Abstract

Understanding the inconsistency in the effects of monsoon changes on drought and flood occurrences would allow scientists to identify useful indicators in the prediction and early warning of regional drought and flood. Based on the calculation of the Standardized Precipitation Index, monsoon indices, and water vapor fluxes from 1956 to 2015, the relationships between drought–flood occurrences and monsoons in different regions of the Lancang River Basin were investigated. Drought and flood occurrences had spatial differences. Areas located in the lower basin had high drought and flood occurrences. The frequencies of drought and flood occurrences have no obvious regional differences and mainly varied periodically at 3–5, 8–15, and 20–25 years. Because the impact and strength of the Tibetan Plateau Monsoon (TPM) and South Asian monsoon (SAM) are limited, the TPM and SAM are the key factors that affect the occurrences of drought and flood in the upstream and downstream regions of the Lancang River Basin, respectively. The TPM and SAM are potentially useful indicators in the prediction of drought and flood occurrences. These results are of great scientific merit in developing an effective mitigation strategy to reduce the impacts of drought–flood disasters in the Lancang River Basin.

INTRODUCTION

The frequency of extreme weather events has increased due to climate change, which has implications for the lives of people and the sustainable development of regions. This problem has attracted special attention from scientists, regulators, and governments (Lesk et al. 2016; Stott 2016; Hussain et al. 2018). Evidence strongly indicates that extreme weather events, such as droughts and floods, are the most severe natural disasters globally. These disasters inflict frequent and consequential economic damage. Therefore, the accurate prediction, spatiotemporal distribution, and early warning of regional drought and flood have become important scientific issues (Yan et al. 2012; Roudier et al. 2016; Tarawneh & Khalayleh 2016).

The frequency of high-intensity, high-frequency, and long-duration droughts and floods has increased recently. These extreme events have impeded the economic and social development of the Lancang River Basin (Pokhrel et al. 2018). Parts of the Lancang River Basin had suffered from its worst-ever drought in 2010, leaving millions of people with insufficient water and rendering large areas of farmland barren. The drought resulted in a direct economic loss of billions of RMB in the Lancang River Basin (Li et al. 2013a). In 2005, the worst flood in recorded history caused record-breaking destruction across the lower Lancang River Basin. The flood left millions of people homeless and many others destitute. Droughts and floods adversely affect water resources and the ecological environment (Liu et al. 2012) as well as foreign and security relations in the region (Stone 2010; Li et al. 2013a). The problem of droughts and floods in the Lancang River Basin has been the constant focus of scholarly attention in the region and elsewhere.

The Lancang River Basin is located in a transitional zone where multiple monsoon systems interact with each other. Monsoons are one of the major controlling factors in the formation of droughts and floods (Zhao et al. 2012). The basin is a key area that reflects the effects of monsoon systems on drought and flood occurrences. Moreover, because the Lancang–Mekong River is an important international river in Asia, understanding the teleconnections between drought–flood and monsoons in the Lancang River Basin can ameliorate the grievances that downstream states have regarding drought and flood occurrences (Xue et al. 2011). To date, most scholarly attention on the Lancang River Basin has been focused on spatiotemporal variations of drought and flood occurrences. Li et al. (2013b) investigated the spatial and temporal characteristics of meteorological droughts in the Lancang River Basin from 1960 to 2005 by using the Standardized Precipitation Index (SPI). More than 60% of the drought variation in the Lancang River Basin is associated with fluctuations in precipitation. Liu et al. (2015) analyzed the characteristics of drought variation at different timescales and explored the relationship between vegetation characteristics and droughts. However, the inconsistency of the effects of monsoon changes on drought and flood occurrences across regions have not been carefully analyzed thus far. Understanding this inconsistency would allow scientists to identify useful indicators in the prediction and early warning of regional drought and flood in the Lancang River Basin.

Therefore, the aim of this study was to investigate drought–flood variations and their teleconnections with monsoons in the Lancang River Basin during 1956–2015. Specifically, this study aimed to answer two questions: How do drought and flood occurrences relate to the monsoons in different regions of the Lancang River Basin? And how do drought and flood occurrences vary over spatiotemporal scales in this relationship?

STUDY AREA AND DATA

The Lancang–Mekong River has its source in the Tibetan plateau and its mouth in the southern region of Vietnam. Its Chinese section is called the Lancang River, and its Southeast Asian section is called the Mekong River. The Lancang–Mekong River is an important international river in Asia. The river has a total length of 1,240 km and a drainage basin area of 91,000 km2 in the Yunnan Province. Monsoons considerably influence the weather and climate conditions in the Lancang River Basin. The rainfall is unevenly distributed throughout the year. The wet season lasts from June to October, and the dry season lasts from November to May of the following year (Jacobs 2002).

