Res., 113, D14128, https://doi.org/10.1029/2007JD009547.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Drumond, A., R. Nieto, and L. Gimeno, 2011: Sources of moisture for China and their variations during drier and wetter conditions in 2000–2004: A Lagrangian approach. Climate Res., 50, 215225, https://doi.org/10.3354/cr01043.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gimeno, L., A. Drumond, R. Nieto, R. M. Trigo, and A. Stohl, 2010: On the origin of continental precipitation. Geophys. Res. Lett., 37, L13804, https://doi.org/10.1029/2010GL043712.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gustafsson, M., D. Rayner, and D. L. Chen, 2010: Extreme rainfall event in southern Sweden: Where does the moisture come from? Tellus, 62A, 605616, https://doi.org/10.1111/j.1600-0870.2010.00456.x.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Huang, Y. J., and X. P. Cui, 2015: Moisture sources of torrential rainfall events in the Sichuan basin of China during summers of 2009–13. J. Hydrometeor., 16, 19061917, https://doi.org/10.1175/JHM-D-14-0220.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Huang, R. H., J. L. Chen, and Y. Liu, 2010: Interdecadal variation of the leading modes of summertime precipitation anomalies over eastern China and its association with water vapor transport over East Asia (in Chinese). Chin. J. Atmos. Sci., 35, 589606.

    • Search Google Scholar
    • Export Citation
  • James, P., A. Stohlet, N. Spichtinger, and S. Eckhardt, 2004: Climatological aspects of the extreme European rainfall of August 2002 and a trajectory method for estimating the associated evaporative source regions. Nat. Hazards Earth Syst. Sci., 4, 733746, https://doi.org/10.5194/nhess-4-733-2004.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Liu, J. R., X. F. Song, G. F. Yuan, X. M. Sun, X. Liu, F. Chen, Z. M. Wang, and S. Q. Wang, 2008: Characteristics of δ18O in precipitation over Northwest China and its water vapor sources (in Chinese). Acta Geogr. Sin., 63, 1222.

    • Search Google Scholar
    • Export Citation
  • Ma, S. H., and Y. W. Xi, 1997: Some regularities of storm rain fall in Xinjiang, China (in Chinese). Acta Meteor. Sin., 55, 239248.

  • Newell, R. E., N. E. Newell, Y. Zhu, and C. Scott, 1992: Tropospheric rivers? A pilot study. Geophys. Res. Lett., 19, 24012404, https://doi.org/10.1029/92GL02916.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Nieto, R., and L. Gimeno, 2019: A database of optimal integration time for Lagrangian studies of atmospheric moisture sources and sinks. Sci. Data, 6, 59, https://doi.org/10.1038/s41597-019-0068-8.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Pan, Y., Y. Shen, J. J. Yu, and P. Zhao, 2012: Analysis of the combined gauge-satellite hourly precipitation over China based on the OI technique (in Chinese). Acta Meteor. Sin., 70, 13811389.

    • Search Google Scholar
    • Export Citation
  • Shen, Y., Y. Pan, J. J. Yu, P. Zhao, and Z. J. Zhou, 2013: Quality assessment of hourly merged precipitation product over China (in Chinese). Trans. Atmos. Sci., 36, 3746.

    • Search Google Scholar
    • Export Citation
  • Shi, Y. G., and Z. B. Sun, 2008: Climate characteristic of water vapor transportation and its variation over Xinjiang (in Chinese). Plateau Meteor., 27, 310319.

    • Search Google Scholar
    • Export Citation
  • Sodemann, H., C. Schwierz, and H. Wernli, 2008: Interannual variability of Greenland winter precipitation sources: Lagrangian moisture diagnostic and North Atlantic Oscillation influence. J. Geophys. Res., 113, D03107, https://doi.org/10.1029/2007JD008503.

