Impact of climate-induced extreme events and demand – supply gap on water resources in Bangladesh

Agriculture, domestic, and industry rely on water resources systems for ful ﬁ lling water demand, while water resources systems face both climate-induced extreme events and management and governance problems. These constraints lead to a mismatch between demand and supply of water for those sectors. This study applies central tendency and variability to analyze data and mixed methods approach to interpret the result. From 1999 to 2019, the growth rates of population, gross domestic product, and urban population are ­ 1.354, 6.084, and 3.70%, respectively, contributing to increased water demand. However, the average groundwater depletion increased from 2.455 km 3 (1989 – 1990) to 4.9016 km 3 , while the average ﬂ ood-affected areas enhanced by 8,644 km 2 in 2014 – 2018 compared to 1987 – 1991. Furthermore, salt-affected areas incremented by 222,810 ha in 2009 contrasted to 1973, whereas the mean wind speed of cyclones increased by 30.02 km in 2015 – 2020 compared to 1988 – 1995. The mean sea-level rise increased by 16.8 and 169.2 cm in 1995 – 2000 compared to 1979 – 1983 in Cox ’ s Bazar and the Char Changa station, respectively. The Mann – Kendall test was applied to detect the trend. This study urges policymakers, water experts, and academics to promote rainwater harvesting that is sustainable to govern rainwater and miti-gate water and economic poverty. intensity of extreme that affects water resources by analyzing national-level data. It will help the national and global community advance water resources management for climate change adaptation and sustainable economic development. It will also urge ﬁ nding and accommodating a sustainable approach to reducing the water demand – supply gap. This study is relevant for var-ious government organizations dealing with water resources management, 4 policymakers, and professionals in water resources management and governance. This study follows a mixed-method approach to analyze the data. This study employs central tendency and a measure of variability to analyze data. Measures of central tendency include mean and median, while variability attributes to standard deviation, minimum, and maximum. These components help to understand the intensity, frequency, and average change in climate-induced extreme events that affect water resources. To identify the trend, the Mann – Kendall (MK) test (Mann 1945; Kendall 1975) was applied that widely employed in hydrological (Yue & Wang 2004; Machiwal & Jha 2012) and climatological (Mavromatis & Stathis 2011) time-series data. As per this test, the null hypothesis (H 0 ) is that there is no trend, while the alternative hypothesis (H 1 ) is that there is a presence of trend (Onoz & Bayezit 2012). The ‘ R ’ software was employed to perform the MK test. This study uses the following hypotheses for this test.


INTRODUCTION
Water is a vital component of the planet, and water resources are intricately connected to climate change dynamics (Bates et al. 2008;IPCC 2014). Water systems are vulnerable to man-made, natural, and climatic extreme events such as floods, cyclones, drought, heatwaves, rainfall, storm surges, and sea-level rise. However, over the past 100 years, global water consumption altogether in household, industry, and agriculture has increased by a factor of six and continue to increase steadily at a rate of 1% per year due to population growth, change in consumption patterns, and economic advancement (Wada & Bierkens 2014;UN-Water 2020). However, the supply of water is more erratic and uncertain (UNU Water 2013;FAO 2017;IPCC 2018) due to climate change that directly and indirectly changes the hydrological systems that are disturbing water quality, volume, and change in demand of water . This mismatch between demand and supply of water (Liu et al. 2017) leads to the water crisis that is more severe in Bangladesh.
Therefore, the water crisis is a crucial challenge in Bangladesh due to (a) growing demand throughout the country that creates immense pressure on groundwater sources (Chowdhury 2010); (b) an alarming increase in water pollution due to unplanned industrialization and dire sanitation circumstances (World Bank 2005); (c) water resource management that is the sole responsibility of a few organizations 1 and does not have good coordination between them (Chan et al. 2016); (d) upstream water diversion through different dams and barrages 2 in the transboundary rivers which inhibits the ability to mitigate water crisis (Chan et al. 2016); and (e) adverse impacts of climate change 3 (Asaduzzaman et al. 2010).
There is a lack of investigation in Bangladesh by analyzing the national-level data and quantifying the level of change in extreme events that affect water resources. A study of the whole country is needed and important for better understanding the factors responsible for rising water demand and supply-side response. The main objective of this paper is to advance the literature on factors responsible for rising water demand, quantify water sources response to this, and quantify the intensity of extreme that affects water resources by analyzing national-level data. It will help the national and global community advance water resources management for climate change adaptation and sustainable economic development. It will also urge finding and accommodating a sustainable approach to reducing the water demand-supply gap. This study is relevant for various government organizations dealing with water resources management, 4 policymakers, and professionals in water resources management and governance.

