Extreme Hydrometeorological Conditions and Changes in the Amu Darya River Basin in Central Asia

Amirkhamza Murodov aKey Laboratory of Tibetan Environment Changes and Land Surface Processes, Institute of Tibetan Plateau Research, Chinese Academy of Sciences, Beijing, China
bUniversity of Chinese Academy of Sciences, Beijing, China
cState Key Laboratory of Tibetan Plateau Earth System and Resources Environment (TPESRE), Institute of Tibetan Plateau Research, Chinese Academy of Sciences, Beijing, China

Search for other papers by Amirkhamza Murodov in
Current site
Google Scholar
PubMed
Close
,
Lan Cuo aKey Laboratory of Tibetan Environment Changes and Land Surface Processes, Institute of Tibetan Plateau Research, Chinese Academy of Sciences, Beijing, China
bUniversity of Chinese Academy of Sciences, Beijing, China
cState Key Laboratory of Tibetan Plateau Earth System and Resources Environment (TPESRE), Institute of Tibetan Plateau Research, Chinese Academy of Sciences, Beijing, China
dCenter for Excellence in Tibetan Plateau Earth Sciences, Chinese Academy of Sciences, Beijing, China

Search for other papers by Lan Cuo in
Current site
Google Scholar
PubMed
Close
,
Ning Li aKey Laboratory of Tibetan Environment Changes and Land Surface Processes, Institute of Tibetan Plateau Research, Chinese Academy of Sciences, Beijing, China
bUniversity of Chinese Academy of Sciences, Beijing, China
cState Key Laboratory of Tibetan Plateau Earth System and Resources Environment (TPESRE), Institute of Tibetan Plateau Research, Chinese Academy of Sciences, Beijing, China

Search for other papers by Ning Li in
Current site
Google Scholar
PubMed
Close
,
Davlatkhudzha Murodov eKey Laboratory of Western China’s Environmental Systems (Ministry of Education), College of Earth and Environmental Sciences, Lanzhou University, Lanzhou, China
fInstitute of Geology, Earthquake Engineering and Seismology of National Academy of Sciences of Tajikistan, Dushanbe, Tajikistan

Search for other papers by Davlatkhudzha Murodov in
Current site
Google Scholar
PubMed
Close
,
Mei Hou aKey Laboratory of Tibetan Environment Changes and Land Surface Processes, Institute of Tibetan Plateau Research, Chinese Academy of Sciences, Beijing, China
bUniversity of Chinese Academy of Sciences, Beijing, China
cState Key Laboratory of Tibetan Plateau Earth System and Resources Environment (TPESRE), Institute of Tibetan Plateau Research, Chinese Academy of Sciences, Beijing, China

Search for other papers by Mei Hou in
Current site
Google Scholar
PubMed
Close
, and
Gulfam Hussain bUniversity of Chinese Academy of Sciences, Beijing, China
gTaiwan International Graduate Program, Earth System Science, Academia Sinica, Taipei, Taiwan

Search for other papers by Gulfam Hussain in
Current site
Google Scholar
PubMed
Close
Free access

Abstract

The Amu Darya contributed 70% of the flow to the Aral Sea in central Asia before the 1960s, when the Amu Darya streamflow to the Aral Sea started to dwindle. The severe environmental and socioeconomic disaster happened mainly due to intensified water abstraction with the backdrop of climate change. However, knowledge of up to the most recent extreme climate conditions and their changes, as well as their relations to streamflow in the basin, is still lacking. This study aims to understand extreme hydrometeorological conditions and their changes, as well as their relations in the past several decades, especially in the upper Amu Darya basin. The spatial patterns of the means of all extreme temperature indices followed the elevation gradient. The majority of the basin showed an increasing trend in extreme warm events but a decreasing trend in extreme cold events. The north of the upper basin had over 1000 mm annual precipitation, and the east had less than 300 mm annual precipitation. Overall, the upper Amu Darya basin underwent a wetting and warming annual trend. Annual streamflow in the upper subbasins was less than 750 m3 s−1, but together they produced over 1500 m3 s−1 flow in the middle reach and basin outlet. Streamflow change varied among subbasins. Correlations between climatic factors and streamflow at annual time steps were weak but distinct at monthly time steps with lagged effects. In highland subbasins with high coverage of glaciers and snow, temperature minima and maxima impacts were opposite and overwhelmed precipitation, whereas in lowland subbasins, precipitation was more important.

© 2023 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: Lan Cuo, lancuo@itpcas.ac.cn

Abstract

The Amu Darya contributed 70% of the flow to the Aral Sea in central Asia before the 1960s, when the Amu Darya streamflow to the Aral Sea started to dwindle. The severe environmental and socioeconomic disaster happened mainly due to intensified water abstraction with the backdrop of climate change. However, knowledge of up to the most recent extreme climate conditions and their changes, as well as their relations to streamflow in the basin, is still lacking. This study aims to understand extreme hydrometeorological conditions and their changes, as well as their relations in the past several decades, especially in the upper Amu Darya basin. The spatial patterns of the means of all extreme temperature indices followed the elevation gradient. The majority of the basin showed an increasing trend in extreme warm events but a decreasing trend in extreme cold events. The north of the upper basin had over 1000 mm annual precipitation, and the east had less than 300 mm annual precipitation. Overall, the upper Amu Darya basin underwent a wetting and warming annual trend. Annual streamflow in the upper subbasins was less than 750 m3 s−1, but together they produced over 1500 m3 s−1 flow in the middle reach and basin outlet. Streamflow change varied among subbasins. Correlations between climatic factors and streamflow at annual time steps were weak but distinct at monthly time steps with lagged effects. In highland subbasins with high coverage of glaciers and snow, temperature minima and maxima impacts were opposite and overwhelmed precipitation, whereas in lowland subbasins, precipitation was more important.

© 2023 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: Lan Cuo, lancuo@itpcas.ac.cn

1. Introduction

Climate change has been causing severe disasters that are becoming more detrimental to societies and ecosystems in recent years. It is now widely known that climate warming is accompanied by enhanced extreme precipitation and temperature events across the globe since the 1950s (Alexander et al. 2006; Brohan et al. 2006; Arias et al. 2021). Extreme temperature events such as heat waves bring severe consequences on the social economy and human livelihood (Robine et al. 2008; Gu et al. 2016). Global warming intensifies the water cycle and aggravates the frequency and intensity of extreme precipitation events causing floods or droughts, which has a significant negative impact on the economy, environment, and society (Zhang et al. 2017).

As a typical arid and semiarid region, central Asian (CA) countries are vulnerable and susceptible to climate change (Lioubimtseva and Henebry 2009; He et al. 2014; Chang et al. 2016). The climate of this region is continental, subtropical, and semiarid with complicated interactions among solar radiation, atmospheric circulation, and complex orography (Mahmadaliev 2009). Unger-Shayesteh et al. (2013) reported statistically significant rates of temperature change ranging from 0.18° to 0.42°C decade−1 in four of the mountains in the CA. According to Huang et al. (2012), the rate of warming trend in the CA during the cold season (November–March) is 0.24°C decade−1, which is higher than the other semiarid regions around the world, whereas the rate during the warm season (May–September) is 0.08°C decade−1. Temperature increase was 0.65°C between 1942–72 and 1973–2003 (Martino et al. 2009), and temperature is predicted to become even higher by the end of the twenty-first century in the CA (Lioubimtseva et al. 2005; Lioubimtseva and Henebry 2009; Mannig et al. 2013; Hu et al. 2014; Aalto et al. 2017). The increasing tendency of temperature was also found in the southwest of Tajikistan (Khakimov et al. 2007; Kayumov 2010; Chevallier et al. 2014; Wang et al. 2016; Aalto et al. 2017; Gulakhmadov et al. 2020), while the slightest decrease was detected in northern Tajikistan.

Based on long-term precipitation datasets, many studies have revealed that the global land area has experienced an overall increasing trend in precipitation during the last century (Diaz et al. 1989; Hartmann et al. 2013; Gu and Adler 2015). However, precipitation change displays regional variations during the past century (Dai et al. 1997; Trenberth et al. 2007). The precipitation pattern in the CA is primarily the result of the transition between two circulation systems. The westerlies supply most of the precipitation during the winter and spring months, while the Indian summer monsoon provides summer and autumn precipitation (Nezlin et al. 2004; Holmes et al. 2009; Wang et al. 2010; Fuchs et al. 2013; Pohl et al. 2017; Gulakhmadov et al. 2020). Most of the moisture coming from the Indian Ocean is blocked by the high Pamir, Hindu Kush, Tianshan, and Himalayan mountains to the southeast of the CA (Schiemann et al. 2008). Central Asia’s precipitation, especially seasonal precipitation, is highly spatially heterogeneous. Large amount of precipitation usually occurs in the mountains open to moisture-laden jet streams from the southwest of the Pamir (Finaev et al. 2016). Annual precipitation falls fairly evenly in all seasons in the north, whereas precipitation falls mainly during winter and spring in the south, accounting for up to 80% of annual total precipitation during 1930–2009 (Chen et al. 2011).

The Amu Darya is the largest river in the CA and a substantial proportion of the annual streamflow of the Amu Darya River basin (ADRB) is from snow and glacier melt in the high Pamir and Zarafshan mountains (Froebrich and Kayumov 2004; Sidike et al. 2016; Wang et al. 2016; Armstrong et al. 2018). An increasing tendency in streamflow was recorded in the Kafarnigan River basin (a subbasin of the ADRB), during the spring, while a decreasing trend appeared during the winter and autumn season (Lobanova and Iulii 2019; Gulakhmadov et al. 2020). The GIS modeling suggest that streamflow decreased by 5.1% in the Vakhsh River and 1.8% in the Pyandj River during 1970–2009 (Finaev et al. 2016); in the upper Pyandj River, streamflow decreased by 15.5% due to reduction in precipitation during 1951–2007 (Wang et al. 2016). It has been predicted that by the end of the twenty-first century, increasing temperatures are likely to substantially affect snow cover in the upstream mountains and lead to reduction in streamflow by 20% in the ADRB (White et al. 2014). However, there are limited reports on the streamflow conditions and their relations to climate factors at annual and monthly time steps in the entire ADRB up to the most recent period.

Extreme climate indices that include both precipitation and temperature extremes have been studied in many regions across the globe (Vincent et al. 2005; Brown et al. 2008; Toreti and Desiato 2008; Donat et al. 2016; Yosef et al. 2019; Ding et al. 2018; Poudel et al. 2020). However, limited studies of up to the most recent extreme climate indices were conducted in the CA. Most of the studies in the ADRB focused on a few glacierized subbasins, while in-depth knowledge of extreme climate and hydrology in all subbasins of the ADRB are still lacking. Moreover, no study has been conducted to assess the relationships between precipitation/temperature changes and streamflow change in the upper ADRB, where streamflow originates. Hence, answers to how extreme temperature and precipitation indices have changed in the ADRB up to the most recent and which climate parameters are the crucial driving factor for streamflow change are essentially unknown but are very important for society, ecosystem, and water resources planning and management to ensure sustainable social development. Hence, the objectives of the current study are 1) to reveal up to the most recent extreme and mean climate conditions based on gridded products and in situ observations in the ADRB, 2) to investigate streamflow change in the ADRB, and 3) to determine the relationships between climatic factors and streamflow during the past several decades in the upper ADRB. This paper is organized as follows: study area and data are introduced in section 2. Section 3 presents the results and discussion regarding the means and trends, as well as the relationships between the three climate factors and streamflow in the upper ADRB. Section 4 draws the conclusions.