Precipitation data and National Centers for Environmental Prediction (NCEP) reanalysis data were used in this study. Monthly precipitation data were obtained from the Chinese National Meteorological Center (http://cdc.cma.gov.cn). The data were obtained from eight national meteorological stations in China (Deqin, Dali, Jingdong, Lincang, Lancang, Jiangcheng, Jinghong, and Mengla) (Figure 1). All the reanalysis data used in the calculation processes were obtained from the US government's National Oceanic and Atmospheric Administration's Earth System Research Laboratory (http://www.esrl.noaa.gov/psd/data/gridded/data.ncep.reanalysis.pressure.html). Due to data availability constraints, quality-controlled data from 1956 to 2015 were used in this study.

Figure 1

Sketch map of the study area.

Figure 1

Sketch map of the study area.

METHODS

Standardized Precipitation Index

Compared with the Z and Palmer drought indices, the SPI is more commonly used to characterize drought and flood in hydrometeorological systems (Zhang et al. 2017; Bahrami et al. 2018). The SPI is useful in the spatial analysis of drought and flood. The index only requires precipitation data and can compare drought and flood across climatic regions and timescales.

Because precipitation is typically not normally distributed, the SPI overcomes this disadvantage by first fitting an incomplete gamma distribution to the data and then transforming it into a normal distribution. As a normalized index, drought and flood can be represented and monitored using the dimensionless SPI. Negative values indicate drought, and positive values indicate flood conditions. The SPI is defined as follows:
formula
(1)
where xi is the rainfall, xm is the mean precipitation, and s is the standard deviation of precipitation.

The SPI can be applied at different timescales (e.g., 1-, 3-, 6-, and 12-month timescales). Comparing the SPI with other timescales, the SPI with a 12-month timescale (SPI-12) is less affected by short-term variation in precipitation and can thus more accurately represent interannual drought and flood variations (Paulo et al. 2016). In this study, SPI-12 was calculated from the monthly precipitation data from 1956 to 2015. In our calculation method, the data of the first 11 months were excluded from the SPI-12 series. To facilitate comparisons between the SPI-12 series, the drought and flood analysis was first performed with the data for 1957. According to the drought and flood categories proposed by McKee et al. (1993), an SPI less than −1 indicates drought and an SPI larger than 1 indicates flood. According to these drought and flood categories, the numbers of drought and flood events that occurred each year were calculated over the annual, wet season, and dry season timescales. According to the occurrences of drought and flood events with respect to each category, the change characteristics of drought and flood occurrences in the Lancang River Basin were analyzed over the annual, wet season, and dry season timescales.

Monsoon indices

The Lancang River Basin is located in the transitional zone where drought and flood occurrences are influenced by multiple monsoons (Li et al. 2011; Xue et al. 2011). To clarify the effects of monsoons on drought and flood occurrences, three crucial and relatively independent monsoons, namely the Tibetan Plateau monsoon (TPM), South Asian monsoon (SAM), and East Asian monsoon (EAM), were considered in this study.

The TPM indices were obtained from the NCEP reanalysis data (a geopotential height of 600 hPa) by using the plateau monsoon index formula of Tang et al. (1984). The EAM indices were calculated from the NCEP reanalysis data (meridional wind of 850 hPa averaged over 20–30°N and 110–130°E) by applying the formula of Li & Zeng (2005). The SAM indices were calculated using the formula of Goswami et al. (1999) (the differences in the values between the 850- and 200-hPa zonal winds were averaged over 10–30°N and 70–110°E). The aforementioned definitions, which are straightforward and have a clear physical meaning, best reflect the variation characteristics of monsoons. The variations in the aforementioned monsoon indices and the variations in precipitation agreed well with each other. Thus, we can use the aforementioned monsoon indices to explore the relationships between drought–flood occurrences and monsoons.

Spatiotemporal changes in drought and flood occurrences

The cluster analysis, which is based on the Euclidean distance, is widely used in partitioning research, because it can efficiently handle long-term time series data (Oliveira et al. 2017; Raziei 2018). We used the cluster analysis to explore the regional differences in the characteristics of drought and flood occurrences in the Lancang River Basin.

The Mann–Kendall method, which is a nonparametric statistical test recommended by the World Meteorological Organization, has been proved to be versatile in analyzing trends in hydrometeorological data (Tang et al. 2014; Dawood 2017). We used this method to explore the trends in drought and flood occurrences at the annual and seasonal timescales.