METHODOLOGY Study site
Bangladesh is not a large country in terms of area (147,570 km 2 ) but ranked 8th and 5th in the World and Asia, respectively, in terms of population size (163 million). The area attributes to 18,290 km 2 (water) and 130,170 km 2 (land). Figure 1 shows that it locates on the Bay of Bengal in South Asia surrounded by India (west, north, and east) and Myanmar (south-east). The topography of the country characterizes by a broad alluvial plain (southern and rises toward the north), hilly region (northeastern and southeastern), and terrace land (central and northwestern) (Shahid 2010). The topography and geographical locations make the country more susceptible to climate-induced hazards (Bhowmik et al. 2021). A major part of this country (80%) consists of the floodplain of Ganges, Brahmaputra, and Meghna (GBM), and some other rivers (Brouwer et al. 2007).
Water resources in Bangladesh can be categorized into two groups, namely surface and groundwater. The surface water is attributed to transboundary water flow, rainwater, water on seasonal wetlands, and water on standing water bodies (Ahmed & Roy 2007). Water flows from these sources substantially varied between monsoon and dry seasons. Major responsible factors for these variations are water flow in GBM and rainfall patterns. For instance, more than 75% of annual rainfall takes place in monsoon and water management practices of upstream countries (e.g., planned interventions and anthropogenic actions) (Shahid 2010;Kolås et al. 2013).
Since the country stands on the Northern Tropic, its climate is tropically characterized by high temperature, heavy seasonal rainfall, high humidity, and distinguished seasonal variations (Thomas et al. 2013). Rainfall varies from 1,400 mm (west) to .4,300 mm (east) side of the country (Shahid 2010). According to Shahid (2010), seasons can be divided into four categories

METHODS
This study follows a mixed-method approach to analyze the data. This study employs central tendency and a measure of variability to analyze data. Measures of central tendency include mean and median, while variability attributes to standard deviation, minimum, and maximum.
These components help to understand the intensity, frequency, and average change in climate-induced extreme events that affect water resources. To identify the trend, the Mann-Kendall (MK) test (Mann 1945;Kendall 1975) was applied that widely employed in hydrological (Yue & Wang 2004;Machiwal & Jha 2012) and climatological (Mavromatis & Stathis 2011) timeseries data. As per this test, the null hypothesis (H 0 ) is that there is no trend, while the alternative hypothesis (H 1 ) is that there is a presence of trend (Onoz & Bayezit 2012). The 'R' software was employed to perform the MK test. This study uses the following hypotheses for this test.

Analytical framework
Climate change exposes itself as enhancing the frequency and intensity of extreme events that influence the quality and quantity of water in multiple ways (Bates et al. 2008). For example, water quality is degraded through higher water temperature, reduced dissolved oxygen, rising pollutant concentration during drought, water pollution, deterioration of groundwater quality due to saltwater intrusion in coastal settings, disruption of water treatment facilities during floods (IPCC 2014; UNEP 2016), pathogen contamination by floods, and thus a reduced self-purifying capacity of freshwater bodies (Smakhtin et al. 2020), arsenic contamination in groundwater, 6 while erratic rainfall affects the availability of water. These impacts add challenges to the sustainable management of water resources (WWAP/UN-Water 2018) that threaten the enjoyment of human rights to water and sanitation for potentially billions of people (UN Water 2020), particularly in developing countries like Bangladesh.
Different studies have shown the adverse impact of climate change on water resources in Bangladesh (IPCC 2007;WHO 2008;CCC 2009;Rabanni et al. 2012;Abedin et al. 2014;Chan et al. 2016) that lead to the deficiency of water. The water resources system of the country is exposed to vulnerability of different types of regional climate variability such as rising temperature, the erratic nature of rainfall, droughts, floods, the rise of sea level, cyclones and storm surge, saltwater intrusion (Khan et al. 2011;Rabanni et al. 2012), and arsenic contamination (Chan et al. 2016) that adversely affect the supply of water. At the same time, rapid economic and population growth, and urbanization in Bangladesh cause water demand to rise in household, industry, and agriculture. In this context, it is high time to quantify the factors responsible for rising water demand, the response of water sources to this change, and climatic extremes on water resources in Bangladesh. Figure 3 explores the connection between climatic extreme events, mismatch between demand and supply, and water resources systems in Bangladesh.