2. Methodology

a. Study area

The Aral Sea, ranked as the world’s fourth largest inland lake in the mid-twentieth century, was originally fed by the streamflow from the ADRB in the south and the Syr Darya River basin (SDRB) in the north. The Amu Darya River originates in the mountains of Tajikistan and Afghanistan (Pamir and Zarafshan). ADRB, the largest river in the CA, lies over 34°23′–43°23′N and 59°42′–75°05′E and contributed 70% of the water to the Aral Sea before the connection was cut off due to canals and reservoirs. Its catchment occupies 470 000 km2 (of which 2.89% or 13 600 km2 of area is covered by glaciers) with an altitude ranging from 72 m in the western to 7495 m in the eastern (Fig. 1a). The total river length is 1415 km excluding the Pyandj River (Froebrich and Kayumov 2004; Lobanova and Iulii 2019), crisscrossing through Tajikistan, Afghanistan, Turkmenistan, and Uzbekistan. The ADRB consists of five major subbasins, including Pyandj River basin (PYRB), Vakhsh River basin (VKRB), Kafarnigan River basin (KFRB), Zarafshan River basin (ZVRB), and Kunduz River basin (KZRB) (Fig. 1b).

Fig. 1.
Fig. 1.

(a) The geographical location of the Amu Darya River basin (ADRB) and its major tributaries. The green circles show selected climate stations, red circles are hydrological stations, and the thin gray lines mark national transboundaries. (b) The thick black line outlines the upper Amu Darya River basin, which is controlled by the Kerki station, including four primary subbasins: the Pyandj River basin (PYRB), Vakhsh River basin (VKRB), Kafarnigan River basin (KFRB), and Kunduz River basin (KZRB). The subbasin Zarafshan River basin (ZVRB) does not contribute to Kerki. The red circles represent selected hydrological stations in each subbasin that have the same time period records that were used to establish correlations between climate factors and streamflow.

Citation: Journal of Hydrometeorology 24, 2; 10.1175/JHM-D-22-0025.1

From 1950 to the late 1980s, irrigated area increased from 2.9 to 7.2 million hectares (ha) in the Aral Sea basin (Bedford 1996), with consequent increases in water withdrawals. Seasonally, flow regulation provides about 61.4 km3 of water for economic and agricultural use in the ADRB (Stulina and Eshchanov 2013). Turkmenistan and Uzbekistan are the main water consumers of the Amu Darya. Most notably, the Karakum Canal with its head above Kerki and 1400 km long diverts almost a third of the flow of the Amu Darya into Turkmenistan, which resulted in a sharp decline in water levels (Bedford 1996). In 1972 the Nurek Dam (a capacity of 10.5 km3), a hydropower dam, started operation along lower end of the Vakhsh River and supplied irrigation to 70 000 ha of crop land above Kerki (Adenbaev et al. 2015). As a result, the area of the Aral Sea decreased by 55 369 km2 (82.03%) from 1960 to 2011, with the lake splitting into two bodies of water in 1986 (Yang et al. 2020). Other studies, such as Li et al. (2021) and Su et al. (2021), also stated that anthropogenic activity particularly cultivated area expansion was the most important factor leading to the decreasing trend in streamflow. Li et al. (2021) using GlobeLand30 dataset at a 30-m resolution analyzed land use and cover change for the entire Aral Sea region reported that the area of water bodies in the Aral Sea decreased by 51.6%, from 26 280.8 km2 in the 2000s to 12 712.3 km2 in the 2010s. The water body further decreased by 27% in the 2010s and shrank to 9285.2 km2 in 2020. Likewise, Su et al. (2021) stated that 50.38% of the water body dried and was converted into bare area during 1992–2015. It is quite noticeable that since the 1970s, streamflow has dropped dramatically in the downstream due to water withdrawal in the middle and lower reaches and it is 4 times lower than the recorded level in 1930–40s, resulting in unstable ecological streamflow into the Aral Sea (Stulina and Eshchanov 2013). The dwindled river that no longer flows to its historical terminus has resulted in environmental and socioeconomic disasters, including desertification of the dried seabed and the collapsing of the fishing industry near the sea. To exacerbate the situation, in the upper ADRB, climate warming–induced glacier and snow retreat could result in alarming consequences on streamflow in the future, which needs to be investigated thoroughly. Thus, the changes in streamflow from all tributaries in the upper ADRB are due to climate change, whereas streamflow at and below the Kerki station is affected by human activity.

The river is highly dependent on the glacier and snow hydrological processes as nearly all of the water of the ADRB originates from the high ranges of the Pamir Mountains. The contributions to the streamflow include glacier melt (38%), snowmelt (27%), and rainfall (16%) in the high Pamir mountainous region (Agal’tseva et al. 2011; Immerzeel et al. 2012; Hagg et al. 2013). Here, we use the Kerki station as the outlet of the upper ADRB, which constitutes around 288 898 km2, or about 60% of the entire ADRB of which 4.39% (12 681 km2) is covered by glacier (Fig. 1b, Table 3). Based on in situ observations, on a seasonal scale, winter (DJF) precipitation is about 150.0 mm (31.2%), spring (MAM) precipitation is 227.6 mm (48.2%), summer (JJA) precipitation is 32.1 mm (6.82%), and fall (SON) precipitation is 62.3 mm (13.21%) in the upper ADRB. Average annual precipitation exceeds 1000 mm in the mountains, while it is only 100 mm in the foothill and its adjacent (Wang et al. 2016). About 80% of streamflow in the ADRB is concentrated from April to September, and only 13% occurs from December to February (Agal’tseva et al. 2011; Wang et al. 2016). The upper eastern high mountainous region is characterized by permanent snow and ice, while open grasslands dominate middle altitudes in the mountains, and low land downstream is dominated by desert and irrigated croplands. Most of the favorable land in the ADRB used for agriculture and economic activities such as Karshi steppe, Bukhara region, and the southern part of Turkmenistan is far from the main riverbank, where precipitation and water resources are scant.

b. Data

Glacier data were obtained from Randolph Glacier Inventory (RGI 4.0, https://www.glims.org/RGI/randolph40.html). Gridded daily temperature products from Princeton Global Forcing (PGF) (Sheffield et al. 2006) and Climate Prediction Center (CPC) (Chen et al. 2008) were used. Gridded daily APHRODITE (Yatagai et al. 2012) and CHIRPS (Funk et al. 2014) were used for precipitation analysis. Observed streamflow and both grided temperature and precipitation in the same period were used in partial correlation analysis. Temperature (PGF and CPC) and precipitation (APHRODITE and CHIRPS) gridded products were quality checked using quality-controlled station data first and merged together to extend their study period from 1951 to 2019. The merged products of gridded temperature and precipitation were homogenized based on the best quality product of temperature (PGF) and precipitation (APHRODITE) and further adjusted by elevation-dependent lapse rates. The details of the grid product processing procedures are provided in the online supplemental material. PGF, APHRODITE, and CHIRPS have a daily time step and a spatial scale of 0.25°. CPC was downscaled to 0.25° from 0.5° using the nearest neighbor method.

We have collected all available observed streamflow up to the most recent. The observed monthly streamflow in Tajikistan have been obtained from the Hydrometeorological Agency of the Republic of Tajikistan. Fifteen hydrological stations with monthly streamflow in the ADRB in northern Afghanistan have been taken from http://aizon.org/watershed_atlas.htm (last access in November 2018), and two downstream stations (Kerki and Chatly) with daily streamflow in the lower and middle reaches of the ADRB have been obtained from Global Runoff Data Centre (GRDC, https://www.bafg.de/GRDC/EN/01_GRDC/11_rtnle/rationale_node.html). Hydrological observation deteriorated due to civil war and instability in the region after the collapse of the Soviet Union in central Asia. Observation periods of all the selected stations are different, and some observations have long gaps. Because of the limitation in observed streamflow data, we are not able to obtain stations that are in exactly the same recent 30-yr period for climatology analysis. Given the difficulty in accessing and acquiring in situ hydrological data, we were guided by the principle of utilizing as many observation data as possible permitted by data quality in order to understand hydrological conditions in the ADRB because any good quality data are better than no data. Further, that fact that there is no consistent time period that is not up to the most recent data does not mean we cannot study climate. We still need to understand the historical conditions and changes when drastic changes occur in an area that is not well understood and is suffering some ecosystem crisis due to water issue. Based on the above rationale, the records with more than 10 years were used for analysis. In fact, different time periods at individual stations were used to study extreme temperature and precipitation events for global land area by the sixth IPCC report (e.g., Seneviratne et al. 2021) and Sun et al. (2021). Table 1 shows detailed information of the selected hydrological stations. All stations in the upper ADRB are free from human intervention, whereas streamflow measured at Kerki and Chatly was heavily affected by human activity. Because of time period difference, we do not intend to compare the streamflow amount and its change in the ADRB. Instead, we show the hydrological conditions as the data show during their specific periods.

Table 1

Detailed information of selected hydrological stations.

Table 1

In this study, 17 (8 temperature and 9 precipitation indices, Table 2) out of 27 extreme climate indices defined by the Expert Team on Climate Change Detection and Indices (ETCCDI), together with daily temperature maxima and minima, beginning frozen day, end frozen day, number of frozen day and annual total precipitation (23 indices in total) were chosen to assess the means and trends of extreme climate and frozen conditions in the ADRB in the CA. Daily temperature maxima and minima and precipitation were used to calculate the 17 ETCCDI indices using the user-friendly software package RClimDex, which is available for free from the WMO/CLIVAR/JCOMM ETCCDI website (http://cccma.seos.uvic.ca/ETCCDI) and runs in the R programming environment.

Table 2

Definitions of ETCCDI, Tmin, Tmax, annual total precipitation, BFD, EFD, and NFD used in this study.

Table 2

The quality of observed streamflow was checked and controlled at monthly and annual time steps when the time steps were available by the following procedures. If more than 2 months of data were missing in a year, the whole year was treated as missing. For hydrological data, the gaps are filled by using the mean values of the two adjacent proceeding and following months. To fill the missing values of the target station, we used precipitation from the surrounding stations that are close to each other and share similar climate characteristics as reference. The standard ratio scaling method which requires three close stations was applied to fill the gaps. Annual values were averaged (for temperature and streamflow) or aggregated (for precipitation) from monthly or daily time series. Due to space limit, here we report climate conditions and changes at annual time step, whereas streamflow and the relations between streamflow and climate factors are reported at both annual and monthly time steps.