The Mann–Kendall method involves calculating the Kendall rank correlation of a time series. The test statistic S is calculated using the following formula:
formula
(2)
and
formula
(3)
where xi and xj are the values during periods i and j (j > i), respectively, and n is the length of the time series.
The test statistic Z is calculated according to S and Var(S) as follows:
formula
(4)

The Z value is used to calculate a statistically significant trend. A positive Z value indicates an increasing trend, and a negative Z value indicates a decreasing trend (Kendall 1975). An absolute Z value of >1.96 is considered to be statistically significant at the 0.05 level.

Evolving characteristics of drought and flood occurrences

Compared with traditional methods, such as the serial correlation analysis and the Fourier transform, the wavelet analysis method is more effective for the time series analysis. The wavelet analysis method has been widely applied for analyzing variability modes at multitime scales (Sang 2013; Thiombiano et al. 2017). Thus, we used the wavelet analysis method to analyze the time series of drought and flood occurrences in the Lancang River Basin.

Wavelet transform is the decomposition of a target time series n by a mother wavelet (Farge 1992). Specifically,
formula
(5)
where Wn(s,t), the wavelet coefficient, denotes the wavelet transform of the time series at scale s and time t. The parameter is derived through the contraction and expansion of the mother wavelet.
The Morlet wavelet is commonly used to extract variability modes in time series data particularly, because it can provide an optimal tradeoff between localization in time and frequency (Grinsted et al. 2004). It is defined as follows:
formula
(6)
where w0 is the nondimensional frequency and w0 is set as 6 (Farge 1992).

Teleconnections between monsoons and drought–flood occurrences

Compared with the nonlinear correlation method, the linear correlation method is more commonly used to explore the teleconnections between hydrometeorological factors and large-scale atmospheric circulation systems. The linear correlation method is efficient (Chiew & McMahon 2002). Therefore, we used a linear correlation to identify the teleconnections between drought–flood occurrences and monsoons in the Lancang River Basin.

Water vapor fluxes are a crucial part of regional energy and water cycle processes and are critical factors in the prediction of regional drought and flood occurrences (Ryu et al. 2015; Liu et al. 2016). An analysis of water vapor fluxes can reveal the characteristics of regional drought and flood occurrences. We adapted a commonly used method, which was first proposed by Rasmusson (1968), to calculate water vapor fluxes (Li et al. 2018) from the NCEP reanalysis data. The formulas in the adopted method are as follows:
formula
formula
formula
(7)
where Q is the mean water vapor content of a column; g is the acceleration due to gravity; ps and pu are the lower and upper limits of the air pressure, respectively; q is the specific humidity; V is the wind velocity vector; Qλ and Qφ are the zonal and radial components of the vertically integrated total mean vapor fluxes, respectively; i and j are the unit vectors directed positively to the east and north, respectively; and u and v are the zonal and radial winds, respectively.

RESULTS

Spatiotemporal changes in drought and flood occurrences in the Lancang River Basin

The spatial distribution characteristics of drought and flood were obtained using the cluster analysis on the SPI-12 data of eight meteorological stations in the Lancang River Basin. The cluster analysis method could clearly categorize the SPI-12 data of the eight stations into three categories, with each category containing data from more than one station (Figure 2). The first category included the Deqin and Dali stations, which are located in the upper basin. The second category included the Jingdong, Lincang, Lancang, and Jiangcheng stations, which are located in the middle basin. The third category included the Jinghong and Mengla stations, which are located in the lower basin.

Figure 2

Clustering results for SPI-12 in the Lancang River Basin.

Figure 2

Clustering results for SPI-12 in the Lancang River Basin.

To better present the differences among the three categories, the average SPI-12 values were calculated for the three categories. The average SPI-12 values were −0.02, −0.004, and −0.006 for the first, second, and third categories, respectively. In the Lancang River Basin, the southeast part has a relatively wet condition, whereas the northwest part is relatively dry.

According to the clustering results of the SPI-12 data, the Deqin, Lincang, and Jinghong stations fell into three distinct categories and were therefore selected as the representative stations.

Table 1 details the descriptive statistics of drought and flood occurrences at the representative stations in the Lancang River Basin. The occurrences of drought events over the annual timescale during the past 60 years were ranked in the following order: Jinghong > Deqin > Lincang. The occurrences of drought events at the representative stations were higher during the dry season than during the wet season. Drought events were frequent during the dry season, especially in the late spring and early summer (Xie et al. 2004). The occurrences of flood events over the annual timescale during the past 60 years were ranked in the following order: Jinghong > Lincang > Deqin. The occurrences of flood events at the representative stations were higher during the dry season than during the wet season. The occurrences of flood events decreased gradually when moving from downstream locations to upstream ones.