Demand side of water
The current population of Bangladesh is about 160 million and Population Reference Bureau (PRB) estimates 250 million by 2050 based on the natural growth rate is 1.90% per annum (Streatfield & Karar 2008). This growth rate is based on crude death rate (CDR) 8/1000 and crude birth rate (CBR) 27/1000, which is more realistic (Streatfield & Karar 2008). Figure 4 presents that the total population increases each year that commenced at 2.02% in 1999 and shrunk to 1.04% in 2019. It grows on an average of 1.354% per year. This growing population is taking a toll on water demand in two different ways. First, household water demand where they need water for their use includes washing, drinking, bathing, and so on. Secondly, the extra population needs extra food and other products and services that also require water to produce. Figure 4 shows the growth rate in total population, GDP, urban population, and irrigated areas that increase the demand for water in Bangladesh. Figure 4 shows that the urban population increases by an average of 3.708% over the last 21 years. At the same time, GDP (Gross Domestic Product) grew on an average of 6.0846% per annum that was 4.67% in 1999, and rise to 8.15% in 2019. The GDP goes up more than 6% every year after 2010. This high growth of the economy intensifies the pressure on water resources to supply more water to maintain the growth. However, the growth of the urban population along with economic  wealth can raise the demand for water-intensive foods such as meat, milk, egg, pulse, nuts, and cereals which ultimately lead to increased water demand. However, the country uses 51.346% of the total land in 2004 for agricultural production and it rises to 59.711% in 2016 to ensure food security. This growth reveals that water demand also increases in the agriculture sector to ensure food security for the extra population. To fulfill the demand for water for the growing population, economy, and industry, it needs to ensure a stable water supply from different sources. Otherwise, it will create a water crisis and constraint economic development.

Supply side of water
Transboundary river water flow The Ganges, Brahmaputra, Meghna (GBM), and Southeastern Basin (SEB) drain large volumes of rainwater that occurred within and outside of the country. During this passage, it creates and records the peak level of water in each basin every year. This peak level data use and analyze to interpret the stability of the water supply from this source. Table 2 shows the central tendency and variability of peak water flow for the Brahmaputra River basins from 2012 to 2018. It shows that the mean peak water level was 19.  Extreme events H 0 : there is no trend in climate-induced extreme events that limit water supply and affect water resources in Bangladesh H 1 : there is a trend in climate-induced extreme events that limit water supply and affect water resources in Bangladesh monsoon seasons, respectively, which reveal that water flow during the dry season is very low and during monsoon is very high. This high difference between maximum points out unstable water supply and makes it difficult to fulfill water demand through this source. Another attribute is the standard deviation that points out that water flow is not stable. This high standard deviation means that the actual flow is far away (À/þ) from the mean. Table 2 shows some indicators that measure unstable water supply in the Ganges River Basin in terms of peak water flow. The first one is the mean that began at 13.007 m in 2012 and reached 17.758 m in 2017 and 16.7472 m in 2018. It is a big difference between 2018 and 2017 compared to 2012. The following indicator is the standard deviation, and its increasing trend indicates that river water flow is going more abnormal per year, particularly in recent years (2017 and 2018) with respect to 2012. The third indicator is the sum of transboundary water flow each year that commenced at 130.07 m in 2012 and reached the highest at 443.95 m in 2017 and the second highest at 418.68 m in 2018. However, the gap between maximum and minimum was 29.14 m in 2012 and reached 69.95 m in 2018 that more than double contrasted to 2012. This large extension of the gap is not a good sign for a stable supply of water for the basin people, aquatic, marine communities, and the economy.
The water flow in the Meghna basin is more stable than other basins in terms of standard deviation, range, and mean but a big change shows in total volume. The first measure average peak water flows less than 12 m but more than 11 m except in 2012, while standard deviation ranges from 3 to 6 m. However, total peak water flow increases significantly from 147.86 m in 2012 to 310.45 m in 2017 and 289.34 m in 2018. Another significant change is in the difference between maximum and minimum indicates that widen the gap between monsoon and dry season water flow. For instance, it started at 12.24 m in 2012 and reached 20.5 m in 2018. This extension of the gap is not a good sign for water supply from this river because it suggests increasing water supply during monsoon and a decrease in the dry season.
The central tendency and variability of the Southeastern Basin in Table 2 show that no big change has been taking place regarding standard deviation, while it takes place big in total water flows, the difference between the maximum and minimum peak water flow. The mean peak level of water flow was reduced by 7.07% in 2018 balanced to 2012. The volume of peak water flow ranges from 87.43 to 106.22 m which is not a good sign. The range is started at 9.45 m in 2012 and reached 18.36 m in 2018 that is increased by 94.29% compared to 2012. This high variability in peak water flow in the river basin along with erratic rainfall increases the probability of more reliance on groundwater that also responsible for causing a mismatch between demand and supply of water.

Groundwater
Groundwater is another main source of water. The use of this source is more intense during the dry season for fulfilling the demand than during the monsoon season. Easier accessibility and less expensive make this source more useful than other sources. Due to these advantages, people are more reliant and overexploiting this source. This overexploitation makes this source no more reliable, and the groundwater table is going down with time. Table 3 shows the historical depletion of groundwater at different time intervals. The intensity of depletion of the country is very high and it is growing. The mean and median levels of depletion enlarge by 100 and 132% in 2008-2009, respectively, in comparison to 1989-1990. At the same time, the range 7 positively adjusted with 41.49% in 2008-2009. All of these attributes indicate the intensity of growth in groundwater exhaustion. This depletion reduces the availability of water for different sectors and makes it more challenging to meet up the demand.