Three indices in a frozen year, i.e., the period from 1 July of current calendar year to 30 June of next calendar year, were defined for representing the entire freeze–thaw cycle in the ADRB: beginning frozen day (BFD), ending frozen day (EFD), and number of frozen days (NFD). BFD is defined as the first day of first 3 consecutive days when air temperature is less than or equal to 0°C. EFD is defined as the last day of last 3 consecutive days when temperature is less than or equal to 0°C. NFD is number of days when temperature is less than or equal to 0°C between BFD and EFD with the days when temperature greater than 0°C excluded. NFD represents actual number of frozen days. For the ease of computation, all of the dates are expressed as the number of days in a frozen year, e.g., 1 July of 1951 is day 1 of frozen year 1951, 2 July of 1951 is day 2 of frozen year 1951, 30 June of 1952 is day 365 of frozen year 1951, etc.

c. Analysis

1) Mann–Kendall test and Sen’s estimator

The trends of extreme climate indices and streamflow were calculated using Sen’s slope (Sen 1968) and their significance were tested using the nonparametric Mann–Kendall (hereafter MK; Mann 1945; Kendall 1975). The MK test is nonparametric; hence, there is no requirement of the normality of the distribution of the samples, thereby less sensitive to extreme sample values, and is independent of the hypothesis about the nature of the trend, whether linear or not. The advantage of the test is its low sensitivity to abrupt breaks due to inhomogeneous time series (Jaagus 2006). It tests whether to reject the null hypothesis (H0) or accept the alternative hypothesis (Ha), where H0 means no significant trend is present and Ha means monotonically increasing or decreasing trend is present (Mann 1945; Kendall 1975). Sen’s nonparametric estimator developed by (Sen 1968) is used to estimate the true existing slope in n pairs of the data. A moving t test is used to detect the abrupt changing points. The calculations of the MK test, Sen’s estimator, and moving t test are in the supplemental material.

2) Partial correlations

Maximum and minimum temperatures and precipitation are all correlated with streamflow because of the nature of the water cycle in a cold and dry environment. However, all these variables are in complex relationships confounded by their coexistence and are also affected by the locations of the subbasins. Partial correlation can quantify the correlation between two variables when conditioned on other variables, meaning that the confounding effects of other variables on the two-variable relationship can be removed. Partial correlation analysis method was used to determine the response of streamflow to precipitation and temperature changes at annual and 0–4-month lags. Correlations at 0–4-month lags could reflect the fact that streamflow should have a delayed response to temperature and precipitation where annual precipitation concentrates in winter and spring that melts when weather becomes warm several months later and contributes to streamflow.

To carry out partial correlation analysis, we used gridded monthly and annual Tmax, Tmin, precipitation, and observed streamflow in the same period for primary subbasins and the entire upper ADRB which is controlled by Kerki hydrological station. Hydrological stations with more than 10 years of consecutive observations in each subbasin were selected (see streamflow stations in Fig. 1b). Except Kerki, most hydrological stations are in remote locations with limited human activity. The time series of spatially averaged Tmax, Tmin, and precipitation and the sum of streamflow of stations in each subbasin were used in the partial correlation analysis. Partial correlation coefficient calculation is provided in the supplemental material. Streamflow change above Kerki is predominantly affected by climate change, whereas streamflow changes at the Kerki and Chatly stations at the ADRB outlet are the results of the combination of climate change and human activity.

3. Results and discussion

a. Temperature

1) Mean annual temperature indices

Based on gridded temperature, the average annual of TNn, TXn, TN10p, TX10p, TXx, TNx, TN90p, TX90p, Tmax, and Tmin (see definitions in Table 2) were analyzed for the entire ADRB for the 1951–2019 period (Fig. 2). The spatial distribution of extreme mean annual temperature indices displayed similarity, that is, high temperatures existed in the middle and lower reaches, while the eastern high mountainous area endured low temperatures. TXx, TXn, TNx, TNn, Tmax, and Tmin range from 27.6° to 46.6°C, from −25.5° to 0.9°C, from 10.1° to 29.6°C, from −42.7° to −13.2°C, from −3.7° to 25.4°C, and from −8.1° to 11.7°C, respectively (Figs. 2a–f).

Fig. 2.
Fig. 2.

The spatial distribution of (a) mean annual maximum value of daily maximum temperature (TXx), (b) mean annual maximum value of daily minimum temperature (TXn), (c) mean annual minimum value of daily maximum temperature (TNx), (d) mean annual minimum value of daily minimum temperature (TNn), (e) mean annual daily maximum temperature Tmax, (f) mean annual daily minimum temperature Tmin, (g) percentage of time when daily maximum temperature < 10th percentile (TX10p), (h) percentage of time when daily maximum temperature > 90th percentile (TX90p), (i) percentage of time when daily minimum temperature < 10th percentile (TN10p), and (j) percentage of time when daily minimum temperature > 90th percentile (TN90p) during 1951–2019. Note, the upper ADRB above Kerki is outlined with a black line. The same notation applies in the following figures.

Citation: Journal of Hydrometeorology 24, 2; 10.1175/JHM-D-22-0025.1

Annually, TX10p was not uniformly distributed in the basin, with low values in the eastern and central basin and high values in the west, which is different from TX90p spatial pattern. Annual TN10p is high in the north and east, indicating colder conditions than in the south and west. Annual TN90p is essentially opposite to TN10p, high values in the west and low values in the mountains (Figs. 2g–j). In general, these temperature indices follow the temperature change with elevation; high mountains have colder weather, and vice versa.

2) Trends in extreme temperature indices

Mann–Kendall trends of gridded observed annual temperature extreme indices across the ADRB are shown in Fig. 3. For TXx, statistically significant positive trends were located mostly in the western areas. The eastern part of the domain including eastern PYRB and KZRB exhibited insignificant decreasing trend. In all, 81.4% of the area showed positive trend in annual TXx with domain averaged statistically significant increasing trend of 0.007°C yr−1 (Figs. 3a, 4a). For TXn, 67.9% of the area exhibited positive trend and statistically significant increasing trends were located in the eastern high mountains of the VKRB and PYRB with basin averaged statistically significant increasing trend of 0.01°C yr−1, whereas the northwest and the KZRB in the south experienced an insignificant negative trend (Figs. 3b, 4b). TNx displayed increasing trends in 94.4% area, and its statistically significant positive trends were mostly in eastern high mountains and northwestern domain, with area averaged statistically significant increasing trend of 0.023°C yr−1 (Figs. 3c, 4c). The decreasing TNn trends were located in the west, accounting for 53.1% of the domain, with basin averaged insignificant decreasing trend of −0.008°C yr−1 (Figs. 3d, 4d). The Tmax and Tmin displayed increasing trends in 97.8% and 99.1% areas, respectively. Statistically significant positive trends were observed in the majority of basins for Tmax. However, the increase in Tmin was stronger than that in Tmax with the rates of 0.026° and 0.019°C yr−1, respectively (Figs. 3e,f, 4e,f).

Fig. 3.
Fig. 3.

Annual trends of (a) TXx, (b) TXn, (c) TNx, (d) TNn, (e) Tmax, (f) Tmin, (g) TX10p, (h) TX90p, (i) TN10p, and (j) TN90p during 1951–2019. Points represent significant trends at α = 0.05.

Citation: Journal of Hydrometeorology 24, 2; 10.1175/JHM-D-22-0025.1

Fig. 4.
Fig. 4.

Annual trends of ADRB averaged of (a) a TXx, (b) TXn, (c) TNx, (d) TNn, (e) Tmax, (f) Tmin, (g) TX10p, (h) TX90p, (i) TN10p, and (j) TN90p during 1951–2019 based on gridded data. The Z value represents significance of the time series.

Citation: Journal of Hydrometeorology 24, 2; 10.1175/JHM-D-22-0025.1

Annual TX10p (cold day) and TN10p (cold night) had generally decreasing trends over the domain, with 99.4% and 97.3% areas showing reductions, respectively. Statistically significant decreasing trends covered the majority of basin, accounting for 79.4% and 85.4%, respectively (Figs. 3g,i). Basin averaged annual TX10p and TN10p decreased significantly at −0.14% and −0.183% yr−1, respectively (Figs. 4g,i), with TX10p statistically significant. On the contrary, annual TX90p (warm day) and TN90p (warm night) displayed increasing trends over the basin, with significantly large increasing trends all over the basin for TN90p. About 97.9% and 99.9% of the domain experienced increasing trends for TX90p and TN90p, respectively. Significant trends accounted for about 66.0% and 99.6% for TX90p and TN90p, respectively (Figs. 3h,j). Nonsignificant increasing trends were noted around the central part to the south with even decreasing trends in KZRB for TX90p (Fig. 3h). Basin averaged TX90p and TN90p increased at 0.106% and 0.280% yr−1, respectively (Figs. 4h,j). The analysis clearly shows a reduction in the number of cold days and nights but an increase in warm days and nights. The trends on cool and warm days appeared to be smaller than those on cool and warm nights. Cool days and nights showed increasing trends before the 1970s, but sharp decreases afterward, which is in agreement with the results from Klein Tank et al. (2006).

3) Climatology and changes of freeze–thaw cycles

The annual mean BFD in the basin ranged from 123 (31 October) in the mountainous northeast to 169 (16 December) in the dry and low-elevation southwest, with a regional average of 150 (27 November, Figs. 5a, 6a). Approximately, 82% of gridded cells showed positive BFD trends of which only 3% were statistically significant at p < 0.05 (Fig. 6b). A basin-averaged insignificant positive BFD trend of 0.095 day yr−1 was seen during 1951–2019 (Fig. 9a). The mean annual EFD ranged from 189 (5 January) in the central-southern part to 234 (19 February) in the eastern high mountains, with a regional average of 215 (1 February) (Figs. 5c, 6b). Early EFD occurred primarily in the source of the Kunduz River with low elevation (Fig. 5c). Approximately, 61.5% of gridded cells showed increasing EFD trends of which 28% was statistically significant at p < 0.05) (Fig. 5d). Significant negative EFD trends occurred in the west high mountains, with a basin-averaged insignificant trend of −0.118 day yr−1 seen during 1951–2019 (Fig. 6b). The mean annual NFD ranged from 1.2 in the southwest above the Kerki station to 183 at the source of the Vakhsh River, with a regional average of 92 (Fig. 5e). Small NFD occurred primarily in the southwest, especially the source of the Kunduz River basin in low elevations, while great NFD was found in the upper high mountainous areas (Fig. 5e). About 79% of the basins showed decreasing NFD trends of which 44% was statistically significant occurring in the eastern high mountains and the western desert. A basin-averaged significant trend at −0.171 day yr−1 was found during 1951–2019 (Fig. 6c). The changes in BFD, EFD, and NFD clearly show that the frozen period is shortening in the basin, especially in the upper ADRB.

Fig. 5.
Fig. 5.

Spatial and temporal characteristics of annual (a),(b) beginning frozen day (BFD), (c),(d) ending frozen day (EFD), and (e),(f) number of frozen days (NFD). (left) Mean of the annual time series and (right) trend of the annual time series. Dots in the right column indicate that the trends are significant at p > 0.05.

Citation: Journal of Hydrometeorology 24, 2; 10.1175/JHM-D-22-0025.1

Fig. 6.
Fig. 6.

Basin averaged means and trends of (a) BFD, (b) EFD, and (c) NFD during July 1951 and June 2019. Note, an asterisk indicates that the slope is significant at p > 0.05.