Table 1

Drought and flood occurrences at representative stations in the Lancang River Basin

StationsDrought occurrences
Flood occurrences
AnnualMonsoon seasonNon-monsoon seasonAnnualMonsoon seasonNon-monsoon season
Deqin 126 50 76 96 39 57 
Lincang 107 50 57 110 40 70 
Jinghong 137 51 86 123 53 70 
StationsDrought occurrences
Flood occurrences
AnnualMonsoon seasonNon-monsoon seasonAnnualMonsoon seasonNon-monsoon season
Deqin 126 50 76 96 39 57 
Lincang 107 50 57 110 40 70 
Jinghong 137 51 86 123 53 70 

Table 2 lists the trend analysis results of drought and flood occurrences at the representative stations in the Lancang River Basin during the past 60 years. Over the annual timescale, drought occurrences had an increasing trend and the trend of the Deqin station (located in the upper basin) was statistically significant at the 0.05 level. During the wet season, increasing trends were observed in the series of drought occurrences; however, these trends were not statistically significant at the 0.05 level. During the dry season, drought occurrences had an increasing trend and the trend of the Deqin station was statistically significant at the 0.05 level. Since 2000, drought occurrences in Southwest China have increased. The trend analysis results confirmed prior observations of the changes in the characteristics of drought occurrences in Southwest China (He et al. 2011; Liu et al. 2015). Finally, flood occurrences had a decreasing trend over the annual and seasonal timescales; however, the trend was not statistically significant at the 0.05 level.

Table 2

Mann–Kendall trend test results for drought and flood occurrences at representative stations in the Lancang River Basin

StationsDrought occurrences
Flood occurrences
AnnualMonsoon seasonNon-monsoon seasonAnnualMonsoon seasonNon-monsoon season
Deqin 1.98* 0.36 2.05* −0.34 −0.16 −0.02 
Lincang 0.99 0.97 1.35 −0.52 −0.25 −0.38 
Lancang 1.55 1.13 1.54 −1.17 −1.07 −0.62 
StationsDrought occurrences
Flood occurrences
AnnualMonsoon seasonNon-monsoon seasonAnnualMonsoon seasonNon-monsoon season
Deqin 1.98* 0.36 2.05* −0.34 −0.16 −0.02 
Lincang 0.99 0.97 1.35 −0.52 −0.25 −0.38 
Lancang 1.55 1.13 1.54 −1.17 −1.07 −0.62 

‘*’ indicates that it is statistically significant at the 0.05 level.

Drought is a prominent meteorological disaster in the Lancang River Basin. Drought and flood occurrences have seasonal differences. The occurrences of drought and flood events at the representative stations were higher during the dry season than during the wet season. Drought and flood occurrences also have spatial differences. The Jinghong station (located in the lower basin) recorded a high occurrence of drought and flood, which indicated that Jinghong is a drought- and flood-prone area (Duan et al. 2007; Wu & Yan 2014).

Evolving characteristics of drought and flood occurrences in the Lancang River Basin

To better reveal the evolving characteristics of drought and flood occurrences in the Lancang River Basin, the real part of the wavelet coefficients of the series of drought and flood occurrences over the annual timescale was calculated, as illustrated in Figures 3 and 4.

Figure 3

Contour maps of the real part of the wavelet transformation coefficients for the flood frequency at the representative sites in the Lancang River Basin: (a) Deqin, (b) Lincang, and (c) Jinghong.

Figure 3

Contour maps of the real part of the wavelet transformation coefficients for the flood frequency at the representative sites in the Lancang River Basin: (a) Deqin, (b) Lincang, and (c) Jinghong.

Figure 4

Contour maps of the real part of the wavelet transformation coefficients for drought occurrences at the representative stations in the Lancang River Basin: (a) Deqin, (b) Lincang, and (c) Jinghong.

Figure 4

Contour maps of the real part of the wavelet transformation coefficients for drought occurrences at the representative stations in the Lancang River Basin: (a) Deqin, (b) Lincang, and (c) Jinghong.

The flood occurrences at the Deqin station (located in the upper basin) had two distinct periods: 5–10 and 20–30 years. The periodic changes of these two periods have persisted throughout the past 60 years. The flood occurrences at the Lincang station (located in the middle basin) had three distinct periods: 3–8, 10–16, and 20–30 years. The periodic changes in the 3–8-year period have appeared in the series of flood occurrences since the late 1990s. The periodic changes in the 5–10- and 15–25-year periods have persisted throughout the past 60 years. The periodic changes in the 3–8-, 10–18-, and 20–30-year periods have appeared in the series of flood occurrences at the Jinhong station (located in the lower basin). The periodic changes in the 3–8-year period primarily occurred during the 1950s–1970s and 2000s–2010s, and the other two periodic changes have persisted throughout the past 60 years (Figure 3).