Rainfall
Rainfall is one of the main weather factors that indicate climate change and is also one of the main sources of water. A deeper understanding of rainfall variability in Bangladesh can help to address the demand and supply gap of water at different times over the year. Figure 5 points out that a major portion of rainfall takes place in JJA (June, July, and August) and SON (September, October, and November) and remains in DJF (December, January, and February) and MAM (March, April, and May). It is assumed that JJA and SON are called 'wet', whereas DJF and MAM are 'dry' seasons. Figure 6 postulates that the wet and dry seasons are getting more and less rainfall since the mean rainfall in wet and dry seasons increases and decreases by 119.94 and 84.13 mm, respectively, in 2011-2015 compared to 1991-1995.  1989-1990 1994-1995 1999-2000 2004-2005 2008-2009 -1995, 1996-2000, 2001-2005, 2006-2010, and 2011-2015 (Source: World Bank Group Climate Change Knowledge Portal, https://climateknowledgeportal.worldbank.org/download-data (accessed 21 April 2020)). The gap between minimum and maximum rainfall in each season is extending over the period 8 that indicates a rise in intensity. It also demonstrates that precipitation is going less evenly distributed over the season and more erratic day by day.

Flood
Flood is a rainfall-driven climate-induced problem in Bangladesh. The probability of occurrence of rainfall-driven climate hazards (e.g., flood) is getting higher and higher due to changes in the pattern of rainfall that adversely affects the hydrological process. To determine the impact of floods and to address, it needs to understand the pattern and intensity of floods. Table 4 shows the flood-affected areas (in km 2 ) in a 5-year time span from 1988 to 2018. It can be assumed that if flood-affected areas crossed 30,000 km 2 or 20% of total areas of the country, it can be considered extreme floods. With this indication, Table 4 exhibits that the country faces extreme floods in every 5-year time period 9 except 1999-2003 and 2009-2013 after 1987. Moreover, if we consider the maximum value as an extreme flood indicator with the assumption, it attributes that country faces extreme floods at least once in every period. Table 4

Salinity level
Different levels of salinity in the soil, surface water, and groundwater are tolerable for vegetation, crops, and aquatic communities. For instance, soil salinity level of ,4 dS/m 10 is within the range of tolerance for crops and vegetation (SRDI 2001). For surface water, an electric conductivity of 5 dS/m considers as bearable for freshwater vegetation and aquatic communities as per the national standard of Bangladesh. However, the salinity level in groundwater is 600 mg. Chloride is acceptable but ESCAP/UN (1987) recommended 1,000 mg/l of chloride equal to a threshold level of 2 dS/m. Table 5 shows the mean and median soil salinity levels are 8.656 and 6 dS/m, respectively. These values indicate that soil is not suitable for crops and vegetation. In a large part of the region, people cannot cultivate their land due to salinity. The difference between the maximum and minimum soil salinity levels (18.67 dS/m) points out that a major part of land shifted into barren and will add more in the coming future. However, the mean surface water salinity level (4.65 dS/m) is within the tolerant level (5 dS/m) for freshwater vegetation and marine communities. If we add the standard deviation (3.65 dS/m) with this mean value that crosses the limit that indicates that some places where surface water is not safe for freshwater vegetation, and maritime society.
However, the value of the mean groundwater salinity level (2.78 dS/m) is beyond the tolerant level (2 dS/m) for vegetation and aquatic communities. The gap between the highest and lowest level (11.97 dS/m) demonstrates that the salinity level in groundwater is high in some places that severely damaging the quality of groundwater and making it unusable. This high level of salinity penetrates to a lower level or no salinity and converts the lower level into a high level and no salinity to lower-level salinity. In this way, it is possible to facilitate the salinity problem and it will go beyond the coastal region of the country. In this context, the analysis of the different years in terms of areas and class of salinity will give a clearer picture and provide evidence about the overall situation of salinity regarding the conversion of low salinity areas into high salinity areas, and the growth of salt-affected areas in coastal Bangladesh. Table 6 Figure 6 points out the salinity areas in terms of salinity classes in the coastal setting of Bangladesh. The first issue is that the salinity area is increasing. For instance, the mean area of salinity was 15,960 ha in 1973 and reached 18,250 ha in 2009, while the total area increased from 287,370 to 328,430 ha in the same year for Salinity Class (S1). On the other hand, the mean level of S2 reduced from 23,690 to 15,230 ha in 2009. It can reasonably assume that S1 salinity areas convert into S2. To get a more specific understanding, we want to see two more Salinity Classes S3 and S4. Figure 6 shows that the mean level of S3 increased from 4,350 to 19,590 ha, while the range 11 of salinity-affected areas increased from 34,500 ha in 1973 to 69,720 ha in 2009. More importantly, the areas of S3 increased by more than four times in 2009 compared to 1973. Both indicators in S3 demonstrate that salinity-affected areas are converting from lower levels to higher levels of salinity. However, the mean and total areas were 2,210 and 39,900 ha but rise to 4840 and   1987-1991 1992-1998 a 1999-2003 2004-2008 2009-2013 2014-2018  87,140 ha in 2000, respectively, for S4. In the next 9 years, the mean level and total areas increased by 1,150 and 14,780 ha, respectively, in 2009 which highlights the high intensity. Moreover, maximum areas of S3 and S4 rise from 34,500 and 14,000 ha to 69,720 and 30,570 ha in 2009 compared to 1973. It indicates that areas of these salinity levels are more than doubled during that time (from 1973 to 2009). Therefore, some factors are responsible for increasing this salt-affected area with this high intensity. Tidal fluctuation is one of them.