Citation: Journal of Hydrometeorology 24, 2; 10.1175/JHM-D-22-0025.1

b. Precipitation

1) Mean annual precipitation indices

The driest portion of the ADRB was the eastern mountains and northwestern plains. The scale difference between annual total precipitation (Fig. 7a) and PRCPTOT (Fig. 7b) was large indicating that the region had frequent days with daily precipitation ≤ 1 mm which were omitted by PRCPTOT. The highest values were in the north, center, and south of the upper ADRB, while the eastern and western regions had low precipitation. Mean annual total precipitation and PRCPTOT were around 10–300 mm in the west and east, essentially desert regions, whereas the north-center part had more than 500 mm. Mean CDD shows the opposite spatial pattern to that of annual total precipitation, and PRCPTOT and CWD were high in the west and east but low in the north, center, and south of the upper ADRB. Mean annual R10mm was mostly more than 20 days in the wet area and fewer than 10 days in the west and east deserts. Mean annual R20mm was mostly more than 15 days in the wet area but fewer than 3 days in the deserts. The spatial patterns of R95p, R99p, Rx1day, and Rx5day demonstrated high values in the central ADRB but low values in the rest of the basin. The mean annual R95p and (R99p) had the highest value of 150 mm and (40 mm) primarily in the central basin, but less than 50 mm and (10 mm) in the east and west areas. Likewise, high values of more than 30 mm (60 mm) of mean annual RX1day and (RX5day) were found in the central basin, while values less than 15 mm and (30 mm) occurred in the eastern and western regions.

Fig. 7.
Fig. 7.

The spatial distributions of mean annual precipitation indices, (a) annual total precipitation, (b) annual total precipitation from days ≥ 1 mm (PRCPTOT), (c) consecutive dry days (CDD), (d) consecutive wet days (CWD), (e) annual count of days when precipitation ≥ 10 mm (R10mm), (f) annual count of days when precipitation ≥ 20 mm (R20mm), (g) annual total precipitation from days > 95th percentile (R95p), (h) annual total precipitation from days > 99th percentile (R99p), (i) annual maximum 1-day precipitation (RX 1day), and (j) annual consecutive 5-day precipitation (RX 5day).

Citation: Journal of Hydrometeorology 24, 2; 10.1175/JHM-D-22-0025.1

2) Trends in extreme precipitation indices

About 67.55%, 69.77%, 74.66%, 59.33%, 52.86%, and 53.61% of grid cells showed increasing trends for annual total precipitation, PRCPTOT, CWD, R95p, RX1day, and RX5day, with basin-averaged increasing trends of 0.426, 0.431, 0.004, 0.097, 0.005, and 0.060 mm yr−1, respectively, and none of the averaged trends were significant at α = 0.05 (Figs. 8, 9a,b,d,g,h,i,j). On the contrary, about 60.45% and 55.56% of the basin had decreasing trends in annual CDD and R10mm with a basin average of −0.087 days yr−1 and −0.002 mm yr−1, respectively. There were 88.66% and 94.88% grid cells without changing trends for R20mm and R99p, which were mainly distributed in the middle and lower reaches of the ADRB (Figs. 8f,h). Generally, the wet part of the UADRB became drier while the other part of the UADRB became wetter during the study period.

Fig. 8.
Fig. 8.

The grid based and station observed trends of (a) annual total precipitation, (b) PRCPTOT, (c) CDD, (d) CWD, (e) R10mm, (f) R20mm, (g) R95p, (h) R99p, (i) RX 1day, and (j) RX 5day. Dots represent significant trends at α = 0.05.

Citation: Journal of Hydrometeorology 24, 2; 10.1175/JHM-D-22-0025.1

Fig. 9.
Fig. 9.

Basin averaged (a) annual total precipitation, (b) PRCPTOT, (c) CDD, (d) CWD, (e) R10mm, (f) R20mm, (g) R95p, (h) R99p, (i) RX 1day, and (j) RX 5day.

Citation: Journal of Hydrometeorology 24, 2; 10.1175/JHM-D-22-0025.1

c. Streamflow

Again, we stress that due to the time period difference among the hydrological stations, means, and trends of streamflow of these stations cannot be compared. Only the actual values are described as they appear in order to understand hydrological conditions. Mean annual streamflow was 2.9 m3 s−1 at Sayad in the Kulm River in the north of Afghanistan and was 1627.3 m3 s−1 at Kerki in the middle reach of the ADRB (Fig. 10a). The annual streamflows at a majority of stations in the UADRB above Kerki were less than 300.0 m3 s−1. The annual trend ranged from a significant trend of −25.111 m3 s−1 at Kerki to 1.438 m3 s−1 at Sayad in the Kunduz River in the northern Afghanistan (Fig. 10b). Out of 30 hydrological stations, 12 (18) stations streamflow showed positive (negative) annual trends out of which 2 (3) were statistically significant at the α = 0.05 level. A majority of stations (3 out of 4) in the Vakhsh subbasin showed small positive trends (VKRB, Fig. 1b) where glaciers cover 16.5% of area (Table 3), whereas 4 out of 5 stations with decreasing trends occurred in the Pyandj subbasin, the largest subbasin where glaciers cover 8.0% of area (PYRB, Figs. 1b, 10b).

Fig. 10.
Fig. 10.

Spatial patterns of (a) mean annual discharge and (b) annual discharge trend (m3 s−1 yr−1). Stars represent statistically significant trends at α = 0.05.

Citation: Journal of Hydrometeorology 24, 2; 10.1175/JHM-D-22-0025.1

Table 3

Partial correlations between annual maximum, minimum temperature, precipitation, and streamflow in five subbasins and the entire upper Amu Darya River basin (UADRB); Q: discharge, P: precipitation. The bold number is statistically significant at the α = 0.05 level.

Table 3

The monthly streamflow is characterized by strong seasonality with high flow occurring in late spring and summer from May to September, reaching the highest in July, clearly showing a delayed response to precipitation, while the lowest value was observed from December to February (Fig. S1 in the online supplemental material). Mean monthly flow ranged from 1.8 m3 s−1 in January at Sayad in the Khulm River to 3533.4 m3 s−1 in July at Kerki in the main branch. Streamflow trends ranged from −49.101 m3 s−1 in August at Kerki to 9.083 m3 s−1 in April in the Kunduz River at Khojagar in northern Afghanistan (Fig. S2). A large portion of stations had positive trends in the UADRB, mostly occurring from December to May and reaching their maximum number of stations with positive trend in April (23 out of 30 stations). Negative trends mostly occurred during the dry season (summer and early fall), reaching the maximum number of stations in August (20 out of 30 stations) (Fig. S2). Great and significant decreasing trends occurred from July to October at Kerki, and from August to October at Chatly.

Streamflow changes at the Kerki and Chatly stations were the result of both climate change and intense human activity. Given that at both annual and monthly scales, streamflow showed statistically significant (α = 0.05) strong decreasing trends at Chatly and Kerki (Fig. S2, Figs. 10b, 11a,b), which had relative long periods, and thus these two stations were selected for abrupt changepoint detection that may show when intensive human activity started to disrupt the natural variation of streamflow at the two stations. The moving t test revealed that the break points appeared in 1953 and 1955 at Chatly and in 1973 at Kerki when t values exceeded a critical t value at α = 0.05 for both steps 10 and 12 consistently (Figs. 11c,d).

Fig. 11.
Fig. 11.

(a),(b) Annual time series and trend and (c),(d) moving t test at hydrological stations Kerki and Chatly.

Citation: Journal of Hydrometeorology 24, 2; 10.1175/JHM-D-22-0025.1

The detected abrupt changing points and observed decreasing trend in the downstream were consistent with the conclusions by Stulina and Eshchanov (2013). Since the 1960s, water abstraction from the Amu Darya River into irrigation channels or reservoir operation were the main reasons that led to a decreasing trend in streamflow in the downstream. Annually, on average, 13.5 km3 of water has been withdrawn from the Amu Darya into the Karakum Chanel in lower basin since it was constructed in Turkmenistan in the 1950s (Zonn 2014), which explains why Chatly had an abrupt changepoint around 1953–55. In 1972 the Nurek Dam (a capacity of 10.5 km3) started operation along the lower end of the Vakhsh River that connects to the main branch and supplied irrigation to 70 000 ha of crop land above Kerki (Adenbaev et al. 2015), which could explain the abrupt changepoint at Kerki. Furthermore, downstream of the Amu Darya is Tayamuyun where human activities, e.g., agriculture, industries, urbanization, ecosystem conservation, and reservoir operations are intense, thus drastically affecting natural streamflow (Stulina and Eshchanov 2013). In the UADRB, however, Wang et al. (2016) stated that streamflow exhibited a significant negative trend in PYRB, VKRB, and KFRB but a slight negative trend was detected in KZRB, which are different from this study, while a small positive trend was detected in ZVRB, which is in agreement with our study. The difference is primarily due to the different stations and study periods examined.

d. Relationships between climate factors and streamflow

Based on monthly climate gridded products and observed streamflow, correlations between climate factors and streamflow were examined at annual and 0–4-month lagged time scales using the partial correlation method for five primary upper subbasins including PYRB, VKRB, KFRB, KZRB, and ZVRB, and the Kerki-controlled UADRB (Fig. 1b). Climate factors were Tmax, Tmin, and precipitation. At an annual scale, positive correlations between Tmax and streamflow were seen in PYRB, VKRB, and KZRB, whereas negative correlations existed in KFRB and ZVRB (Table 3). The Tmin was positively correlated with streamflow in PYRB, VKRB, and ZVRB, which all contain more than 6.3% glacier area, but negatively correlated with streamflow in KFRB and KZRB, which contain less than 2.5% glacier area (Table 3). Correlation coefficients were small except in KZRB, and all were not significant. There were positive correlations between annual precipitation and streamflow in all subbasins, and only statistically significant in KFRB and UADRB. For the entire UADRB, annual precipitation had a stronger influence on annual streamflow than Tmax and Tmin did. Given that streamflow at Kerki has been affected anthropogenically to some extent, the correlations could have been stronger without human influence.

Here streamflow lagged 0–4 months behind precipitation and Tmax and Tmin were correlated with 0-month precipitation and Tmax and Tmin. For example, January precipitation/temperatures were correlated with January (lag 0), February (lag 1), March (lag 2), April (lag 3), and May (lag 4) streamflow. The same procedure applies to the other months as well. Here, results in VKSH with 16.5% glacier and average elevation of 3925 m and KFRB with 2.1% glacier and average elevation of 2291 m are presented in Figs. 12a–f. Results in the other basins are in the supplemental material and Fig. S3.

Fig. 12.
Fig. 12.

Correlation coefficients between monthly climate factors and 0–4-month lagged streamflow in VKRB and KFRB of the upper Amu Darya Basin. Hatched bars mean significance at the α = 0.05 level.