According to Figure 4, the drought occurrences at the Deqin station (located in the upper basin) had three distinct periods: 3–5, 8–10, and 15–30 years. The periodic changes in the 3–5-year period mainly occurred during the 1970s and 1980s. Since the late 1980s, the periodic changes in the 8–10-year period have appeared in the series of drought occurrences. The periodic changes in the 15–30-year period have persisted throughout the past 60 years. The drought occurrences at the Lincang station (located in the middle basin) also had three distinct periods: 3–4, 5–10, and 15–25 years. The periodic changes in the 3–4-year period have mainly appeared since the late 1990s. The periodic changes in the 5–10- and 15–25-year periods have persisted throughout the past 60 years. The drought occurrences at the Jinhong station (located in the lower basin) had three distinct periods: 3–5, 8–10, and 15–25 years. The periodic changes in the 3–5-year period have mainly occurred since the mid-1960s, and the other two periodic changes have persisted throughout the past 60 years.

Teleconnections between monsoons and drought–flood occurrences in the Lancang River Basin

The linear correlation method was used to explore the teleconnections between SPI-12 and the monsoon indices during the corresponding period of drought and flood occurrences. As detailed in Table 3, SPI-12 was negatively correlated with the monsoon indices during drought occurrences. The correlation coefficients of the Deqin and Lincang stations in descending order for the monsoons were as follows: SAM > TPM > EAM. The correlation coefficients of the Jinghong station for the monsoons in descending order were as follows: SAM > EAM > TPM. The SAM had an important role in the drought occurrences in the Lancang River Basin. SPI-12 was positively correlated with monsoon indices during flood occurrences. The correlation coefficients at the three stations for the monsoons in descending order were as follows: SAM > TPM > EAM.

Table 3

Coefficients of correlation between SPI-12 and the monsoon indices at the representative sites in the Lancang River Basin

StationsSPI-12EAMSAMTPM
Deqin Drought occurrences −0.09 −0.18* −0.12 
Flood occurrences 0.07 0.17 0.16 
Lincang Drought occurrences −0.03 −0.27* −0.09 
Flood occurrences 0.02 0.22* 0.11 
Jinghong Drought occurrences −0.07 −0.34* −0.03 
Flood occurrences 0.02 0.25* 0.10 
StationsSPI-12EAMSAMTPM
Deqin Drought occurrences −0.09 −0.18* −0.12 
Flood occurrences 0.07 0.17 0.16 
Lincang Drought occurrences −0.03 −0.27* −0.09 
Flood occurrences 0.02 0.22* 0.11 
Jinghong Drought occurrences −0.07 −0.34* −0.03 
Flood occurrences 0.02 0.25* 0.10 

‘*’ indicates that it is statistically significant at the 0.05 level.

The convergence of water vapor transport is associated with large-scale atmospheric circulation systems and is a key condition in the formation of precipitation. Therefore, the driving effects of monsoon systems on drought and flood occurrences can be indicated by the evolution of regional water vapor fluxes. The characteristics of water vapor fluxes over a geopotential height of 500 hPa are a good indication of the precipitation distribution in the Lancang River Basin (Yan et al. 2007). Exploring the water vapor fluxes during the driest and wettest year can help us understand the effects of monsoons systems on drought and flood occurrences in the Lancang River Basin. The methods of calculating the areal average precipitation include the Thiessen polygon, arithmetic average, and isohyet methods. The Thiessen polygon method is mostly used when calculating the areal average precipitation in cases where rainfall stations are sparse. Therefore, we adopted this method to calculate the areal average precipitation. Extreme precipitation values were recorded in 1970 (most rainfall, the wettest year) and 2003 (least rainfall, the driest year) in the Lancang River Basin.

Figure 5 illustrates the distributions of water vapor fluxes at a geopotential height of 500 hPa in 1970 and 2003. The vector arrows of the wind fields indicate that a southwest airflow and a northwest airflow enter into the Lancang River Basin. The southwest airflow field in the downstream regions of the Lancang River Basin indicates that the prevailing winds in these regions blow from the southwest. The SAM brings a southwest airflow into the basin through the water vapor channel from the Indian Ocean and constitutes the dominant airflow direction and water vapor source in the region (Xue et al. 2011; Zhao et al. 2012). The northwest airflow field in the upstream regions of the Lancang River Basin indicates that the prevailing winds in these regions blow from the northwest. The wind flow may be influenced by the thermodynamic effect of the Qinghai–Tibet Plateau (Xue et al. 2011). The TPM brings a southwest airflow into the basin and forms the dominant airflow direction and water vapor source in the region. Due to the barrier effects of the topography of the Lancang River Basin (latitudinal barrier), the water vapor fluxes carried by the EAM have difficulty in flowing into the Lancang River Basin. The presence of the southwest airflow is therefore a good explanation of the relatively strong correlation between SPI-12 and the SAM in the downstream regions of the Lancang River Basin. Similarly, the presence of the northwest airflow is a good explanation of the relatively strong correlation between SPI-12 and the TPM in the upstream regions of the Lancang River Basin. The differences in water vapor fluxes between the driest and wettest years are mainly reflected in the influence of the intensity and range. During the wettest year, increased water vapor fluxes were brought into the Lancang River Basin by the southwest and northwest airflows.