Tidal movement
Tidal movement is the fluctuation between high and low levels of water in the coastal region. To analyze data regarding the tidal movement in coastal settings, it can be categorized as regular tidal movement (between 0.3 and 1 m), high tidal movement (1-2 m), and very high tidal movement (.2 m). The latter two are most responsible for causing saltwater intrusion and floods that deter water supply. To calculate the tidal movement in coastal Bangladesh, 0.3 m is considered as the threshold value. Out of 147 subdistricts, 63 fall in the range between 2.38 and 3.07 m tidal movement. Figure 7 presents a histogram of tidal movement in coastal Bangladesh. It demonstrates that the major part (42.85%) of coastal Bangladesh is in the very high tidal movement category. However, 46 and 38 out of 147 subdistricts fall in the high tidal and regular tidal movement category, respectively. The latter is a less responsible factor for water crisis than the former, but climate change can change the equation and turn the regular tidal movement into high and high can shift to a very high tidal movement area category.
To analyze historical data regarding temperature in Bangladesh, this study uses central tendency and variability that give a better understanding regarding monthly temperature over the year. It also gives a clear picture concerning the distribution of temperature over the period. To compare the monthly temperature over the year, data divide into five periods that is the first (1991-1995), second (1996)(1997)(1998)(1999)(2000), third (2001)(2002)(2003)(2004)(2005), fourth (2006)(2007)(2008)(2009)(2010), and fifth (2011-2015) period. Table 7 reveals that temperature is not in a similar pattern throughout 1991-2015. The mean and median temperature levels escalate by 0.91 and 0.953°C in the fourth period compared to the first, respectively. At the same time, mean and median temperature levels decline by 0.88 and 1.08 in 2011-2015 compared to 2006-2010. The temperature variation is increasing since the difference between the minimum and maximum increases by 1. 5427 and 0.1676 in 2011-2015 compared to 2006-2010 and 1991-1995, respectively. The water demand facilitates since temperature variation and average temperature increase.

Cyclone
Due to global warming, cyclones are more intense and frequent, and are directly and indirectly responsible for the water crisis. First, it damages the water management and supplies infrastructure that constraint the water supply. Secondly, it pollutes water (e.g., facilitating saltwater intrusion through flooding and storm surges). This intensity of cyclones is held responsible for damaging water supply, management infrastructure, and polluting water that cause scarcity of water.
The result in Table 8 shows that the mean and median wind speeds rise by 22.16 and 21.21% in 2015-2020 compared to 1988-1995. It is reasonably assumed that when the maximum wind speed is above 150 km/h, it considers as an extreme level cyclone. With this assumption, we can observe that during the first 14 years, five extreme-level tropical cyclones take place where the average speed is 192.03 km/h. During the last 14 years, five cyclones hit the country where the average wind speed is 223.1 km/h. It implies that the intensity of cyclones increased significantly in the latter compared to the earlier.

Sea-level rise
Several contributing factors are responsible for changes in sea level that include the expansion of seawater by melting glaciers and ice sheets due to global warming. This study includes four stations such as Hiron Point, Khepupara, Cox's Bazar, and Char Changa to measure sea level. Table 9 shows that the mean and median sea levels rise by 73.55 and 80 cm in 1999-2003 compared to 1983-1986, respectively, at Hiron Point. However, Table 10 displays the central tendency and variability  of ocean levels in the Khepupapra station. The result shows that the sea level heightens following the preceding period. For instance, the mean sea level increased by 58.2, 70, and 90.05 cm in 1984-1988, 1989-1994, and 1994-1999, respectively, connected to 1979-1983. Moreover, the median sea level goes up by 298 cm in 1994-1999 compared to 1979-1983. Table 11 exposes that the average sea level increased by 16.8 cm in 1995-2000 compared to 1979-1983, while the median sea level adjusted by 17 cm at Cox's Bazar station. This station is steadier than the other station concerning the mean and median of sea-level rise. At the same station and time, the maximum water level in the ocean enhanced by 17 cm in 1995-2000 related to 1979-1983, while standard deviation positively evolves by 0.34. Table 12 presents the central tendency and variability of sea level at the Char Changa station for 1979 and 2000. It reveals that the mean and median sea levels go up by 169.2 and 103 cm in 1995-2000 compared to 1979-1983. It indicates that the sea-level rise in this station is more intense than in the other station. This rise facilitates due to climate change in the coming future.