Citation: Journal of Hydrometeorology 24, 2; 10.1175/JHM-D-22-0025.1

Correlations between Tmax and 0–4-month-lagged streamflow in VKRB are shown in Fig. 12a. There were five significant negative correlation coefficients in April–August which were lag 2, lag 3, and lag 4. This indicates that April and June Tmax negatively affected July and September streamflow, respectively; May Tmax affected July streamflow; and July and August Tmax influenced streamflow 4 months later (Fig. 12a). This relationship demonstrated months-long lasting negative effects of warm season daytime peak temperature on streamflow which might have to do with intensified evapotranspiration during daytime overwhelming melt contribution. Nonsignificant mixed correlations are found in autumn, winter, and early spring. On the other hand, Tmin was positively correlated to lagged streamflow in most months (Fig. 12b). In fact, Tmin in one month could affect several lagged streamflows. For example, Tmin in June significantly influenced concurrent and lag-3 (September) streamflow; Tmin in July significantly affected lag-1, lag-3, lag-4 streamflow; Tmin in August significantly affected concurrent, lag-2, and lag-4 streamflow; and Tmin in September significantly affected lag-1 and lag-3 streamflow. The analysis clearly shows that summer Tmin increase significantly benefited concurrent and late season’s streamflow (Fig. 12b). Spring Tmin significantly positively affected summer streamflow. This correlation clearly demonstrates the positive control of Tmin on melt contribution. Usually, Tmin exists before dawn when evapotranspiration is also low but melt could be intense in a basin with large coverage of snow and glacier. High Tmin thus facilitates melt, which contributes to streamflow. The relationships between precipitation and streamflow are mixed and insignificant mostly in the first half of the year. Significant correlations with lag-2 and lag-4 streamflow occurred in May only. In the later part of the year, most correlations turned negative, which were all small and insignificant (Fig. 12c). During June and July, there was essentially too little precipitation to affect streamflow. Obviously, among the three climate factors, Tmin was the most important impactor in VKRB where 16.5% of the area is covered by glaciers (Table 3).

In KFRB, Tmax and streamflow displayed mostly mixed and nonsignificant correlations across 0–4-month lag spectrum except November and December. April Tmax and streamflow had significant positive correlation, whereas November and December had significant negative correlations with lag-0, lag-1, and lag-2 streamflow (Fig. 12d). For Tmin and streamflow, most correlations were nonsignificant, except for lag-3 and lag-4 streamflow in September, which were negative, lag-1 streamflow in November and lag-1, lag-2, and lag-3 streamflow in December, which were positive (Fig. 12e). There were 10 significant positive correlations between precipitation streamflow during spring and summer (Fig. 12f). The fact that precipitation positively correlated to streamflow demonstrated that at a lower elevation basin and with very limited glacier compared to VKRB and PYRB, rainfall and snowfall both contributed to streamflow directly in both concurrent and lagged ways in KFRB. Similar relationships were also identified in the other four subbasins (Fig. S3). The 0–4-month lagged correlations shed light on the impacts of climate factors on streamflow, which were weak at annual time step correlations in the UADRB.

Hydrological processes are sensitive to climate change especially in cold and dry regions. Three processes are functioning here: 1) positive contribution to streamflow from glacier and snowmelt, which benefits from increasing temperature; 2) negative contribution to streamflow from evapotranspiration consumption, which also increases with temperature and precipitation; and 3) precipitation impact overwhelming temperature impacts. Essentially, the trade-off between meltwater and evapotranspiration determines streamflow in upland basins, which are controlled by different temperatures. Partial correlation analysis indirectly demonstrates that in the glacier basins (e.g., VKRB and ZVRB), Tmin benefits streamflow by enhancing melt manifested at both monthly and annual time steps, whereas Tmax reduces streamflow by boosting evapotranspiration especially at monthly time step. Using averaged daily air temperature in such analysis could smooth out the opposite effects and thus provides confounding results in such basins. This finding was not reported before. Whereas, in lowland basins with little glacier accumulation, the balance of evapotranspiration and precipitation determines streamflow, where Tmax and Tmin behave differently from highland basins. In lowland basins with little glacier coverage but large amounts of snow, Tmax contributes to snowmelt and its impact on streamflow is positive (e.g., Tmax in KZRB in Fig. S3) but Tmin has a negative influence. Further, when the snow amount is limited in lowland basins, rainfall could largely determine how much water is available for streamflow and temperature impacts are less significant (e.g., KFRB). Furthermore, with a warming climate, more precipitation could fall as rainfall rather than snowfall in spring, thus precipitation could even influence streamflow directly in the early portion of the year in the future in highland basins.

The difference between this study and the study of Wang et al. (2016) is that this study has examined the spatiotemporal patterns of annual extreme temperature and precipitation in the ADRB by using 23 extreme temperature and precipitation indices and frozen period indices, and annual and monthly streamflow in the entire ADRB. Also, this study established the relationships between climate factors (Tmax, Tmin, and precipitation) and streamflow at 0–4 lagged months in five major subbasins (PYRB, VKRB, KFRB, ZVRB, and KZRB) and the entire UADRB and found that streamflow responses to Tmax, Tmin, and precipitation were basin specific, and there is indeed complex long-lasting climate influence on streamflow at the monthly time scale that could not be revealed at the annual time scale. In glacier- and snowmelt-dominated basins (PYRB, VKRB, and ZVRB), temperature (Tmax, Tmin) impacts are prominent, whereas in rainfall- and snowfall-dominated basins (KFRB and KZRB) positive precipitation effects are more important, which is different from Wang et al. (2016), who stated that warming temperature had much less effect than reducing precipitation on streamflow decrease. This study is the first step to diagnose the hydroclimate conditions in the ADRB.

The hydrological modeling results from Hou et al. (2023) showed that in the high-elevation subbasins, such as the Vakhsh, Pyandj, and Zarafshan, snow and glaciers together contributed around 80% of the annual streamflow. Whereas, in low-elevation subbasins such as Kafarnigan and Kunduz, rainfall contributed more than 50%, and snow accounted for another 40% of the annual streamflow. These streamflow compositions from the VIC simulation provide evidence to support our partial correlation analysis results, in that, in high-elevation basins, temperature impacts dominate because of great amount of meltwater contribution, whereas in low-elevation basins, precipitation effects are more important than temperature.

Besides the updated results in the conditions and changes of extreme climate indices and streamflow in the ADRB, the new findings from this study also include 1) that not only do climate factors have lagged effects on streamflow at the monthly time step, but also the lag effects are complex and basin specific, and 2) there is a difference in the roles of Tmax and Tmin in affecting streamflow in basins with different amount of glacier coverage, which was not reported before. In high glacier-covered basins, Tmin positively affects streamflow, whereas Tmax negatively affects streamflow. On the other hand, in lower basins with much less glacier coverage, the importance of Tmin is diminished. The Tmax and precipitation have a very important influence on streamflow. The revelation of the different effects of Tmax and Tmin on streamflow is especially relevant to the undergoing global warming because this study and many others found asymmetric warming in Tmax and Tmin (Karl et al. 1993; Ding et al. 2018; Poudel et al. 2020), which will have implications on streamflow change. Unfortunately, the different influences of Tmax and Tmin on streamflow currently cannot be quantified due to the structure of most hydrological models in that average daily or hourly temperature (not Tmax and Tmin specifically) is used to drive snow or glacier simulations in most cases. To better understand the roles of Tmax and Tmin on streamflow, model structures need to be improved in the future.

4. Conclusions

Streamflow in the Amu Darya Basin has been decreasing since the 1960s. The decrease in streamflow and the shrinking Aral Sea sounded an alarm and resulted in strict water use policies in the downstream arid region that strongly depends on the Amu Darya. Because of the change in streamflow, it is imperative to understand how the climate has changed and how it alone has influenced streamflow in the upper basins where water resources are produced and human activities are relatively limited. Based on the merged PGF and CPC products, the spatial patterns of all annual extreme temperature indices were very similar. In general, means of all temperature indices showed elevation dependence where higher mountains were colder, and vice versa. There were positive annual trends in all warm temperature indices including TXx, TNx, TN90p, and TX90p for the entire basin in about 95% of gridded cells except TXx, but mixed changes in cold temperature indices including TXn, TNn, TN10p, and TX10p. Upper subbasins showed positive trends in most temperature indices but negative trends in cool days and cool nights, indicating prevailing warming in the mountains. Basin-averaged annual trends showed the strongest increase in TN90p, but insignificant trends in TNn and TXn.

Generally, the spatial patterns of annual precipitation indices such as annual total precipitation, PRCPTOT, CWD, R10mm, R20mm, R95p, R99p, Rx1day, and Rx5day suggested that the north, which is the source of KFRB, was the wettest region, while the east and west had less precipitation. A majority of the basin became wetter, whereas the wet area in the basin became drier during 1951–2019. Combining the spatial and temporal patterns of both temperature and precipitation, it is clear that the wet portion of upper ADRB became warmer and drier, whereas the other portion of the ADRB became warmer and wetter.

Streamflow displayed decreasing trends during various time periods at 18 stations and increasing trends at 12 stations. The majority of stations exhibited increasing trends in the VKRB, KFRB, and ZVRB. However, decreasing trends were detected at a large proportion of stations in the PYRB and KZRB. Also, sharp decreasing trends occurred at Kerki in the middle stream and downstream at Chatly. Abrupt changepoints in 1953 and 1955 at Chatly, and in 1973 at Kerki were detected due to human activity. Monthly streamflow increased in the northern basin during winter and early spring.

The influences of climatic factors on streamflow were generally weak at the annual time step but distinct and complex at the monthly time step with lagged effects. Climate factors affected streamflow differently in the upland and lowland basins. The Tmax and Tmin impacted streamflow in opposite ways in highland basins (e.g., VKRB and ZVRB), which have high glacier and snow coverage, with Tmin (Tmax) having had predominantly positive (negative) impacts on streamflow that were seldom reported previously, while precipitation influence was mixed. Whereas in lowland basins (e.g., KZRB and KFRB) with little glacier coverage, Tmax and Tmin behaved differently from the upland basins, and precipitation and streamflow were positively correlated predominantly. This implies that in glacier- and snowmelt-dominated basins (VKRB and ZVRB), temperature (Tmax and Tmin) impacts were prominent, whereas in rainfall- and snowfall-dominated basins (KFRB and KZRB) the precipitation effect was more important. Also, the lagged effect in the region is caused by high winter and spring precipitation brought by the westerlies, which mismatches the intra-annual temperature cycle, and the lagged time period varied among basins. Water resources management needs to pay close attention to the combination of the temperature and precipitation in the several months preceding streamflow.

Acknowledgments.

This work was supported by the Strategic Priority Research Program of the Chinese Academy of Sciences (XDA20060202) and Second Tibetan Plateau Scientific Expedition and Research Program (STEP, 2019QZKK0203).

Data availability statement.

Hydrological stations with monthly streamflow in the ADRB in northern Afghanistan were obtained from Global Runoff Data Centre, http://aizon.org/watershed_atlas.htm and https://www.bafg.de/GRDC/EN/01_GRDC/11_rtnle/rationale_node.html). Daily maximum and minimum temperature and precipitation for the different periods at 34 weather stations were obtained from the Global Historical Climatological Network (GHCN), https://www.ncdc.noaa.gov/ghcn-daily-description. Glacier data were obtained from the Randolph Glacier Inventory (RGI 4.0, https://www.glims.org/RGI/randolph40.html). Daily temperature maxima and minima and precipitation were used to calculate the 17 ETCCDI indices using the user-friendly software package RClimDex, which is available for free from the WMO/CLIVAR/JCOMM ETCCDI website (http://cccma.seos.uvic.ca/ETCCDI) and runs in the R programming environment. Detailed information about data availability and processing are given in section 2b. Except for six temperature stations data and streamflow data from the Hydrometeorological Agency of Tajikistan that were purchased under a nondisclosure agreement, the other data are freely available from respective websites listed in the text and here.