Figure 5

Distributions of water vapor flux at 500 hPa for the driest year (2003) and wettest year (1970) in the Lancang River Basin: (a) driest year and (b) wettest year.

Figure 5

Distributions of water vapor flux at 500 hPa for the driest year (2003) and wettest year (1970) in the Lancang River Basin: (a) driest year and (b) wettest year.

DISCUSSION

Drought and flood occurrences have spatial differences in the Lancang River Basin. The drought and flood occurrences in the Lancang River Basin exhibited a distinct segmentation into three subregions upon the cluster analysis based on the Euclidean distance. This segmentation revealed the behavior pattern of the precipitation regime in the Lancang River Basin. The annual precipitation in the Lancang River Basin is more than 1,600 mm in the southeast and less than 250 mm in the northwest. Moreover, the precipitation decreases gradually from the south to the north (Shi et al. 2013; Wang et al. 2015).

Minor regional differences were observed in the periodic characteristics of the drought and flood occurrences in the Lancang River Basin. The drought and flood occurrences varied periodically mainly at 3–5, 8–15, and 20–25 years. The periods of 3–5, 8–15, and 20–25 years also exhibited changes in precipitation (Liu et al. 2008; Yu et al. 2015). The drought and flood occurrences were calculated using precipitation data. Therefore, the variations in drought-flood occurrences and the variations in precipitation agreed well with each other.

The map of wind field characteristics in the Lancang River Basin for 1970 and 2003 confirmed the assumptions regarding regional differences in the relationships between drought and flood occurrences and monsoons. The SAM brings a southwest airflow into the downstream regions of the Lancang River Basin and forms the dominant airflow direction and water vapor source in the region. Therefore, changes in the SAM are a key factor affecting drought and flood occurrences in the downstream regions of the Lancang River Basin. The prevailing winds in the upstream regions of the Lancang River Basin blow from the northwest. The TPM constitutes the dominant airflow direction and water vapor source in the upstream regions of the Lancang River Basin. Therefore, the TPM is a key factor affecting drought and flood occurrences in the downstream regions of the Lancang River Basin. The impact and strength of the SAM and TPM are limited. The relationship between SPI-12 and the SAM weakens gradually from downstream to upstream, whereas the relationship between SPI-12 and the TPM weakens gradually from upstream to downstream in the Lancang River Basin.

The SAM and TPM are potentially useful indicators in the prediction of drought and flood occurrences. By using certain indicators and soft computing techniques for hydrological prediction (Wu & Chau 2011; Chau 2017; Yaseen et al. 2019), the precise prediction and early warning of drought and flood occurrences can be achieved to assist policymakers in developing effective mitigation strategies that reduce the impact of drought and flood disasters.

CONCLUSIONS

According to our calculation of the SPI, monsoon indices, and water vapor fluxes from 1956 to 2015, drought–flood occurrences and their teleconnections with monsoons in different regions of the Lancang River Basin were investigated. We drew the following conclusions:

  • (1)

    Drought and flood occurrences had spatial differences. Areas located in the lower basin had high drought and flood occurrences. Therefore, these areas are prone to drought and flood.

  • (2)

    The frequencies of drought and flood occurrences had no obvious regional differences and mainly varied periodically at 3–5, 8–15, and 20–25 years.

  • (3)

    Regional differences existed in the effects of monsoon changes on drought and flood occurrences. Because the impact and strength of the TPM and SAM are limited, the TPM and SAM are the key factors that affect the occurrences of drought and flood in the upstream and downstream regions of the Lancang River Basin, respectively.

Although our study provides an insight into drought and flood occurrences, it has some shortcomings. Many factors, such as land use change and other interventions, may lead to the inhomogeneity of drought and flood occurrences. These factors may have increased the noise in the relationships between drought–flood occurrences and monsoons. Further study is required on superior methods for predicting and quantifying the effects of monsoons on drought and flood occurrences.