MK test
The MK test statistics indicate that water demand factors GDP has an increasing monotonic trend, while population and urban population growth have a decreasing trend. However, the MK test statistics in Table 13 present that water supply sources such as rainfall in JJA and transboundary river water do not show a trend but rise, while rainfall in DJF, MAM, and SON is shrinking. At the same time, groundwater depletion is also rising. The climatic disasters such as flood-affected  1979-1983 1984-1988 1989-1994 1994-1999 1983-1986 1987-1990 1991-1994 1995-1998 1999-2003 1979-1983 1984-1989 1990-1994 1995-2000 Mean areas, SLR (Hiron Point, Cox's Bazar), cyclone, salt-affected areas, and salinity class (S1, S3, and S4) do not have a trend but rise. At the same time, SLR (Khepupara, Char Changa) has a trend and rising, while S2 does not have a trend and shrinking. The S statistics for Khepupara is stronger than the other station.  1979-1983 1984-1989 1990-1994 1995-2000

DISCUSSION
This research contributes to the present understanding of how water supply sources respond to the demand and climateinduced extreme limit supply and affect water resources in Bangladesh. Water demand in Bangladesh is increasing due to population, economic, and urban population growth. The results are similar to Ahmed et al. (2015), Asaduzzaman et al. (2010), Chowdhury (2010), and World Bank (2005), who found that water resources in Bangladesh have been facing some challenges that include but are not limited to (a) rise in pollution due to unplanned urbanization and industrial growth; (b) adverse impact of climate change that roughly affects 160 million people directly and indirectly; and (c) growing demand of water that creates pressure on groundwater. The factors such as economic growth, population, and urban population are responsible for increasing water demand and putting more pressure on water resources. The MK test result (Table 13) expresses that GDP growth is increasing trend and statistically significant. This result leads to accepting the alternative hypothesis. However, population and urban population are still growing (Figure 4) but decreasing trend (Table 13) and statistically significant. Therefore, it rejects the null hypothesis. These three factors are constantly laying down more stress on water supply systems. The finding is similar to WWAP/UN Water (2018), who argued that rising water demand follows economic and population growth. The finding enriches the argument of Liu et al. 2017. They mentioned 'population growth, economic development, and dietary shift (toward more animal products) have resulted in ever increasing water demand, and consequently pressures on water resources' (Liu et al. 2017, p. 545).
With an increase in water demand, supply from different water sources is decreasing, unstable, and unreliable in Bangladesh. As a result, the groundwater has been overexploited since its depletion almost doubled (99.66%) from 1989-1990 to 2008-2009 in Bangladesh. This depletion intensifies the pressure on aquifers and ultimately provides less time to reduce the gap between withdrawal and recharge, limiting this source's water supply. This result is similar to Collins (2008), Hidalgo et al. (2009), andPiao et al. (2010), who investigated the adverse impact on aquifer recharge in the Karst region and found that the groundwater level has been going down.
Concerning the water supply from rainfall, the result indicates that rainfall has been more erratic in recent years than in the past. For example, the range 12 of rainfall in the JJA 13 and DJF 14 seasons increased by 102.706% and shrank by 55.66% in 2011-2015 compared to 1991-1995. The result advances literature such as Goswami et al. (2006), Lau & Wu (2007), and Rajeevan et al. (2008), who found a pattern of change in rainfall that includes an increase in heavy precipitation and a decrease in moderate rainfall. Moreover, this finding also enriches the literature by Lu & Fu (2010) and Chen et al. (2012), where they found a significant rise in interannual rainfall variability and precipitation will be more extremes. However, the temperature increased by 0.5745°C in 2001-2010 compared to 1991-2000. This result is consistent with Frich et al. (2002), who observed that every decade has been hotter than the previous decade since the later part of the 20th century. In addition, the result is also similar to IPCC (2013), which found that the average global air temperature surged by 0.85°C (0.65-1.06°C) over the period 1880-2012.