REFERENCES

  • Aalto, J., M. Kämäräinen, M. Shodmonov, N. Rajabovc, and A. Venäläinena, 2017: Features of Tajikistan’s past and future climate. Int. J. Climatol., 37, 49494961, https://doi.org/10.1002/joc.5135.

    • Search Google Scholar
    • Export Citation
  • Adenbaev, Y. B., F. H. Hikmatov, and O. A. Kxaydarova, 2015: Hydrological regime of the lower reaches of Amudarja River in the conditions of intensive economical use of water resources. Eur. Sci. Rev., 910, 18–21.

    • Search Google Scholar
    • Export Citation
  • Agal’tseva, N. A., M. V. Bolgov, T. Yu. Spektorman, M. D. Trubetskova, and V. E. Chub, 2011: Estimating hydrological characteristics in the Amu Darya River basin under climate change conditions (in Russian). Russ. Meteor. Hydrol., 36, 681689, https://doi.org/10.3103/S1068373911100062.

    • Search Google Scholar
    • Export Citation
  • Alexander, L. V., and Coauthors, 2006: Global observed changes in daily climate extremes of temperature and precipitation. J. Geophys. Res., 111, D05109, https://doi.org/10.1029/2005JD006290.

    • Search Google Scholar
    • Export Citation
  • Arias, P., and Coauthors, 2021: Technical summary. Climate Change 2021: The Physical Science Basis, V. Masson-Delmotte et al., Eds., Cambridge University Press, 33–144.

  • Armstrong, R. L., and Coauthors, 2018: Runoff from glacier ice and seasonal snow in High Asia: Separating melt water sources in river flow. Reg. Environ. Change, 19, 12491261, https://doi.org/10.1007/s10113-018-1429-0.

    • Search Google Scholar
    • Export Citation
  • Bedford, D. P., 1996: International water management in the Aral Sea Basin. Water Int., 21, 6369, https://doi.org/10.1080/02508069608686491.

    • Search Google Scholar
    • Export Citation
  • Brohan, P., J. J. Kennedy, I. Harris, S. F. B. Tett, and P. D. Jones, 2006: Uncertainty estimates in regional and global observed temperature changes: A new data set from 1850. J. Geophys. Res., 111, D12106, https://doi.org/10.1029/2005JD006548.

    • Search Google Scholar
    • Export Citation
  • Brown, S. J., J. Caesar, and C. A. T. Ferro, 2008: Global changes in extreme daily temperature since 1950. J. Geophys. Res., 113, D05115, https://doi.org/10.1029/2006JD008091.

    • Search Google Scholar
    • Export Citation
  • Chang, J., H. Zhang, Y. Wang, and Y. Zhu, 2016: Assessing the impact of climate variability and human activities on streamflow variation. Hydrol. Earth Syst. Sci., 20, 15471560, https://doi.org/10.5194/hess-20-1547-2016.

    • Search Google Scholar
    • Export Citation
  • Chen, F. H., W. Huang, L. Y. Jin, J. H. Chen, and J. S. Wang, 2011: Spatiotemporal precipitation variations in the arid Central Asia in the context of global warming. Sci. China Earth Sci., 54, 18121821, https://doi.org/10.1007/s11430-011-4333-8.

    • Search Google Scholar
    • Export Citation
  • Chen, M., W. Shi, P. Xie, V. B. S. Silva, V. E. Kousky, R. W. Higgins, and J. E. Janowiak, 2008: Assessing objective techniques for gauge-based analyses of global daily precipitation. J. Geophys. Res., 113, D04110, https://doi.org/10.1029/2007JD009132.

    • Search Google Scholar
    • Export Citation
  • Chevallier, P., B. Pouyaud, M. Mojaïsky, M. Bolgov, O. Olsson, M. Bauer, and J. Froebrich, 2014: River flow regime and snow cover of the Pamir Alay (Central Asia) in a changing climate. Hydrol. Sci. J., 59, 14911506, https://doi.org/10.1080/02626667.2013.838004.

    • Search Google Scholar
    • Export Citation
  • Dai, A., I. Y. Fung, and A. D. Del Genio, 1997: Surface observed global land precipitation variations during 1900-88. J. Climate, 10, 29432962, https://doi.org/10.1175/1520-0442(1997)010<2943:SOGLPV>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Diaz, H. F., R. S. Bradley, and J. K. Eischeid, 1989: Precipitation fluctuations over global land areas since the late 1800’s. J. Geophys. Res., 94, 11951210, https://doi.org/10.1029/JD094iD01p01195.

    • Search Google Scholar
    • Export Citation
  • Ding, J., L. Cuo, Y. Zhang, and F. Zhu, 2018: Monthly and annual temperature extremes and their changes on the Tibetan Plateau and its surrounding during 1963–2015. Sci. Rep., 8, 11860, https://doi.org/10.1038/s41598-018-30320-0.

    • Search Google Scholar
    • Export Citation
  • Donat, M. G., L. V. Alexander, N. Herold, and A. J. Dittus, 2016: Temperature and precipitation extremes in century-long gridded observations, reanalyses, and atmospheric model simulations. J. Geophys. Res. Atmos., 121, 11 17411 189, https://doi.org/10.1002/2016JD025480.

    • Search Google Scholar
    • Export Citation
  • Finaev, F. A., L. Shiyin, B. Weijia, and J. Li, 2016: Climate change and water potential of the Pamir mountains. Geogr. Environ. Sustainability, 9, 88105, https://doi.org/10.15356/2071-9388_03v09_2016_06.

    • Search Google Scholar
    • Export Citation
  • Froebrich, J., and O. Kayumov, 2004: Water management aspect of Amu Darya. Dying and Dead Seas Climatic Versus Anthropic Causes, J. C. J. Nihoul, P. O. Zavialov, and P. P. Micklin, Eds., NATO Science Series: IV: Earth and Environmental Sciences, Vol. 36, Springer, 49–76, https://doi.org/10.1007/978-94-007-0967-6_2.

  • Fuchs, M. C., R. Gloaguen, and E. Pohl, 2013: Tectonic and climatic forcing on the Panj river system during the quaternary. Int. J. Earth Sci., 102, 19852003, https://doi.org/10.1007/s00531-013-0916-2.

    • Search Google Scholar
    • Export Citation
  • Funk, C., A. Hoell, S. Shukla, I. Bladé, B. Liebmann, J. B. Roberts, F. R. Robertson, and G. Husak, 2014: Predicting East African spring droughts using Pacific and Indian Ocean Sea surface temperature indices. Hydrol. Earth Syst. Sci., 18, 49654978, https://doi.org/10.5194/hess-18-4965-2014.

    • Search Google Scholar
    • Export Citation
  • Gu, G., and R. F. Adler, 2015: Spatial patterns of global precipitation change and variability during 1901–2010. J. Climate, 28, 44314453, https://doi.org/10.1175/JCLI-D-14-00201.1.

    • Search Google Scholar
    • Export Citation
  • Gu, S., C. Huang, L. Bai, C. Chu, and Q. Liu, 2016: Heat-related illness in China, summer of 2013. Int. J. Biometeorol., 60, 131137, https://doi.org/10.1007/s00484-015-1011-0.

    • Search Google Scholar
    • Export Citation
  • Gulakhmadov, A., X. Chen, N. Gulahmadov, T. Liu, R. Davlyatov, S. Sharofiddinov, and M. Gulakhmadov, 2020: Long-term hydro-climatic trends in the mountainous Kofarnihon River basin in Central Asia. Water, 12, 2140, https://doi.org/10.3390/w12082140.

    • Search Google Scholar
    • Export Citation
  • Hagg, W., M. Hoelzle, S. Wagner, E. Mayr, and Z. Klose, 2013: Glacier and runoff changes in the Rukhk catchment, upper Amu-Darya basin until 2050. Global Planet. Change, 110, 6273, https://doi.org/10.1016/j.gloplacha.2013.05.005.

    • Search Google Scholar
    • Export Citation
  • Hartmann, D. L., and Coauthors, 2013: Observations: Atmosphere and surface. Climate Change 2013: The Physical Science Basis, T. F. Stocker et al., Eds., Cambridge University Press, 159–254.

  • He, Z. H., J. Parajka, F. Q. Tian, and G. Blöschl, 2014: Estimating degree-day factors from MODIS for snowmelt runoff modeling. Hydrol. Earth Syst. Sci., 18, 47734789, https://doi.org/10.5194/hess-18-4773-2014.

    • Search Google Scholar
    • Export Citation
  • Holmes, J. A., E. R. Cook, and B. Yang, 2009: Climate change over the past 2000 years in western China. Quat. Int., 194, 91107, https://doi.org/10.1016/j.quaint.2007.10.013.

    • Search Google Scholar
    • Export Citation
  • Hou, M., L. Cuo, A. Murodov, J. Ding, Y. Luo, T. Liu, and X. Chen, 2023: Streamflow composition and the contradicting impacts of anthropogenic activities and climatic change on streamflow in the Amu Darya basin, central Asia. J. Hydrometeor., 24, 185201, https://doi.org/10.1175/JHM-D-22-0040.1.

    • Search Google Scholar
    • Export Citation
  • Hu, Z., C. Zhang, Q. Hu, and H. Tian, 2014: Temperature changes in central Asia from 1979 to 2011 based on multiple datasets. J. Climate, 27, 11431167, https://doi.org/10.1175/JCLI-D-13-00064.1.

    • Search Google Scholar
    • Export Citation
  • Huang, J., X. Guan, and F. Ji, 2012: Enhanced cold-season warming in semi-arid regions. Atmos. Chem. Phys., 12, 53915398, https://doi.org/10.5194/acp-12-5391-2012.

    • Search Google Scholar
    • Export Citation
  • Immerzeel, W. W., A. F. Lutz, and P. Droogers, 2012: Climate change impacts on the upstream water resources of the Amu and Syr Darya River Basins. FutureWater Rep. 107, 103 pp., https://www.futurewater.eu/wp-content/uploads/2012/03/Upstream_Report_FW_web.pdf.

  • Jaagus, J., 2006: Climatic changes in Estonia during the second half of the 20th century in relationship with changes in large-scale atmospheric circulation. Theor. Appl. Climatol., 83, 7788, https://doi.org/10.1007/s00704-005-0161-0.

    • Search Google Scholar
    • Export Citation
  • Karl, T. R., and Coauthors, 1993: A new perspective on recent global warming: Asymmetric trends of daily maximum and minimum temperature. Bull. Amer. Meteor. Soc., 74, 10071024, https://doi.org/10.1175/1520-0477(1993)074<1007:ANPORG>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kayumov, A., 2010: Glaciers resources of Tajikistan in condition of the climate change. Tech. Rep., 32 pp., https://s3.amazonaws.com/media.archnet.org/system/publications/contents/6892/original/DPC3768.pdf?1384804666.