ACKNOWLEDGEMENTS

This study was supported and funded by the National Natural Science Foundation of China (grant no. 51609008), Key Special Project of the National Key Research and Development Program (grant no. 2016YFC0402309), and the Natural Science Foundation of Hubei Province (grant no. 2016CFA092). The authors greatly appreciate the excellent comments of the anonymous reviewers for improving the manuscript quality.

REFERENCES

Bahrami
M.
Bazrkar
S.
Zarei
A. R.
2018
Modeling, prediction and trend assessment of drought in Iran using Standardized Precipitation Index
.
Journal of Water and Climate Change
10
(
1
),
181
196
.
Chau
K. W.
2017
Use of meta-heuristic techniques in rainfall-runoff modelling
.
Water
9
(
3
),
186
.
https://doi.org/10.3390/w9030186
.
Chiew
F. H. S.
McMahon
T. A.
2002
Global ENSO-streamflow teleconnection, streamflow forecasting and interannual variability
.
Hydrological Sciences Journal
47
(
3
),
505
522
.
Duan
C. C.
Zhu
Y.
You
W. H.
2007
Characteristic and formation cause of drought and flood in Yunnan Province rainy season
.
Plateau Meteorology
26
(
2
),
402
408
.
(in Chinese)
.
Farge
M.
1992
Wavelet transforms and their applications to turbulence
.
Annual Review of Fluid Mechanics
24
(
1
),
395
458
.
Goswami
B. N.
Krishnamurthy
V.
Annmalai
H.
1999
A broad-scale circulation index for the interannual variability of the Indian summer monsoon
.
Quarterly Journal of the Royal Meteorological Society
125
(
554
),
611
633
.
Grinsted
A.
Moore
J. C.
Jevrejeva
S.
2004
Application of the cross wavelet transform and wavelet coherence to geophysical time series
.
Nonlinear Processes in Geophysics
11
,
561
566
.
He
J. Y.
Zhang
M. J.
Wang
P.
Wang
S. J.
Wang
X. M.
2011
Climate characteristics of the extreme drought events in southwest China during recent 50 years
.
Acta Geographica Sinica
66
(
9
),
1179
1190
.
(in Chinese)
.
Hussain
A.
Rasul
G.
Mahapatra
B.
WahidSabarnee
S.
Tuladhar
S.
2018
Climate change-induced hazards and local adaptations in agriculture: a study from Koshi River Basin, Nepal
.
Natural Hazards
2
,
1
19
.
Kendall
M.
1975
Rank Correlation Methods
.
Charles Griffin
,
London
.
Lesk
C.
Rowhani
P.
Ramankutty
N.
2016
Influence of extreme weather disasters on global crop production
.
Nature
529
(
7584
),
84
.
Li
J. P.
Zeng
Q. C.
2005
A new monsoon index, its interannual variability and relation with monsoon precipitation
.
Climatic and Environmental Research
10
(
3
),
351
365
.
(in Chinese)
.
Li
B.
Li
L. J.
Li
H. B.
Liang
L. Q.
Li
J. Y.
Liu
Y. M.
Zeng
H. W.
2011
Changes in precipitation extremes in Lancang River Basin, 1960–2005
.
Progress in Geophysics
30
(
3
),
290
298
.
(in Chinese)
.
Li
H.
Gupta
J.
Van Dijk
M. P.
2013a
China's drought strategies in rural areas along the Lancang River
.
Water Policy
15
(
1
),
1
18
.
Li
B.
Su
H.
Chen
F.
Li
S. G.
Tian
J.
Qin
Y. C.
Zhang
R. H.
Chen
S. H.
Yang
Y. M.
Rong
Y.
2013b
The changing pattern of droughts in the Lancang River Basin during 1960–2005
.
Theoretical and Applied Climatology
111
(
3–4
),
401
415
.
Li
L.
Zhang
R.
Wen
M.
J.
2018
Effect of the atmospheric quasi-biweekly oscillation on the vortices moving off the Tibetan Plateau
.
Climate Dynamics
50
(
3–4
),
1193
1207
.
Liu
G.
Liu
J. P.
Zhao
R.
2008
Analysis of periodical variation of Lancang River based on wavelet transform
.
Computer Engineering and Applications
44
(
8
),
236
237
.
(in Chinese)
.
Liu
H.
Feng
C. L.
Chen
B. S.
Wang
Z. S.
Xie
X. Q.
Deng
Z. H.
2012
Overcoming extreme weather challenges: successful but variable assisted colonization of wild orchids in southwestern China
.
Biological Conservation
150
(
1
),
68
75
.
Liu
S. L.
Tian
Y. Y.
An
N. N.
Zhao
H. D.
Dong
S. K.
2015
Effects of climate change on normalized difference vegetation index based on the multiple analysis of standardized precipitation evapotranspiration index methods in the Lancang River Basin