In terms of the salinity problem in coastal Bangladesh, salt-affected areas enhanced by 26.733% from 1973 to 2009. This result is consistent as well as contrary in terms of intensity to Shrivastava & Kumar (2014), who argued that salinized areas are increasing at the rate of 10% per year due to low precipitation, poor cultural practices, high surface evaporation, and irrigation with saline water. If the rate 15 of salt-affected areas remains the same in Bangladesh, it will rise to 1,338,598.298 ha in 2045. This is 17.222% of the total arable land 16 in Bangladesh. At the same time, salt-affected areas of S3 and S4 incremented by 329.05 and 170.532% in 2009 compared to 1973, respectively. The result is consistent with Jamil et al. (2011) who estimated that more than 50% of arable land will be salinized by 2050. The result also strengthens the salinity study by SRDI (2010) and Rasel et al. (2013). The high salinity-affected areas S3 and S4 increased by 354.57 and 155.43% in 2009 compared to 1973. This finding enriches the literature of Shrivastava and Kumar (2015). They observed that 20% of total cultivated and 33% of irrigated agricultural land areas are affected by high salinity worldwide. The result indicates that salt-affected areas 12 Difference between maximum and minimum. 13 Considers as monsoon/rainy season. 14 Considers as dry season. 15 Salt-affected areas in 2009 compared to 1973. 16 As per World Bank data, total arable land was 7,772,300 ha in 2018. https://data.worldbank.org/indicator/AG.LND.ARBL.HA?locations¼BD.
will not be limited to coastal areas. Moreover, the intensity of growth will not same due to climate change impact and geographical boundary.
For the transboundary river water flow, climate change is a concern for transboundary river water flow (Milman et al. 2013) and has a significant impact on the water supply from those rivers. Moreover, Čerkasova et al. (2018) conducted a study on the Vilija River basin using climate scenarios based on IPCC AR5 RCP4.5 model (Collins et al. 2013). They found that yearly transboundary river water discharge increased to 53.7%, where expected rise to 47.6%. However, the present finding concerning transboundary water flow is contrary to Collins et al. (2013) since mean water flows decreased by 1.212, 7.329, and7.056% in 2018 compared to 2012 in Brahmaputra, Meghna, andSoutheastern Basin, respectively. The average and intensity of flood increased since mean and maximum flood-affected areas increased by 93. 357 and 107.98% in 2014-2018 compared to 2009-2013. This is in line with the previous finding. Tabari (2020) found that flood intensity enhances in all climate regimes. However, the intensity of cyclone facilitates since maximum wind speed facilitates by 19.17% in 2015-2020 contrasted to 1988-1995, while the range positively adjusted with 21.03%. The output (Table 8) is in agreement with previous studies by Kossin et al. (2020), Emanuel (2000), and Kossin et al. (2013).
Nevertheless, the sea level has risen with different intensities in Bangladesh. Sequentially, the mean sea level increased by 8.02 and 0.795 cm per year during 1979-2000 in Char Changa and Cox's Bazar. Moreover, the rate is more intense in Khepupara (16.3624 cm) and Hiron Point (3.667 cm). The result is consistent with Douglas (1991), Cazenave & Llovel (2010), Golledge (2019), and Rahmstorf (2007) but indicates different intensities. Moreover, this result reinforces the finding of Glennon (2017), who estimated that 3 feet of the sea level will rise by 2050 but with a different magnitude. For example, if the sea level continues, seawater's height will rise by 2 m by 2013 in Khepupara, 2025 in Char Changa, 2055 in Hiron Point, and 2252 in Cox's Bazar.
This study provides insight to understand further the link between factors for rising water demand, water supply sources, and extreme events that limit water supply in Bangladesh. It will help decision-makers take action for climate change adaptation, displacement, water crisis, and economic stability. Furthermore, it can be an ideal study to anticipate the future scenario for other climate-vulnerable countries (e.g., Philippines, Myanmar, Vietnam, India, Indonesia, and Pakistan) concerning the linkage between climatic extreme events and water. For instance, the salinity level and the intensity of the cyclone are not only increasing in coastal Bangladesh but also possible to extend in the Indian coastal region. It will help them prepare for adaptation since they understand the intensity of extremes and the interlinking sectors. It also urges to find out alternative sources of water, efficient and effective use of limited water resources. Finally, it will draw the attention of policymakers, water managers, and academics for further investigation into coping with extreme climate events that affect water sources, displaced people, and damage economic and environmental resources.