  • Kendall, M. G., 1975: Rank Correlation Methods. 4th ed. Charles Griffin, 202 pp.

  • Khakimov, F. X., O. B. Mirzokhonov, and S. O. Mirzokhonova, 2007: The trend of air temperature change in the western of Pamir in aspect of global warming (in Russian). Rep. Acad. Sci. Repub. Tajikistan, 776–785, https://cyberleninka.ru/article/n/tendentsiya-izmeneniya-temperatury-vozduha-na-zapadnom-pamire-v-aspekte-globalnogo-potepleniya-klimata.

  • Klein Tank, A. M. G., and Coauthors, 2006: Changes in daily temperature and precipitation extremes in central and south Asia. J. Geophys. Res., 111, D16105, https://doi.org/10.1029/2005JD006316.

    • Search Google Scholar
    • Export Citation
  • Li, Q., X. Li, Y. Ran, M. Feng, Y. Nian, M. Tan, and X. Chen, 2021: Investigate the relationships between the Aral Sea shrinkage and the expansion of cropland and reservoir in its drainage basins between 2000 and 2020. Int. J. Digit. Earth, 14, 661677, https://doi.org/10.1080/17538947.2020.1865466.

    • Search Google Scholar
    • Export Citation
  • Lioubimtseva, E., and G. M. Henebry, 2009: Climate and environmental change in arid Central Asia: Impacts, vulnerability, and adaptations. J. Arid Environ., 73, 963977, https://doi.org/10.1016/j.jaridenv.2009.04.022.

    • Search Google Scholar
    • Export Citation
  • Lioubimtseva, E., R. Cole, J. M. Adams, and G. Kapustin, 2005: Impacts of climate and land-cover changes in arid lands of Central Asia. J. Arid Environ., 62, 285308, https://doi.org/10.1016/j.jaridenv.2004.11.005.

    • Search Google Scholar
    • Export Citation
  • Lobanova, A., and D. Iulii, 2019: Analysis of the water quality parameters in the Amudarya River. CAREC Analytical Rep., 60 pp., http://www.cawater-info.net/amudarya-knowledge-base/pdf/amu-wat-qual-en.pdf.

  • Mahmadaliev, 2009: Improving hydrometeorological support in the Republic of Tajikistan (in Russian). World Bank Rep., 113 pp., http://meteo-new.tj/files/doc/The%20report%20of%20the%20world%20bank%20rus%2030-09-09.pdf.

  • Mann, H. B., 1945: Non-parametric test against trend. Econometrica, 13, 245259, https://doi.org/10.2307/1907187.

  • Mannig, B., and Coauthors, 2013: Dynamical downscaling of climate change in Central Asia. Global Planet. Change, 110, 2639, https://doi.org/10.1016/j.gloplacha.2013.05.008.

    • Search Google Scholar
    • Export Citation
  • Martino, L. D., and Coauthors, 2009: Climate Change in Central Asia: A Visual Synthesis. Swiss Federal Office for the Environment (FOEN), 79 pp.

  • Nezlin, P. N., A. G. Kostianoy, and S. A. Lebedev, 2004: Interannual variations of the discharge of Amu Darya and Syr Darya estimated from global atmospheric precipitation. J. Mar. Syst., 47, 6775, https://doi.org/10.1016/j.jmarsys.2003.12.009.

    • Search Google Scholar
    • Export Citation
  • Pfeffer, W. T., and Coauthors, 2014: The Randolph glacier inventory: A globally complete inventory of glaciers. J. Glaciol., 60, 537552, https://doi.org/10.3189/2014JoG13J176.

    • Search Google Scholar
    • Export Citation
  • Pohl, E., R. Gloaguen, C. Andermann, and M. Knoche, 2017: Glacier melt buffers river runoff in the Pamir Mountains. Water Resour. Res., 53, 24672489, https://doi.org/10.1002/2016WR019431.

    • Search Google Scholar
    • Export Citation
  • Poudel, A., L. Cuo, J. Ding, and A. R. Gyawali, 2020: Spatio-temperoal variability of the annual and monthly extreme temperature indices in Nepal. Int. J. Climatol., 40, 49564977, https://doi.org/10.1002/joc.6499.

    • Search Google Scholar
    • Export Citation
  • Robine, J. M., S. L. K. Cheung, S. Le Roy, H. Van Oyen, C. Griffiths, J. P. Michel, and F. R. Herrmann, 2008: Death toll exceeded 70 000 in Europe during the summer of 2003. C. R. Biol., 331, 171178, https://doi.org/10.1016/j.crvi.2007.12.001.

    • Search Google Scholar
    • Export Citation
  • Schiemann, R., D. Lüthi, P. L. Vidale, and C. Schär, 2008: The precipitation climate of Central Asia—Intercomparison of observational and numerical data sources in a remote semiarid region. Int. J. Climatol., 28, 295314, https://doi.org/10.1002/joc.1532.

    • Search Google Scholar
    • Export Citation
  • Sen, P. K., 1968: Estimates of the regression coefficient based on Kendall’s Tau. J. Amer. Stat. Assoc., 63, 13791389, https://doi.org/10.1080/01621459.1968.10480934.

    • Search Google Scholar
    • Export Citation
  • Seneviratne, S. I., and Coauthors, 2021: Weather and climate extreme events in a changing climate. Climate Change 2021: The Physical Science Basis, V. Masson-Delmotte et al., Eds., Cambridge University Press, 1513–1766, https://doi.org/10.1017/9781009157896.013.

  • Sheffield, J., G. Goteti, and E. F. Wood, 2006: Development of a 50-year high-resolution global dataset of meteorological forcings for land surface modeling. J. Climate, 19, 30883111, https://doi.org/10.1175/JCLI3790.1.

    • Search Google Scholar
    • Export Citation
  • Sidike, A., X. Chen, T. Liu, D. Khaidar, and Y. Huang, 2016: Investigating alternative climate data sources for hydrological simulations in the upstream of the Amu Darya River. Water, 8, 441, https://doi.org/10.3390/w8100441.

    • Search Google Scholar
    • Export Citation
  • Stulina, G., and O. Eshchanov, 2013: Climate change impacts on hydrology and environment in the Pre-Aral region. Quat. Int., 311, 8796, https://doi.org/10.1016/j.quaint.2013.07.015.

    • Search Google Scholar
    • Export Citation
  • Su, Y., and Coauthors, 2021: High agricultural water consumption led to the continued shrinkage of the Aral Sea during 1992–2015. Sci. Total Environ., 777, 145993, https://doi.org/10.1016/j.scitotenv.2021.145993.

    • Search Google Scholar
    • Export Citation
  • Sun, Q., X. Zhang, F. Zwiers, S. Westra, and L. V. Alexander, 2021: A global, continental, and regional analysis of changes in extreme precipitation. J. Climate, 34, 243258, https://doi.org/10.1175/JCLI-D-19-0892.1.

    • Search Google Scholar
    • Export Citation
  • Trenberth, K. E., and Coauthors, 2007: Observations: Surface and atmospheric climate change. Climate Change 2007: The Physical Science Basis, S. Solomon et al., Eds., Cambridge University Press, 235–336.

  • Toreti, A., and F. Desiato, 2008: Changes in temperature extremes over Italy in the last 44 years. Int. J. Climatol., 28, 733745, https://doi.org/10.1002/joc.1576.

    • Search Google Scholar
    • Export Citation
  • Unger-Shayesteh, K., S. Vorogushyn, B. Merz, and H. G. Frede, 2013: Introduction to “Water in Central Asia – Perspectives under global change.” Global Planet. Change, 110, 13, https://doi.org/10.1016/j.gloplacha.2013.09.016.

    • Search Google Scholar
    • Export Citation
  • Vincent, L. A., and Coauthors, 2005: Observed trends in indices of daily temperature extremes in South America 1960–2000. J. Climate, 18, 50115023, https://doi.org/10.1175/JCLI3589.1.

    • Search Google Scholar
    • Export Citation
  • Wang, X., Y. Luo, L. Sun, C. He, Y. Zhang, and S. Liu, 2016: Attribution of runoff decline in the Amu Darya River in Central Asia during 1951–2007. J. Hydrometeor., 17, 15431560, https://doi.org/10.1175/JHM-D-15-0114.1.

    • Search Google Scholar
    • Export Citation
  • Wang, Y., X. Liu, and U. Herzschuh, 2010: Asynchronous evolution of the Indian and East Asian summer monsoon indicated by Holocene moisture patterns in monsoonal central Asia. Earth-Sci. Rev., 103, 135153, https://doi.org/10.1016/j.earscirev.2010.09.004.

    • Search Google Scholar
    • Export Citation
  • White, J. C., W. T. Tanton, and W. D. Rycroft, 2014: The impact of climate change on the water resources of the Amu Darya Basin in central Asia. Water Resour. Manage., 28, 52675281, https://doi.org/10.1007/s11269-014-0716-x.

    • Search Google Scholar
    • Export Citation
  • Yang, X., N. Wang, A. Chen, J. He, T. Hua, and Y. Qie, 2020: Changes in area and water volume of the Aral Sea in the arid Central Asia over the period of 1960–2018 and their causes. Catena, 191, 104566, https://doi.org/10.1016/j.catena.2020.104566.

    • Search Google Scholar
    • Export Citation
  • Yatagai, A., K. Kamiguchi, O. Arakawa, A. Hamada, N. Yasutomi, and A. Kitoh, 2012: APHRODITE: Constructing a long-term daily gridded precipitation dataset for Asia based on a dense network of rain gauges. Bull. Amer. Meteor. Soc., 93, 14011415, https://doi.org/10.1175/BAMS-D-11-00122.1.

    • Search Google Scholar
    • Export Citation
  • Yosef, Y., E. Aguilar, and P. Alpert, 2019: Changes in extreme temperature and precipitation indices: Using an innovative daily homogenized database in Israel. Int. J. Climatol., 39, 50225045, https://doi.org/10.1002/joc.6125.

    • Search Google Scholar
    • Export Citation
  • Zhang, M., Y. Chen, Y. Shen, and Y. Li, 2017: Changes of precipitation extremes in arid Central Asia. Quat. Int., 436, 1627, https://doi.org/10.1016/j.quaint.2016.12.024.

    • Search Google Scholar
    • Export Citation
  • Zonn, S. I., 2014: Karakum canal: Artificial river in a desert. The Turkmen Lake Altyn Asyr and Water Resources in Turkmenistan, I. Zonn and A. Kostianoy, Eds., Handbook of Environmental Chemistry, Vol. 28, Springer, 95–106, https://doi.org/10.1007/698_2012_194.

Supplementary Materials

Save
  • Aalto, J., M. Kämäräinen, M. Shodmonov, N. Rajabovc, and A. Venäläinena, 2017: Features of Tajikistan’s past and future climate. Int. J. Climatol., 37, 49494961, https://doi.org/10.1002/joc.5135.