    • Search Google Scholar
    • Export Citation
  • Stohl, A., and P. James, 2004: A Lagrangian analysis of the atmospheric branch of the global water cycle. Part I: Method description, validation, and demonstration for the August 2002 flooding in central Europe. J. Hydrometeor., 5, 656678, https://doi.org/10.1175/1525-7541(2004)005<0656:ALAOTA>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Stohl, A., and P. James, 2005: A Lagrangian analysis of the atmospheric branch of the global water cycle. Part II: Moisture transports between Earth’s ocean basins and river catchments. J. Hydrometeor., 6, 961984, https://doi.org/10.1175/JHM470.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Stohl, A., C. Forster, A. Frank, P. Seibert, and G. Wotawa, 2005: Technical note: The Lagrangian particle dispersion model FLEXPART version 6.2. Atmos. Chem. Phys., 5, 24612474, https://doi.org/10.5194/acp-5-2461-2005.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Stohl, A., F. Forster and H. Sodemann, 2008: Remote source of water vapor forming precipitation on the Norwegian west coast at 60 N°—A tale of hurricanes and an atmospheric river. J. Geophys. Res., 113, D05102, https://doi.org/10.1029/2007JD009006.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sun, B., and H. J. Wang, 2014: Moisture sources of semiarid grassland in China using the Lagrangian particle model FLEXPART. J. Climate, 27, 24572474, https://doi.org/10.1175/JCLI-D-13-00517.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sun, B., and H. J. Wang, 2015: Analysis of the major atmospheric moisture sources affecting three sub-regions of East China. Int. J. Climatol., 35, 22432257, https://doi.org/10.1002/joc.4145.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Trenberth, K. E., 1998: Atmospheric moisture residence times and cycling: Implications for rainfall rates and climate change. Climatic Change, 39, 667694, https://doi.org/10.1023/A:1005319109110.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Trenberth, K. E., and D. P. Stepaniak, 2003: Covariability of components of poleward atmospheric energy transports on seasonal and interannual timescales. J. Climate, 16, 36913705, https://doi.org/10.1175/1520-0442(2003)016<3691:COCOPA>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • van der Ent, R. J., and O. A. Tuinenburg, 2017: The residence time of water in the atmosphere revisited. Hydrol. Earth Syst. Sci., 21, 779790, https://doi.org/10.5194/hess-21-779-2017.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, K. L., H. Jiang, and H. Y. Zhao, 2005: Atmospheric water vapor transport from westerly and monsoon over the Northwest China (in Chinese). Shuikexue Jinzhan, 16, 432438.

    • Search Google Scholar
    • Export Citation
  • Wang, X. R., X. D. Xu, and W. G. Wang, 2007: Characteristic of spatial transportation of water vapor for Northwest China’s rainfall in spring and summer (in Chinese). Plateau Meteor., 26, 749758.

    • Search Google Scholar
    • Export Citation
  • Wernli, H., 1997: A Lagrangian-based analysis of extratropical cyclones II: A detailed case-study. Quart. J. Roy. Meteor. Soc., 123, 16771706, https://doi.org/10.1002/qj.49712354211.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Yang, L. M., X. Li, and G. X. Zhang, 2011: Some advances and problems in the study of heavy rain in Xinjiang (in Chinese). Climatic Environ. Res., 16, 188198, https://doi.org/10.3878/j.issn.1006-9585.2011.02.08.

    • Search Google Scholar
    • Export Citation
  • Yang, L. M., Y. H. Zhang, and H. Tang, 2012: Analyses on water vapor characteristic in three heavy rainstorm processes of Xinjiang in July 2007 (in Chinese). Plateau Meteor., 31, 963973.

    • Search Google Scholar
    • Export Citation
  • Zhang, Y. H., B. X. Yu, Z. K. Wang, and L. H. Jia, 2018: Dynamic mechanism and water vapor transportation characteristics of two extreme rainstorm events in Ili River valley in summer of 2016 (in Chinese). Torrential Rain Disasters, 37, 435444.

    • Search Google Scholar
    • Export Citation
Save