CONCLUDING REMARKS
This study and its underlying research objective are to find out factors that contribute to rising water demand, the response of water sources to fulfill this demand, and the impact of climatic extreme events on water resources in Bangladesh. Economic development, urbanization, and population growth cause more demand for water. The total population and urban population have grown 1.35 and 3.70% per year over the last 21 years in Bangladesh, while the growth rate of GDP was more than 6% over the last 10 years. The test result in Table 13 shows that there is an increasing trend in GDP growth (131.000), while decreasing trend in population (À181.000) and urban population (À164.000) growth. Accept the alternative hypothesis for these factors. These are the major responsible factors for growing water demand in the country. For this growing demand, the supply side is under stress to fulfill this demand.
Water supply from different river basins decreased periodically. For example, the mean level was 19.37, 12.32, and 12.69 in 2012 of Brahmaputra, Meghna, and Southeastern Basin, respectively, and goes down by 0.23, 1.19, and 0.89 m in 2018. The alternative hypothesis accepts for the entire river basins. It means that there is no trend that implies water supply from this source is unstable. However, the water supply through rainfall is not stable. From 1991 to 1995, the mean level rainfall was 57.12 mm in DJF, 444.57 mm in MAM, 1,257.67 mm in JJA, and 518.20 mm in SON, andit reached 31.99, 385.58, 1,476.08, and419.79 mm, respectively, in 2011-2015. The MK test statistics result in Table 13 show a similar pattern that rainfall surges in JJA (24.000), while goes down in DJF (À74.000), MAM (À32.000), and SON (À58.000). However, groundwater is another major water supplier and its depletion rate doubled within two decades ). The average depletion was 2.455 km 3 in 1989-1990 and rises to 4.90 km 3 in 2008-2009. The MK test result indicates no trend in groundwater depletion but increasing in every region of the country that includes NC (6.000), NE (6.000), NW (6.000), SC (8.000), SE (7.000), and SW (8.000). This result in Table 13 exposes that the intensity of groundwater depletion is more in southern than the northern part of the country.
The intensity of climatic disasters' impact on water resources is increasing. From 1988 to 1993, mean and median floodaffected areas were 37,094 and 28,600 km 2 and reached 45,738 and 47,200 km 2 in 2014-2018. For cyclone, the mean and median levels of the maximum wind speed were 135.51 and 122.1 km in 1988-1995 which go up by 30.03 and 25.9 km in 2015-2020, respectively. In addition, salt-affected areas and the areas of different salinity classes are also increasing. For example, salt-affected areas increased by 222,810 ha in 2009 compared to 1973, while the mean areas of S4 rise from 2,220 ha in 1973 to 5,990 ha in 2009. At the same time, the average area S3 also increased by 15,240 ha. The MK test result exhibited in Table 13 also indicates the rise of these climatic disasters that include cyclones (19.000), salt-affected areas (3.000), and salinity class (S1¼1.000, S3¼3.000, and S4¼3.000).
The sea level rises in the different stations in Bangladesh. At Hiron Point, the mean level of sea level was 7,048.25 cm in 1983-1986 and reached 7,121.8 cm in 1999-2003. This station does not have a rising trend but rise (40.000). Moreover, at Khepupara (119.000) and Char Changa (70.000), the result in Table 13 exposes that the height of sea level has an increasing trend and strong compared to other factors. Therefore, these extreme events are directly and indirectly affected the demand and supply of water as well as affected total water resources systems in Bangladesh.
This study unequivocally presents the five issues for future programs and actions. First, groundwater is depleting fast, rainfall is more erratic, 17 and water supply from the riverine system is not stable. To address these issues, it could be better to use rainwater for domestic, industrial, and agricultural purposes through rainwater harvesting since the country is one of the most rain-intensive countries (average rainfall of 2,400 mm/year). Rainwater harvesting can be promoted and on the priority list to address the water demand-supply gap. This program is more applicable in the southern part of the country since groundwater depletion is more potent than in the northern part. Previous studies found that rainwater harvesting is a sustainable approach to govern rainwater and address economic poverty and water crisis in the southern part of Bangladesh (Islam 2018(Islam , 2019. Secondly, climatic disasters are rising (some have a monotonic rise), affecting water resources, people, the economy, and the environment. It is high time to combine water, climate change, economy, and people as a whole, not each of these as an isolated issue. Moreover, water and climate change issues should not confine to the ministry of water resources and the environment.
Thirdly, decisions are based on data and science. Lack of updated data constraint study leads to a wrong decision regarding water resources management and climate change adaptation. For example, in this study, the latest data for SLR is 2003, while salinity intrusion and groundwater depletion are 2009. Even some stations (e.g., Khepupara, Cox's Bazar, and Char Changa) do not have data for 2003. Recent data could contribute more conclusive findings significantly. The country needs to collect and upload updated data (for the whole country) in the concerned ministry and departmental website to help make better decisions for socioeconomic development, water resources management, and climate change adaptation. Fourthly, more diplomatic effort needs to bring stability in supplying water through the transboundary river. Because agricultural productivity in Bangladesh largely relies on the transboundary river's water supply. The fifth issue is efficiency. As there is a widening gap between demand and supply of water and climatic extreme will facilitate this gap, the efficient use of limited supply could limit this gap. This study urges to boost the water-use efficiency in Bangladesh. This can be done by promoting less waterintensive foods and diversifying crop production that requires less water to produce. For example, meat (beef) and rice are two of the most water-intensive foods and crops. Both of these agricultural products are popular in the country. Since the country faces a demand-supply gap for water and climatic extreme will facilitate it, there is no time to waste to bring efficiency in water use.