    • Search Google Scholar
    • Export Citation
  • Adenbaev, Y. B., F. H. Hikmatov, and O. A. Kxaydarova, 2015: Hydrological regime of the lower reaches of Amudarja River in the conditions of intensive economical use of water resources. Eur. Sci. Rev., 910, 18–21.

    • Search Google Scholar
    • Export Citation
  • Agal’tseva, N. A., M. V. Bolgov, T. Yu. Spektorman, M. D. Trubetskova, and V. E. Chub, 2011: Estimating hydrological characteristics in the Amu Darya River basin under climate change conditions (in Russian). Russ. Meteor. Hydrol., 36, 681689, https://doi.org/10.3103/S1068373911100062.

    • Search Google Scholar
    • Export Citation
  • Alexander, L. V., and Coauthors, 2006: Global observed changes in daily climate extremes of temperature and precipitation. J. Geophys. Res., 111, D05109, https://doi.org/10.1029/2005JD006290.

    • Search Google Scholar
    • Export Citation
  • Arias, P., and Coauthors, 2021: Technical summary. Climate Change 2021: The Physical Science Basis, V. Masson-Delmotte et al., Eds., Cambridge University Press, 33–144.

  • Armstrong, R. L., and Coauthors, 2018: Runoff from glacier ice and seasonal snow in High Asia: Separating melt water sources in river flow. Reg. Environ. Change, 19, 12491261, https://doi.org/10.1007/s10113-018-1429-0.

    • Search Google Scholar
    • Export Citation
  • Bedford, D. P., 1996: International water management in the Aral Sea Basin. Water Int., 21, 6369, https://doi.org/10.1080/02508069608686491.

    • Search Google Scholar
    • Export Citation
  • Brohan, P., J. J. Kennedy, I. Harris, S. F. B. Tett, and P. D. Jones, 2006: Uncertainty estimates in regional and global observed temperature changes: A new data set from 1850. J. Geophys. Res., 111, D12106, https://doi.org/10.1029/2005JD006548.

    • Search Google Scholar
    • Export Citation
  • Brown, S. J., J. Caesar, and C. A. T. Ferro, 2008: Global changes in extreme daily temperature since 1950. J. Geophys. Res., 113, D05115, https://doi.org/10.1029/2006JD008091.

    • Search Google Scholar
    • Export Citation
  • Chang, J., H. Zhang, Y. Wang, and Y. Zhu, 2016: Assessing the impact of climate variability and human activities on streamflow variation. Hydrol. Earth Syst. Sci., 20, 15471560, https://doi.org/10.5194/hess-20-1547-2016.

    • Search Google Scholar
    • Export Citation
  • Chen, F. H., W. Huang, L. Y. Jin, J. H. Chen, and J. S. Wang, 2011: Spatiotemporal precipitation variations in the arid Central Asia in the context of global warming. Sci. China Earth Sci., 54, 18121821, https://doi.org/10.1007/s11430-011-4333-8.

    • Search Google Scholar
    • Export Citation
  • Chen, M., W. Shi, P. Xie, V. B. S. Silva, V. E. Kousky, R. W. Higgins, and J. E. Janowiak, 2008: Assessing objective techniques for gauge-based analyses of global daily precipitation. J. Geophys. Res., 113, D04110, https://doi.org/10.1029/2007JD009132.

    • Search Google Scholar
    • Export Citation
  • Chevallier, P., B. Pouyaud, M. Mojaïsky, M. Bolgov, O. Olsson, M. Bauer, and J. Froebrich, 2014: River flow regime and snow cover of the Pamir Alay (Central Asia) in a changing climate. Hydrol. Sci. J., 59, 14911506, https://doi.org/10.1080/02626667.2013.838004.

    • Search Google Scholar
    • Export Citation
  • Dai, A., I. Y. Fung, and A. D. Del Genio, 1997: Surface observed global land precipitation variations during 1900-88. J. Climate, 10, 29432962, https://doi.org/10.1175/1520-0442(1997)010<2943:SOGLPV>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Diaz, H. F., R. S. Bradley, and J. K. Eischeid, 1989: Precipitation fluctuations over global land areas since the late 1800’s. J. Geophys. Res., 94, 11951210, https://doi.org/10.1029/JD094iD01p01195.

    • Search Google Scholar
    • Export Citation
  • Ding, J., L. Cuo, Y. Zhang, and F. Zhu, 2018: Monthly and annual temperature extremes and their changes on the Tibetan Plateau and its surrounding during 1963–2015. Sci. Rep., 8, 11860, https://doi.org/10.1038/s41598-018-30320-0.

    • Search Google Scholar
    • Export Citation
  • Donat, M. G., L. V. Alexander, N. Herold, and A. J. Dittus, 2016: Temperature and precipitation extremes in century-long gridded observations, reanalyses, and atmospheric model simulations. J. Geophys. Res. Atmos., 121, 11 17411 189, https://doi.org/10.1002/2016JD025480.

    • Search Google Scholar
    • Export Citation
  • Finaev, F. A., L. Shiyin, B. Weijia, and J. Li, 2016: Climate change and water potential of the Pamir mountains. Geogr. Environ. Sustainability, 9, 88105, https://doi.org/10.15356/2071-9388_03v09_2016_06.

    • Search Google Scholar
    • Export Citation
  • Froebrich, J., and O. Kayumov, 2004: Water management aspect of Amu Darya. Dying and Dead Seas Climatic Versus Anthropic Causes, J. C. J. Nihoul, P. O. Zavialov, and P. P. Micklin, Eds., NATO Science Series: IV: Earth and Environmental Sciences, Vol. 36, Springer, 49–76, https://doi.org/10.1007/978-94-007-0967-6_2.

  • Fuchs, M. C., R. Gloaguen, and E. Pohl, 2013: Tectonic and climatic forcing on the Panj river system during the quaternary. Int. J. Earth Sci., 102, 19852003, https://doi.org/10.1007/s00531-013-0916-2.

    • Search Google Scholar
    • Export Citation
  • Funk, C., A. Hoell, S. Shukla, I. Bladé, B. Liebmann, J. B. Roberts, F. R. Robertson, and G. Husak, 2014: Predicting East African spring droughts using Pacific and Indian Ocean Sea surface temperature indices. Hydrol. Earth Syst. Sci., 18, 49654978, https://doi.org/10.5194/hess-18-4965-2014.

    • Search Google Scholar
    • Export Citation
  • Gu, G., and R. F. Adler, 2015: Spatial patterns of global precipitation change and variability during 1901–2010. J. Climate, 28, 44314453, https://doi.org/10.1175/JCLI-D-14-00201.1.

    • Search Google Scholar
    • Export Citation
  • Gu, S., C. Huang, L. Bai, C. Chu, and Q. Liu, 2016: Heat-related illness in China, summer of 2013. Int. J. Biometeorol., 60, 131137, https://doi.org/10.1007/s00484-015-1011-0.

    • Search Google Scholar
    • Export Citation
  • Gulakhmadov, A., X. Chen, N. Gulahmadov, T. Liu, R. Davlyatov, S. Sharofiddinov, and M. Gulakhmadov, 2020: Long-term hydro-climatic trends in the mountainous Kofarnihon River basin in Central Asia. Water, 12, 2140, https://doi.org/10.3390/w12082140.

    • Search Google Scholar
    • Export Citation
  • Hagg, W., M. Hoelzle, S. Wagner, E. Mayr, and Z. Klose, 2013: Glacier and runoff changes in the Rukhk catchment, upper Amu-Darya basin until 2050. Global Planet. Change, 110, 6273, https://doi.org/10.1016/j.gloplacha.2013.05.005.

    • Search Google Scholar
    • Export Citation
  • Hartmann, D. L., and Coauthors, 2013: Observations: Atmosphere and surface. Climate Change 2013: The Physical Science Basis, T. F. Stocker et al., Eds., Cambridge University Press, 159–254.

  • He, Z. H., J. Parajka, F. Q. Tian, and G. Blöschl, 2014: Estimating degree-day factors from MODIS for snowmelt runoff modeling. Hydrol. Earth Syst. Sci., 18, 47734789, https://doi.org/10.5194/hess-18-4773-2014.

    • Search Google Scholar
    • Export Citation
  • Holmes, J. A., E. R. Cook, and B. Yang, 2009: Climate change over the past 2000 years in western China. Quat. Int., 194, 91107, https://doi.org/10.1016/j.quaint.2007.10.013.

    • Search Google Scholar
    • Export Citation
  • Hou, M., L. Cuo, A. Murodov, J. Ding, Y. Luo, T. Liu, and X. Chen, 2023: Streamflow composition and the contradicting impacts of anthropogenic activities and climatic change on streamflow in the Amu Darya basin, central Asia. J. Hydrometeor., 24, 185201, https://doi.org/10.1175/JHM-D-22-0040.1.

    • Search Google Scholar
    • Export Citation
  • Hu, Z., C. Zhang, Q. Hu, and H. Tian, 2014: Temperature changes in central Asia from 1979 to 2011 based on multiple datasets. J. Climate, 27, 11431167, https://doi.org/10.1175/JCLI-D-13-00064.1.

    • Search Google Scholar
    • Export Citation
  • Huang, J., X. Guan, and F. Ji, 2012: Enhanced cold-season warming in semi-arid regions. Atmos. Chem. Phys., 12, 53915398, https://doi.org/10.5194/acp-12-5391-2012.

    • Search Google Scholar
    • Export Citation
  • Immerzeel, W. W., A. F. Lutz, and P. Droogers, 2012: Climate change impacts on the upstream water resources of the Amu and Syr Darya River Basins. FutureWater Rep. 107, 103 pp., https://www.futurewater.eu/wp-content/uploads/2012/03/Upstream_Report_FW_web.pdf.

  • Jaagus, J., 2006: Climatic changes in Estonia during the second half of the 20th century in relationship with changes in large-scale atmospheric circulation. Theor. Appl. Climatol., 83, 7788, https://doi.org/10.1007/s00704-005-0161-0.

    • Search Google Scholar
    • Export Citation
  • Karl, T. R., and Coauthors, 1993: A new perspective on recent global warming: Asymmetric trends of daily maximum and minimum temperature. Bull. Amer. Meteor. Soc., 74, 10071024, https://doi.org/10.1175/1520-0477(1993)074<1007:ANPORG>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kayumov, A., 2010: Glaciers resources of Tajikistan in condition of the climate change. Tech. Rep., 32 pp., https://s3.amazonaws.com/media.archnet.org/system/publications/contents/6892/original/DPC3768.pdf?1384804666.

  • Kendall, M. G., 1975: Rank Correlation Methods. 4th ed. Charles Griffin, 202 pp.

  • Khakimov, F. X., O. B. Mirzokhonov, and S. O. Mirzokhonova, 2007: The trend of air temperature change in the western of Pamir in aspect of global warming (in Russian). Rep. Acad. Sci. Repub. Tajikistan, 776–785, https://cyberleninka.ru/article/n/tendentsiya-izmeneniya-temperatury-vozduha-na-zapadnom-pamire-v-aspekte-globalnogo-potepleniya-klimata.

  • Klein Tank, A. M. G., and Coauthors, 2006: Changes in daily temperature and precipitation extremes in central and south Asia. J. Geophys. Res., 111,