.
Climatic and Environmental Research
20
(
6
),
705
714
.
(in Chinese)
.
McKee
T. B.
Doesken
N. J.
Kleist
J.
1993
The relationship of drought frequency and duration to time scale
. In
Eighth Conference on Applied Climatology
,
Anaheim, CA
.
Oliveira
P. T.
e Silva
C. M. S.
Lima
K. C.
2017
Climatology and trend analysis of extreme precipitation in subregions of Northeast Brazil
.
Theoretical and Applied Climatology
130
(
1–2
),
77
90
.
Paulo
A.
Martins
D.
Pereira
L. S.
2016
Influence of precipitation changes on the SPI and related drought severity. An analysis using long-term data series
.
Water Resources Management
30
(
15
),
5737
5757
.
Pokhrel
Y.
Burbano
M.
Roush
J.
Kang
H.
Sridhar
V.
Hyndman
D. W.
2018
A review of the integrated effects of changing climate, land use, and dams on Mekong river hydrology
.
Water
10
(
3
),
266
.
Raziei
T.
2018
A precipitation regionalization and regime for Iran based on multivariate analysis
.
Theoretical and Applied Climatology
131
(
3–4
),
1429
1448
.
Roudier
P.
Andersson
J. C.
Donnelly
C.
Feyen
L.
Greuell
W.
Ludwig
F.
2016
Projections of future floods and hydrological droughts in Europe under a +2°C global warming
.
Climatic Change
135
(
2
),
341
355
.
Ryu
Y. H.
Smith
J. A.
Bou-Zeid
E.
2015
On the climatology of precipitable water and water vapor flux in the mid-Atlantic region of the United States
.
Journal of Hydrometeorology
16
(
1
),
70
87
.
Shi
W.
Yu
X.
Liao
W.
Wang
Y.
Jia
B.
2013
Spatial and temporal variability of daily precipitation concentration in the Lancang River Basin, China
.
Journal of Hydrology
495
,
197
207
.
Stott
P.
2016
How climate change affects extreme weather events
.
Science
352
(
6293
),
1517
1518
.
Tang
M. C.
Liang
J.
Shao
M. J.
1984
Preliminary analysis on inter-annual variation of Plateau monsoon
.
Plateau Meteorology
3
(
3
),
76
82
.
(in Chinese)
.
Tang
J.
Yin
X. A.
Yang
P.
Yang
Z. F.
2014
Assessment of contributions of climatic variation and human activities to streamflow changes in the Lancang River, China
.
Water Resources Management
28
(
10
),
2953
2966
.
Tarawneh
Z.
Khalayleh
Y.
2016
Improved estimate of multiyear drought for water resources management studies
.
Journal of Water and Climate Change
7
(
4
),
721
730
.
Thiombiano
A. N.
El Adlouni
S.
St-Hilaire
A.
Ouarda
T. B. M. J.
El-Jabi
N.
2017
Nonstationary frequency analysis of extreme daily precipitation amounts in Southeastern Canada using a peaks-over-threshold approach
.
Theoretical and Applied Climatology
129
(
1–2
),
413
426
.
Wu
M. C.
Yan
J. P.
2014
Influence of solar activities and ENSO events on drought and flood in Yunnan Province
.
Bulletin of Soil and Water Conservation
34
(
4
),
280
284
.
(in Chinese)
.
Xie
M. E.
Cheng
J. G.
Fan
B.
2004
Spatiotemporal distribution features of meteorological disasters in Yunnan
.
Journal of Natural Disasters
13
(
5
),
40
47
.
(in Chinese)
.
Yan
H. M.
Wang
L.
Zhou
G. L.
Lu
Y. B.
2007
The relationship between rainfall in Yunnan summer and the circulation in preceding winter
.
Journal of Applied Meteorological Science
18
(
3
),
340
349
.
(in Chinese)
.
Yu
W. J.
Huang
Y. L.
Shao
M. Y.
2015
Research on characteristics of extreme weather disasters and fluctuations trend on Lancang River Basin
.
Acta Ecologica Sinica
35
(
5
),
1378
1387
.
(in Chinese)
.
Zhang
T.
Li
J.
Hu
R.
Wang
Y.
Feng
P.
2017
Drought class transition analysis through different models: a case study in North China
.
Water Science & Technology: Water Supply
17
(
1
),
138
150
.
Zhao
Q. H.
Liu
S. L.
Deng
L.
Dong
S. K.
Yang
J.
Wang
C.
2012
The effects of dam construction and precipitation variability on hydrologic alteration in the Lancang River Basin of southwest China
.
Stochastic Environmental Research and Risk Assessment
26
(
7
),
993
1011
.